Research Handbook On The Law Of Virtual And Augmented Reality 1786438585, 9781786438584

Table of contents :
Front Matter
......Page 2
Copyright......Page 5
Contents......Page 6
List of figures......Page 8
List of contributors......Page 9
Preface......Page 15
PART I: INTRODUCTION TO THE LAW OF VIRTUAL AND AUGMENTED REALITY......Page 20
1 The law of virtual reality and increasingly smart virtual avatars......Page 21
2 Starting up in virtual reality: examining virtual reality as a space for innovation......Page 63
3 Virtual rule of law......Page 83
4 Mixed reality: how the laws of virtual worlds govern everyday life......Page 122
PART II: INTELLECTUAL PROPERTY LAW......Page 172
5 Virtual copyright......Page 173
6 Trademark law and the right of publicity in augmented reality......Page 211
7 Trade dress in virtual worlds......Page 235
PART III: ISSUES OF CONSTITUTIONAL AND CRIMINAL LAW......Page 260
8 The First Amendment, video games, and virtual reality training......Page 261
9 Virtual reality, haptics, and First Amendment protection for sexual sensation......Page 294
10 Augmented and virtual reality, freedom of expression, and the personalization of public space......Page 323
11 Beyond unauthorized access: laws of virtual reality hacking......Page 359
12 The law and ethics of virtual sexual assault......Page 382
13 Criminal liability for intellectual property offenses of artificially intelligent entities in virtual and augmented reality environments......Page 408
PART IV: APPLYING THE LAW TO DIFFERENT APPLICATIONS OFVIRTUAL AND AUGMENTED REALITY......Page 439
14 Advertising legal issues in virtual and augmented reality......Page 440
15 Data privacy legal issues in virtual and augmented reality advertising......Page 490
16 Reordering the chaos of the virtual arena: harmonizing law and framing collective bargaining for avatar actors and digital athletes......Page 532
PART V: CONTRACT, PROPERTY LAW, AND JURISDICTION......Page 586
17 Property rights in virtual and augmented reality: Second Life versus Pokémon Go......Page 587
18 Freedom of contract in augmented reality......Page 623
19 Law and property in virtual worlds......Page 655
20 Legal jurisdiction and the deterritorialization of social life......Page 681
Index......Page 704

Citation preview

RESEARCH HANDBOOK ON THE LAW OF VIRTUAL AND AUGMENTED REALITY

BARFIELD PRINT.indd 1

13/11/2018 11:52

Woody Barfield dedicates this book to his daughter Jessica, his best friend Bailey, his former students, and the future generations which will spend significant amounts of time in virtual and augmented reality. As always, his wonderful parents are thanked for their support, affection, and guidance. Marc Jonathan Blitz dedicates this book to his parents, Bernard and Alma; his brothers, Jeff and Andrew; and his wife, Holly, and daughters, Rachel and Gabrielle. He thanks his students and colleagues for the many conversations that have helped him think about how to address the challenges of applying First Amendment and other law to emerging technologies. Both editors greatly acknowledge Stephen Gutierrez, acquisitions editor at Edward Elgar Publishing for his early support of the project, and throughout. Further, we thank Erin McVicar, Sue Sharp, and Laura Mann from Edward Elgar Publishing for their assistance with the project, and the production team for their work to produce a publication-quality work. We especially thank the authors for producing outstanding and forward-looking chapters and for their professionalism as they responded to our many queries for more information and for meeting deadlines. We also thank our colleagues for the many conversations we had with them on law and virtual and augmented reality and for their contribution to this emerging field of law.

BARFIELD PRINT.indd 2

13/11/2018 11:52

Research Handbook on the Law of Virtual and Augmented Reality

Edited by

Woodrow Barfield Professor Emeritus, USA Editor, Virtual Reality Journal, USA

Marc Jonathan Blitz Alan Joseph Bennett Professor of Law, Oklahoma City University School of Law, USA

Cheltenham, UK • Northampton, MA, USA

BARFIELD PRINT.indd 3

13/11/2018 11:52

© The Editors and Contributors Severally 2018 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical or photocopying, recording, or otherwise without the prior permission of the publisher. Published by Edward Elgar Publishing Limited The Lypiatts 15 Lansdown Road Cheltenham Glos GL50 2JA UK Edward Elgar Publishing, Inc. William Pratt House 9 Dewey Court Northampton Massachusetts 01060 USA

A catalogue record for this book is available from the British Library Library of Congress Control Number: 2018946021 This book is available electronically in the Law subject collection DOI 10.4337/9781786438591

ISBN 978 1 78643 858 4 (cased) ISBN 978 1 78643 859 1 (eBook)

06

Typeset by Servis Filmsetting Ltd, Stockport, Cheshire

BARFIELD PRINT.indd 4

13/11/2018 11:52

Contents List of figuresvii List of contributorsviii Prefacexiv PART I INTRODUCTION TO THE LAW OF VIRTUAL AND AUGMENTED REALITY  1 The law of virtual reality and increasingly smart virtual avatars Woodrow Barfield and Alexander Williams

2

 2  Starting up in virtual reality: examining virtual reality as a space for innovation44 Crystal Nwaneri  3 Virtual rule of law Michael Risch

64

 4 Mixed reality: how the laws of virtual worlds govern everyday life Josh A.T. Fairfield

103

PART II  INTELLECTUAL PROPERTY LAW  5 Virtual copyright Michael Risch and Jack Russo

154

 6 Trademark law and the right of publicity in augmented reality Brian D. Wassom, Amber M. Underhill, and Andrew L. Rossow

192

 7 Trade dress in virtual worlds Jack Russo

216

PART III  ISSUES OF CONSTITUTIONAL AND CRIMINAL LAW  8 The First Amendment, video games, and virtual reality training Marc Jonathan Blitz

242

 9 Virtual reality, haptics, and First Amendment protection for sexual sensation Brooke Lewis

275

10  Augmented and virtual reality, freedom of expression, and the personalization of public space Marc Jonathan Blitz

304

v

BARFIELD PRINT.indd 5

13/11/2018 11:52

vi  Research handbook on the law of virtual and augmented reality 11 Beyond unauthorized access: laws of virtual reality hacking Gilad Yadin

340

12 The law and ethics of virtual sexual assault John Danaher

363

13  Criminal liability for intellectual property offenses of artificially intelligent entities in virtual and augmented reality environments Gabriel Hallevy

389

PART IV  APPLYING THE LAW TO DIFFERENT APPLICATIONS OF VIRTUAL AND AUGMENTED REALITY 14 Advertising legal issues in virtual and augmented reality S.J. Blodgett-Ford, Woodrow Barfield, and Alexander Williams

421

15 Data privacy legal issues in virtual and augmented reality advertising S.J. Blodgett-Ford and Mirjam Supponen

471

Reordering the chaos of the virtual arena: harmonizing law and framing 16  collective bargaining for avatar actors and digital athletes Jon M. Garon

513

PART V  CONTRACT, PROPERTY LAW, AND JURISDICTION Property rights in virtual and augmented reality: Second Life versus Pokémon 17  Go Hannah YeeFen Lim

568

18 Freedom of contract in augmented reality Scott R. Peppet

604

19 Law and property in virtual worlds Wian Erlank

636

20 Legal jurisdiction and the deterritorialization of social life Paul Schiff Berman

662

Index685

BARFIELD PRINT.indd 6

13/11/2018 11:52

Figures  3.1 Virtual concert in Second Life 66  4.1 Milgram’s Reality-Virtuality (RV) Continuum 112  5.1 Human hand simulation 158  5.2 The Sensorama Simulator 159  5.3 Gesture-based control 160  5.4 Data input glove 161  5.5 MeshWerks wire model with CGI additions 170  7.1 “Pinch” bottle and packaging 222  7.2 Taco Cabana restaurant 228  7.3 Qualitex’s “green-gold” dry-cleaning pad 230  7.4 Wind-resistant springloaded display stands 231  7.5 Geometric uniform designs 233  7.6 “Dismaland” by Banksy 236  7.7 “FUCT” t-shirt designs 237 12.1 Classifying virtual sexual assault 368 14.1 A drawing from Magic Leap’s published patent application 449 14.2  The images in dispute for the Vanna White right of publicity case, taken 454 from the Exhibits in the public court filings 14.3  Lindsay Lohan photo and video game marketing material image of “Lacey 456 Jonas”, triggering a right of publicity action 15.1  An augmented reality friendly monster character bursting out of the cereal boxes in the grocery store aisle for a young child who touches a particular box491 15.2 A child finds an augmented reality character 492

vii

BARFIELD PRINT.indd 7

13/11/2018 11:52

Contributors Woodrow Barfield served as Professor of Engineering at the University of Washington’s Department of Industrial and Systems Engineering with a courtesy appointment in the Departments of Environmental Health and Civil Engineering. He holds a PhD, a JD, and an LLM; is currently Associate Editor for the Virtual Reality Journal, is a former senior editor of the MIT journal Presence: Teleoperators and Virtual Environments, and was on the editorial board of the Washington Journal of Law Technology & Arts. Further, he was Visiting Professor in computer science at University College, London, and was on the program committee which established the first conference on wearable computers and augmented reality. He also edited the first academic books on the same topics. He is a recipient of the National Science Foundation’s Presidential Young Investigator Award and was an External Fellow for Stanford’s Center for Internet and Society. His writings, which include more than 300 scholarly papers, discuss the law of virtual and augmented reality, and how the law applies to artificial intelligence and to humans enhanced with implanted technologies such as neuroprosthesis and brain-computer interfaces. His most recent books are Cyber Humans: Our Future with Machines and the edited collection Fundamentals of Wearable Computers and Augmented Reality (2nd edition), and he is coediting, with Professor Ugo Parallo, the forthcoming Research Handbook on Law and Artificial Intelligence. In addition, he frequently lectures on systems of the future at Duke University and publishes in engineering journals and law reviews, actively reviews for several journals, serves on various program committees, and gives invited keynote talks and lectures. Paul Schiff Berman is Walter S. Cox Professor of Law, George Washington University. He is one of the world’s foremost theorists on the effect of globalization on the interactions among legal systems and is the author of more than 60 scholarly works, including  Global Legal  Pluralism: A Jurisprudence of Law beyond Borders, published by Cambridge University Press in 2012. He was also among the first legal scholars to focus on legal issues regarding online activity, and he is coauthor of one of the leading casebooks in the field. In addition to his scholarly work, Professor Berman has extensive experience in university and law school administration, having served as Vice Provost for Online Education and Academic Innovation at The George Washington University, Dean of the George Washington University Law School, and Dean of the Sandra Day O’Connor College of Law at Arizona State University. Professor Berman has previously served as the Jesse Root Professor of Law at the University of Connecticut School of Law and was a Visiting Professor and Visiting Research Scholar at Princeton University in the Program in Law and Public Affairs. Since 2016 he has served as a Distinguished Visiting Professor at the Queen Mary University of London. He has served on the Organizing Committee of the Association for the Study of Law, Culture, and the Humanities and was cochair of the International Law and Technology Interest Group of the American Society of International Law from 2014 to 2016. Prior to academia, he clerked for Chief Judge Harry T. Edwards of the United States Court of Appeals for the District viii

BARFIELD PRINT.indd 8

13/11/2018 11:52

Contributors  ix of Columbia Circuit and for Justice Ruth Bader Ginsburg of the Supreme Court of the United States. Marc Jonathan Blitz is Alan Joseph Bennett Professor of Law at Oklahoma City University School of Law. He is a graduate of Harvard University and has a law degree and a PhD in Political Science from the University of Chicago. His scholarship focuses on constitutional protection for freedom of thought, freedom of expression, and privacy law—and especially on how each of these areas applies to emerging technologies. He has written articles on how First Amendment free speech doctrine and/or Fourth Amendment doctrine on search and seizure law should apply to government use, or regulation, of, public video surveillance (using fixed CCTV camera systems, or drone-mounted cameras, or policy body cameras), biometric identification methods, and electronic information systems. His articles also include analyses of the First Amendment status of video gameplay and other uses of virtual reality technology, and his recent work has focused on the intersection of constitutional law and neuroscience and psychology. Prior to entering academia, he worked as an attorney at Wilmer Cutler Pickering Hale and Dorr in Washington DC, where his work focused on telecommunications, privacy law, computer law, intellectual property, constitutional law, and antiterrorism security measures. While working in Washington, he was also one of the reporters for the Constitution Project’s Liberty and Security initiative and was one of the drafters of its Guidelines on Public Video Surveillance. John Danaher is Lecturer at the National University of Ireland School of Law, Galway, Ireland. He holds a BCL from University College Cork (2006), an LLM from Trinity College Dublin (2007), and a PhD from University College Cork (2011). He was lecturer in law at Keele University in the UK from 2011 until 2014. He joined NUI Galway in July 2014. His research interests lie, broadly, in the areas of philosophy of law and emerging technologies and law. In the past, he has published articles on human enhancement, brain-based lie detection, the philosophy of punishment, and artificial intelligence. He maintains a blog called Philosophical Disquisitions, and he also writes for the Institute for Ethics and Emerging Technologies. Wian Erlank is Associate Professor of Law, North-West University, Potchefstroom, South Africa. His current and previous lecturing experience takes in the areas of property law, land reform, research methodology, legal skills, and legal writing. He was involved with the creation of the first compulsory three-year language and communication course as part of a legal degree. He is a member of many committees, including Teaching, Re-Curriculation, Research, Ethics (as chairperson), the Student Disciplinary Committee, and the Work Group on Teaching and Learning Technology. He is a board member of the Association of Law Property and Society (ALPS) and an advisory board member of the Young Property Lawyers’ Forum (YPLF), and the South African Literary Awards board (SALA). Additionally, he is a founding member and treasurer of the South African Association of Intellectual Property Law and Information Technology Law Teachers and Researchers (AIPITL). Finally, he is a member of the Academy of Science for South Africa’s Committee on Scholarly Publishing in South Africa (CSPiSA), a member of the Scholarly Publisher Editors’ Forum, and a Fellow of the European Law Institute (ELI).

BARFIELD PRINT.indd 9

13/11/2018 11:52

x  Research handbook on the law of virtual and augmented reality Josh A.T. Fairfield is Professor of Law at Washington and Lee University. He is an internationally recognized law and technology scholar of digital property, electronic contract, big data privacy, and virtual communities. He has published articles in top law journals, as well as the New York Times, Forbes, and the Financial Times. He is a Fulbright grant recipient and member of the American Law Institute. S.J. Blodgett-Ford is Member and Chief Privacy Officer, GTC Law Group PC, and Adjunct Professor, Boston College Law School (Privacy Law and Mobile App Development),  CIPP/US. Blodgett-Ford focuses on intellectual property strategy and transactions, mergers and acquisitions, and data privacy and security, and was previously General Counsel, Tetris Online, Inc. and Senior Manager, Intellectual Property, Nintendo of America Inc. Blodgett-Ford taught advertising law in the University of Washington IP LLM program. Jon M. Garon is Dean of Nova Southeastern University Shepard Broad College of Law. He is a nationally recognized authority on technology law and intellectual property, particularly copyright law, entertainment, and information privacy. Prior to joining Nova Southeastern University in 2014, Garon was the inaugural director of the Northern Kentucky University Salmon P. Chase College of Law, Law + Informatics Institute from 2011 to 2014, and has been active in legal education since 1991. He is the author of four books and numerous book chapters and articles, including The Entrepreneur’s Intellectual Property & Business Handbook (Manegiare Publications 2018); Pop Culture Business Handbook for Cons and Festivals (Manegiare Publications 2017); The Independent Filmmaker’s Law & Business Guide to Financing, Shooting, and Distributing Independent and Digital Films (A Cappella Books, 2d Ed. 2009); Own It—The Law & Business Guide to Launching a New Business Through Innovation, Exclusivity and Relevance (Carolina Academic Press 2007); and Entertainment Law & Practice (2nd edition 2014 Carolina Academic Press). Gabriel Hallevy is Full Professor at the Faculty of Law, Ono Academic College, Israel. He lectures in criminal law, criminal justice, evidence law, corporation law, conflict of laws, bankruptcy law, and game theory in law. He has published more than 30 books (as to 2018) in Israel, the US, and Europe, most of which are cited by the Israeli Supreme Court, and a few dozen articles in the US, Australia, Europe, India, and Israel. He earned his LLB and LLM magna cum laude from Tel-Aviv University, and his PhD summa cum laude from Haifa University. The Knesset, the Israeli Parliament, granted him a special honorary prize for research in criminal law. In 2013 he was chosen as one of the 40 most promising Israelis under the age of 40 (“Top 40 under 40”) by Globes magazine in Israel. He is a long-distance runner, holds a pilot’s license, and speaks Hebrew, English, French, and German. Brooke Lewis holds a JD degree from Rutgers Law School, where she graduated with honors and a certificate in Criminal Law and Procedure. At graduation, she was awarded the Rutgers Blaine E. Capehart Award for Legal Writing Excellence. As a research assistant to Professor Ellen P. Goodman, she worked on issues related to “algorithmic transparency” and contributed to the Rutgers Institute for Information Policy and Law blog. She also served as a Senior Notes and Comments Editor for the Rutgers University Law Review.

BARFIELD PRINT.indd 10

13/11/2018 11:52

Contributors  xi Hannah YeeFen Lim is Associate Professor of Law, Nanyang Technological University, Singapore, where she is utilizing her dual expertise in computer science and law in research projects tackling the legal and ethical issues of artificial intelligence (AI) machines and autonomous systems including autonomous vehicles (AVs). Her book Autonomous Vehicles and the Law: Technology, Algorithms and Ethics will be published in 2018 by Edward Elgar Publishing. She graduated with double degrees in Computer Science and Law from the University of Sydney, where she went on to complete a Master of Laws by Research with Honors under a Telstra Scholarship. She was a full-time visiting professor at the Faculty of Law, National University of Singapore from 2007 to 2011. Prior to that, she taught at the University of Sydney and UNSW. In 2013, she was Visiting Scholar at the University of Cambridge; in 2010, she was Visiting Researcher at Berkman Center for Internet & Society, Harvard University. Hannah is an internationally recognized legal scholar in Technology & Internet Law, IP Law, Data Protection and Privacy Law, and Cybercrime Law. She wrote one of the pioneering and seminal books on Internet law, Cyberspace Law (Oxford University Press 2002), and her research has been cited with approval by senior judiciary, most notably by the High Court of Australia. Hannah is the author of six scholarly books published by internationally established publishers. Crystal Nwaneri is a former fellow with the Berkman Klein Center for Internet and Society and is based in Silicon Valley. Her research covers a wide range of topics, including intellectual property law, virtual reality, innovation, and online media. She received her JD from Harvard Law School and her BA from Stanford University in Communication with a minor in Science, Technology and Society. At Stanford, she spent a year working as research assistant at the Virtual Human Interaction Lab, which studies the implications of human interactions in immersive virtual reality. She is the author of “Ready Lawyer One: Legal Issues in the Innovation of Virtual Reality,” published in the Harvard Journal of Law and Technology, and “Internet Retransmissions in the Real World: A Comparative Analysis of the Legal and Economic Effects of Over-the-Top Broadcast Services in Singapore, the United States, and Abroad,” published in the Information & Communications Technology Law Journal. Scott R. Peppet is Professor of Law and Wolf Getches Fellow at the University of Colorado Law School in Boulder, Colorado, where he has taught since 2000. His scholarship is focused on two areas: the ethics of alternative dispute resolution; and privacy, technology, and contract. In the former area he has published both an award-winning book, Beyond Winning (Harvard University Press 2004), and various academic articles on the use of contingency fees by mediators, the use of mediation in transactions as opposed to disputes, and the ethics of collaborative law. His more recent scholarship on privacy, technology, and contract has focused on the privacy problems of self-disclosure, the privacy implications of the Internet of Things, the impact of new technologies on consumer contract law, and the ways in which technology is changing prostitution markets. Michael Risch is Associate Dean of Faculty Research and Development and Professor of Law at Villanova University Charles Widger School of Law. Professor Risch’s teaching and scholarship focus on intellectual property and internet law, with an emphasis on patents, trade secrets, and information access. His articles have appeared in the Stanford Law Review, among others, and his work has been cited by the U.S. Supreme Court. Prior

BARFIELD PRINT.indd 11

13/11/2018 11:52

xii  Research handbook on the law of virtual and augmented reality to joining the Villanova faculty in 2010, he was an Associate Professor at the West Virginia University College of Law, an Olin Fellow in Law at Stanford Law School, and a partner at boutique IP law firm Russo & Hale LLP. Professor Risch graduated from Stanford University with honors and distinction in public policy and distinction in quantitative economics. He earned his law degree at the University of Chicago, where he graduated with high honors and was an Olin Fellow in Law & Economics and a Bradley Fellow in Law & Economics. Andrew L. Rossow is an internet attorney, adjunct Professor of Cyberspace Law at the University of Dayton School of Law, and contributor for many publications including Forbes Crypto, Blockchain Magazine, Thrive Global, and Cyber-Defense Magazine. Andrew concentrates his practice on cyberspace law and internet law, criminal defense, intellectual property law, contractual disputes, and most recently blockchain technology evaluations. Jack Russo is Managing Attorney of ComputerLaw Group, LLP, in Palo Alto, California (www.computerlaw.com). He  specializes in internet, computer law, and intellectual property litigation, in the federal and  state trial and appellate  courts. He is a member of the California, New York, DC, Oregon, Washington and Hawaii State Bars and Chairman of the Foundation for Creativity in Dispute Resolution (www.ffcdr.com). He has undergraduate and graduate degree from CUNY in administration and  computer and information  science and a JD from UCLA School of Law where he  was both  on the law review and the Moot Court. Jack also holds an MBA from Columbia University. He has written many articles on computer law, and he has spoken on IP, Internet, AI, and other computer industry issues to, among others, the American Bar Association, the Practicing Law Institute, the Computer Law Association, and the San Francisco Bay Area Intellectual Property Inn of Court. More about him and his firm can be found on www. computerlaw.com. Mirjam Supponen is Associate at GTC Law Group PC and focuses on  technology, internet and data privacy law. Supponen is a certified Information Privacy Professional in Europe (CIPP/E) and the United States (CIPP/US). Supponen holds a Master of Laws (LLM with a specialization in international data privacy law) from the University of Helsinki in Finland and a JD from the University of Hawaii. Amber M. Underhill is Associate at Warner, Norcross & Judd LLP, Grand Rapids, Michigan. Amber is a member of the firm’s Intellectual Property practice group, specializing in trademarks and copyrights. Before becoming an attorney, she worked for seven years as an IP paralegal. Amber is also a member of her firm’s Emerging Media and Technologies Team, and has significant experience managing the portfolios of companies in the augmented and virtual reality industry, among many others. Amber’s true passion is client service and providing the most practical advice to protect and enhance each client’s valuable intangible assets. Brian D. Wassom is Partner, Warner, Norcross & Judd LLP, Southfield, Michigan. Brian litigates disputes and provides counsel for clients in a wide range of commercial and intellectual property matters. He has developed a particular focus in matters of creative expression, commercial identity, and privacy, which spans the legal doctrines of

BARFIELD PRINT.indd 12

13/11/2018 11:52

Contributors  xiii copyright, trademark, privacy, publicity rights, advertising, journalism, and related fields. A futurist at heart, Brian has become a globally recognized thought leader in the brand new legal issues raised by augmented reality and other cutting edge emerging media. He leads his firm’s groups focused on Emerging Media and Technologies, Media Litigation, and Privacy Litigation, and regularly writes and speaks on these topics. An accomplished litigator, Brian has achieved victories for his clients in federal and state trial and appeal courts across the country. He was the lead counsel for the plaintiff in Candy Lab, Inc. v. Milwaukee County (E.D. Wis. 2017), resulting in the first judicial decision applying First Amendment free speech rights to the medium of augmented reality. Alexander Williams is Research Associate in Law and Technology, Chapel Hill, NC. Alex’s research focuses on the integration of technology into the body, and intellectual property law applied to virtual and augmented reality displays. He holds a degree in philosophy from the University of North Carolina and is trained in Six Sigma and principles of Experiment Design. He is the coauthor, with Dr Woodrow Barfield, of “Law, Cyborgs, and Technologically Enhanced Brains” and “Cyborgs and Enhancement Technologies,” both accessible online at Philosophies. Gilad Yadin is a Senior Research Fellow at the University of Haifa Center for Cyber Law and Policy. He holds a PhD from the Haifa University Center for Law and Technology. Dr Yadin consults on technology law and ethics and teaches at several academic institutions. In the past he drafted cybercrime legislation for the Israel Ministry of Justice, clerked for a criminal court judge, and managed a research and development group at a top cybersecurity firm. His research has been published in leading law and technology journals such as Cardozo AELJ and Vanderbilt JETLaw.

BARFIELD PRINT.indd 13

13/11/2018 11:52

Preface Recent advances in computer display technology are changing the way we perceive and interact with the world—and, by doing so, are challenging the law on several fronts. Not the least of these are disputes that involve the design and use of virtual and augmented reality worlds. Once fledgling technologies that resulted in theoretical discussions of what law “might” apply, virtual and augmented reality are now mainstream, giving rise to legal disputes in the here and now. Courts are already struggling with questions of who has property rights in virtual content and they are beginning to wrestle with questions that arise when our interaction with virtual and augmented reality spills out into “brick and mortar” interactions. The goal of the Research Handbook on the Law of Virtual and Augmented Reality is to provide an up-to-date discussion of such legal questions, as well as those which may arise as the technology continues to develop and find use in more and more spheres of our daily lives. Just as consumers’ embrace of the Internet transformed the way people communicated, worked, shopped, invested their money, and shaped and tended to their reputations—and raised numerous legal questions as it did so—so might virtual and augmented reality change the way we experience public space and interact with each other, and, in doing so, require a rethinking of many areas of law. The transformation of society generated by virtual and augmented reality technologies is likely to bring novel changes. These technologies will allow users to enter worlds which are far beyond our normal experience. When they do so, they can allow people to learn new skills, explore new places, and experience times and events in ways that otherwise would have been impossible. With the aid of virtual reality, an individual can have the experience of flying to, or walking upon, Mars, or scaling down and traveling inside the human body, or defying gravity, or hovering over an imaginary city—or a real city (such as New York or Paris) when they are miles away from it. The virtual world viewed on a headmounted display becomes the world one sees and hears, replacing the physical environment blocked out by the headgear. More recently, newer versions of virtual reality have also provided haptic or olfactory feedback, thus stimulating more of the human senses (and raising further issues for the law to consider). Augmented reality extends this virtual reality technology in another way: it supplements our experience of the physical environment rather than replacing it. It overlays information (or images such as fantasy game creatures) over real-life visual scenes that continue to reach our eyes from the surrounding environment—through the lenses of glasses, smartphone screens, or some other interface that can display what is in front of us, but also blend in virtual pictures or words. Moreover, thanks to such augmented reality technology, what was once tied to a specific physical setting can now be experienced with mobile technology in the form of augmented reality provided by smartphones or low-cost virtual reality gear. A benefit of augmented reality is that it allows computer-generated information to be accessed by the user any time and at any place. Until recently, virtual and augmented reality technologies were used mainly by researchers and test subjects in research laboratories. However, due to dramatic cost reductions xiv

BARFIELD PRINT.indd 14

13/11/2018 11:52

Preface  xv in display technology and a proliferation of applications, virtual and augmented reality are now mainstream technologies used by millions of people around the globe. In fact, virtual and augmented reality technologies have been used for a wide variety of applications. They have been used in hospitals to train physicians to perform complex surgical procedures. They are increasingly used in psychotherapy. They have also been used in online virtual games for play, such as Second Life, a virtual reality world visited by millions of people yearly. In the past few years, technology companies such as Google, Sony, HTC, Microsoft, and Apple have raced to provide consumers virtual reality videogame consoles (such as Oculus Rift, HTC Vive, and Playstation VR) and enter a growing commercial market. Some of the same companies have touted new augmented reality platforms that enable the development of applications that superimpose upon real world images a smartphone camera image—such as an illusory monster (as part of a location-based virtual game), a virtual piece of furniture (that a home owner can thus better envision in her own home before a purchase), or historical information about the neighborhood or building by which the viewer is wandering (as part of a self-directed tour). As noted earlier, this increasing use of augmented and virtual reality has raised a host of questions about how to apply legal doctrines developed for the brick and mortar world, or for earlier information and Internet technologies, to the new, less familiar realm of virtual and augmented reality. Many of the chapters in this Research Handbook discuss some of these ongoing lawsuits. Individuals, for example, have filed suits (alleging nuisance and other tort violations) against users and the designer of Pokémon Go, claiming that this augmented reality game is leading droves of players to trespass onto private land and cause damage or create nuisances there as they hunt for the fictional Pokémon monsters they must target in the game. Milwaukee has likewise taken action to limit such use of augmented reality, focusing not only on Pokémon Go, but on all location-based augmented reality games. More specifically, that state has enacted an ordinance—under legal challenge at the time of writing—restricting the “introduction” of certain virtual and augmented reality games in its public parks. Additionally, Linden Lab, the company that operates the Second Life virtual world, was sued by an individual whom it barred from future access to its virtual world. This, said the individual, effectively deprived the individual of virtual property he had spent years obtaining, and which he had spent substantial money to acquire. Further, individuals in virtual worlds have considered themselves to have been sexually assaulted when subjected to “unconsented to” touching of their virtual world avatar (the character they control within a virtual world), even though the touching occurs in virtual reality and not in physical space. And still other legal questions are on the horizon. Avatars, for example, are becoming increasingly smart and autonomous: They are no longer simply virtual world puppets, controlled in every action by a flesh and blood person. They are artificial intelligences that can generate their own decisions. When, in such situations, do legal rights or liabilities attach to their AI-generated actions? Are they subject to criminal liability for the acts they perform? Do smart avatars receive legal protections against restrictions of their activities? Which activities occurring in virtual and augmented reality should be considered speech under the First Amendment and what degree of protection should courts afford such speech? And how should copyright law apply to virtual reality worlds? Further, when individuals reuse virtual and augmented reality content that is meant to faithfully

BARFIELD PRINT.indd 15

13/11/2018 11:52

xvi  Research handbook on the law of virtual and augmented reality represent the real world, is this an infringement of a copyright owned by the creator of such an image? The Research Handbook on the Law of Virtual and Augmented Reality contains 20 chapters from US and international legal scholars on a wide range of topics, such as whether virtual sex experienced using haptic technology is a form of protected speech under the First Amendment, or the extent to which virtual images projected in real space for advertising purposes can be regulated by the government. The federal and state law of the US is the focus of the book but to provide the reader a broader view of legal issues related to virtual and augmented reality, especially given its global proliferation, this book also contains chapters from legal scholars in Asia and the European Union exploring how the law of these jurisdictions might apply to virtual and augmented reality technology and applications. Additionally, since the use of augmented reality raises challenges to established legal doctrine that may be distinct from those raised by virtual reality technology, several chapters focus specifically on questions about how the law applies to augmented reality—for example, whether a virtual image projected over the real property of another person can be considered a trespass or a nuisance, and how the law of contract might be challenged by the use of augmented reality. One of the central issues addressed in the Research Handbook on the Law of Virtual and Augmented Reality is the extent to which “real-world” legal doctrine (in its existing form) can be applied to activities occurring in virtual and augmented reality—or whether lawmakers (or virtual and augmented reality creators or users) need to instead generate new or modified legal doctrines that are a better fit with this technological landscape. For example, in virtual and augmented reality worlds, the law of tort may be relevant if one’s virtual avatar is defamed by, or has her privacy invaded by, another avatar. But this depends on whether (and when) avatars can be liable for torts, or the subject of a tort— and where they cannot be, the ways in which avatars’ actions can and cannot establish tort liability for the individuals (if any) who exercise some control over them. And in the area of criminal law, as more people use virtual reality for a range of tasks, new vulnerabilities may be created for people within virtual worlds, and new ways for criminals to exploit the virtual environment may develop. But how criminal law will apply here depends in part on what virtual law harms generate criminal liability (and when virtual reality designers, individuals operating virtual avatars, or the avatars themselves) can be liable for virtualworld crimes. Additionally, intellectual property law is especially likely to be relevant for activities occurring in virtual and augmented reality worlds, given that both technologies rely on use of visual images and that the storylines they use to engage participants may be protected by copyright law. If a company’s trademark is visible on a virtual billboard displayed in virtual or augmented reality, and if the image is used without permission, there may be a cause of action for trademark infringement. And if a virtual avatar resembles the appearance of a famous actor, there may be a “right of publicity” cause of action to protect the celebrity from misappropriation of his/her name, likeness, voice, image, or other recognizable element of personal identity. The above examples suggest that the law developed for the real world will likely be applicable for some activities occurring in virtual and augmented reality. But as both technologies evolve and allow more senses to be realistically stimulated, established law may be insufficient to account for the new experiences that virtual and augmented reality will provide for users.

BARFIELD PRINT.indd 16

13/11/2018 11:52

Preface  xvii While the chapters discuss the current law, which applies to virtual and augmented reality, they are also “forward looking” in that they raise interesting questions about how the law may respond to continuing advances in virtual and augmented reality technologies. We hope the book will stimulate discussion on how the law should apply to virtual and augmented reality worlds and that it provides a valuable resource to legal scholars, judges, and legislators in determining how to protect the rights of all parties involved in the design and use of virtual and augmented reality. Woodrow Barfield, Chapel Hill, North Carolina Marc Jonathan Blitz, Oklahoma City, Oklahoma

BARFIELD PRINT.indd 17

13/11/2018 11:52

BARFIELD PRINT.indd 18

13/11/2018 11:52

PART I INTRODUCTION TO THE LAW OF VIRTUAL AND AUGMENTED REALITY

BARFIELD PRINT.indd 1

13/11/2018 11:52

1.  The law of virtual reality and increasingly smart virtual avatars Woodrow Barfield and Alexander Williams

1. INTRODUCTION Advances in virtual and augmented reality technology, and in the software to produce virtual worlds, have allowed virtual avatars to be used with increasing frequency for a range of activities. But with improvements in the technology to create virtual and augmented reality worlds and decreases in equipment costs have come corresponding issues of law and policy applying to the avatars which represent the human presence in virtual environments. This chapter discusses several issues of law which relate to the design and use of virtual avatars, with a specific focus on avatars that are gaining in intelligence and operating with more and more autonomy from humans.1 To introduce this topic we present an overview of legal disputes brought forth by virtual and augmented reality worlds, including those implicated when virtual avatars are used for commercial activities. After the initial discussion focusing on legal disputes within virtual and augmented reality worlds, the remainder of this chapter focuses on issues of law and policy which relate more specifically to the design and use of increasingly smart virtual avatars. For the purposes of this chapter, we note that virtual avatars are increasingly taking a form in which they are not simply computer-generated “puppets” manipulated by real-world users, as is the case in many online virtual reality games, but rather are becoming more autonomous actors, generating their own decisions and solutions to problems which may not always be intelligible or transparent to the real-world users they represent. Further, it is the case that “familiar” virtual avatars controlled almost entirely by real-world users bring up a host of legal issues, such as whether the avatar uses copyrighted or trademarked material and, if it does, who is liable. But while increasingly smart avatars operating with greater autonomy will raise many of the same legal issues as do current avatars, they will also raise further significant legal issues, such as whether the avatars themselves deserve rights. While many current avatars represent alter egos of real-world users in a virtual world, the concept of an avatar is expansive and can refer to any character or perhaps any object in the virtual world—and may even represent no natural person outside the virtual world. A virtual environment (or virtual world) is an interactive computer simulation which lets its participants see, hear, use, and even modify the simulated objects in the computergenerated environment.2 Within a virtual environment, the user may be stimulated by

1   See generally, W. Barfield, Intellectual Property Rights in Virtual Environments: Considering the Rights of Owners, Programmers and Virtual Avatars, 39 Akron L. Rev. 649 (2006); W. Barfield (ed.), 2nd edition, Fundamentals of Wearable Computers and Augmented Reality (CRC Press 2015). 2   W. Barfield and T. Furness (eds), Virtual Environments and Advanced Interface Design (Oxford University Press 1995).

2

BARFIELD PRINT.indd 2

13/11/2018 11:52

The law of virtual reality and increasingly smart virtual avatars  3 a range of sensory information, including spacialized sound, stereoscopic imagery, olfactory cues, and force or tactile feedback. Some commentators have even suggested that developments in virtual environments are occurring so rapidly that humans may “inhabit” these environments within the foreseeable future. Although this may seem like a bold prediction, many people are already spending significant amounts of time in virtual environments engaging in activities such as learning skills or performing recreational activities. One reason for the time spent in virtual and augmented reality is that the participants may experience a sense of presence, which is generally regarded as the suspension of disbelief that one is viewing a simulation, that is, the experience of actually “being there” in the virtual or augmented reality world.3 More realistic virtual environments lead to a higher sense of presence, and it has been shown that one way to increase the realism of a virtual environment is by projecting avatars in the virtual world that have the ability to interact with humans and, in some cases, to serve as an alter ego for a human creator. Participation in a virtual world through the use of a virtual avatar is a compelling experience, especially for those who play massively multiplayer online roleplaying games and for those that experience virtual reality through head-mounted, stereoscopic 3D displays. The virtual avatar(s) representing people enable thousands of participants to interact simultaneously within the same virtual world, and in such settings these participants effectively become the avatars they have created, often looking through their eyes and engaging with other such virtual beings. Virtual avatars can be thought of as computer programs consisting of datasets and algorithms, along with a visual representation. As such, they may receive the legal protection that is awarded software, and the protection awarded images from copyright and trademark law. However, unlike standard software programs, intelligent avatars may deviate from their original programming until they are no longer recognizable to the original programmer(s), raising a host of legal issues beyond the scope of current virtual avatars. These issues are the focus of this chapter.

2.  THE USE OF AVATARS FOR COMMERCIAL ACTIVITIES Since increasingly “smart” avatars are being used more and more in commercial activities, in this section we introduce examples and legal issues which relate to the use of virtual avatars for marketing and branding purposes. We then move on to other issues of law relating to virtual worlds and increasingly smart virtual avatars. Certainly, introducing real-world brands, in some form or another, within virtual worlds is a viable means of engaging in commerce, and is an area that has generated several legal disputes. Additionally, an interesting question of commerce for virtual reality is whether avatars will actually buy real-world products that are marketed in virtual worlds, in effect purchasing real-world goods for their creators, just as those creators buy virtual-world paraphernalia for them. Could an avatar who currently spends Linden dollars to buy a virtual dress from another

3   W. Barfield and S. Weghorst, The Sense of Presence within Virtual Environments: A Conceptual Framework, in Human-Computer Interaction: Applications and Case Studies (G. Salvendy and M. Smith eds) 699–704 (Elsevier 1993); W. Barfield, Musings on Presence, 25 Years After “Being There,” 2 Presence: Teleoperators and Virtual Environments (2016).

BARFIELD PRINT.indd 3

13/11/2018 11:52

4  Research handbook on the law of virtual and augmented reality avatar’s designer clothing store in Second Life be enticed, while visiting an in-world Gap retail outlet, to click on a cash register and use his or her creator’s credit card to buy a real-world Gap sweater that would be shipped to the creator’s doorstep?4 In terms of efforts to market real-world products in virtual worlds, companies are beginning to look beyond the overall market itself and to think about the potential individual customer, which may be the person behind the avatar or, with increasing intelligence, the avatar itself rather than its creator.5 The virtual avatar, as a distinct creation of the user’s psyche, can influence its creator’s purchasing behavior and even make its own purchases of real-world products in the virtual world, deliverable to the user’s real-world door. At the least, avatars offer a window into people’s hidden preferences and a means for achieving sustained consumer engagement with a brand.6 According to writer Paul Hemp, the avatar arguably represents a distinctly different “shadow” consumer, one able to influence its creator’s purchase of real-world products and conceivably make its own real-world purchases in the virtual world.7 At the least, it may offer insights into its creator’s hidden tastes. Further, online virtual worlds offer untapped marketing potential for real-world products and services, particularly because of their ability to generate sustained consumer engagement with a brand. This occurs through interactions with “avatars,” the “virtual beings” users create as representations of themselves and through which they live and relate to others in these worlds. Increasingly smart virtual avatars might also be enlisted to play a direct role in marketing, in that they could use their virtual-world experience in creating virtual property to design products with real-world potential. For example, several Second Life clothing designers have been approached by real-world fashion houses, and at least one business makes real-world versions of furniture based on virtual “furni” designed by Second Life residents.8 Avatar brokers could link real-world companies with virtual landowners willing to rent space for the companies’ marketing initiatives. This implicates the laws of property and of licensing. Avatars ultimately could run virtual-world stores selling realworld products and paid to publicize, overtly or not, those same products. This implicates the law of contract, and the terms of contracts for virtual reality games have already been an issue of dispute within virtual worlds. For example, in Evans v. Linden Research, Inc.9 the plaintiffs were in litigation with Linden Lab over their virtual property rights, arguing that by suspending their accounts Linden had deprived them of their virtual goods. The company asserted that, according to its terms of service agreement, the plaintiffs only had rights to the intellectual property of goods created in-world, not to the goods themselves.10 Further, many commercial brands that are placed in the virtual world will be protected by trademark law. In general, trademark law protects against the unauthorized use of a trademark in a manner that causes a reasonable consumer to believe that the trademark owner either was the source of the goods or endorsed or sponsored such goods. What if      6    7    8    9   10    4

 5

BARFIELD PRINT.indd 4

Paul Hemp, Getting Real about Virtual Worlds, https://hbr.org/2006/06/avatar-based-marketing. Id. Id. Id. Id. Evans v. Linden Research, Inc., 763 F. Supp. 2d 735 (E.D. Pa. 2011). Id.

13/11/2018 11:52

The law of virtual reality and increasingly smart virtual avatars  5 the virtual reality experience lets a person drive a virtual Mercedes? Will the virtual-world participant think that Mercedes created the virtual reality experience or sponsored it? If so, the consumer may believe that Mercedes endorsed the use of their brand in the virtual world—and if the brand is shown in an unfavorable light, a trademark cause of action may be warranted. Further, what if the virtual environment participant and not the virtual reality company imported the trademarked item into the virtual world, and what if the user can sell the trademarked item they imported for virtual or actual cash? Under US law, there are two pieces of legislation that provide significant liability protection for those who host interactive online media such as the virtual game world of Second Life. The Communications Decency Act (CDA) of 1996 (aimed at liability for defamation cases), and specifically Section 230, subsection (c)(1) of the Act, states: “No provider or user of an interactive computer service shall be treated as the publisher or speaker of any information provided by another information content provider.”11 Additionally, under the Digital Millennium Copyright Act (DMCA)12 an internet service provider may not be considered to be the publisher of the content itself, and therefore may not be liable for the actions of their users (this is the so-called safe harbor provision of the DMCA, applicable for copyright infringement).13 The concept behind the safe harbor Section 512 in the DMCA is that providers are shielded from the burdens of knowing and policing the contents of their network. The cumulative protections provided by the CDA and DMCA can potentially eliminate liability for copyright infringement (DMCA) and defamation (CDA) for those who host virtual environments populated by users’ avatars. Under US law the safe harbor provisions of the DMCA do not protect virtual reality companies from trademark claims, but by analogy to general trademark infringement cases, such as Tiffany v. eBay,14 the virtual reality companies should be protected from contributory infringement claims as long as they either are not aware of the conduct or take action to remove the infringing content if they become aware of it. Furthermore, under the US Federal Trademark Anti-Dilution Act,15 nationally known brands can sue if the use of their trademark by others in advertising or branding a product “tarnishes” or “blurs” the trademark. This Act applies whether or not consumers are confused as to the source of the goods, so it is more analogous to a copyright claim than a trademark claim. Thus, virtual reality companies should be diligent not to include any images of nationally known brands in their advertisements, such as online trailers, teasers, or opening web pages. In addition, augmented reality—which superimposes virtual imagery onto real-world environments, as popularized by the Pokémon Go craze—is also exploring the use of virtual images projected in the physical world for advertising purposes. For years, major   47 U.S.C. § 230(c) (2000), Communications Decency Act.   See Pub.L. No. 105-304, 112 Stat. 2860 (1998) (codified in scattered sections of 17 U.S.C.), among others, the DMCA makes it illegal to circumvent measures that copyright owners use to control others’ access to their works. 13   Jonathan Band and Matthew Schruers, Safe Harbors Against the Liability Hurricane: The Communications Decency Act and the Digital Millennium Copyright Act, 20 Cardozo Arts & Ent. L. J. 295 (2002); see also David Sheridan, Zeran v. AOL and the Effect of Section 230 of the Communications Decency Act upon Liability for Defamation on the Internet, 61 Alb. L. Rev. 147 (1997–8). 14   Tiffany Inc. v. eBay Inc., 600 F.3d 93 (2d. Cir. 2010). 15   Federal Trademark Dilution Act of 1995, Pub. L. No. 104-98, 109 Stat. 985 (Jan. 16, 1996), amending the Trademark Act of 1946, codified at 15 U.S.C. § 1125 et seq. 11 12

BARFIELD PRINT.indd 5

13/11/2018 11:52

6  Research handbook on the law of virtual and augmented reality brands have offered mobile augmented reality apps that create interactive 3D digital models of their product. But with augmented reality, some interesting issues of law have arisen as users view the virtual images throughout the real-world environment. For example, Pokémon Go allows users to “capture” Pokémon on private property, which has resulted in trespassing lawsuits and injuries to participants despite the game featuring warnings such as “Remember to be alert at all times. Stay aware of your surroundings.” Another issue is that with augmented reality there is a responsibility to ensure that the digital representations of a product faithfully convey the product’s actual features; if this is not the case, several causes of action are available. For example, if the augmented reality demonstration of a product materially misleads or misinforms the consumer about a product’s features, the advertiser could find itself liable for false advertising. In the US the primary legal authority creating the right to sue for such actions is the Lanham Act, which also regulates the use of trademarks. False advertising claims may be filed by a regulatory agency or by a private party whose interests may be negatively affected by the challenged activity. The difficulty involved in precise 3D rendering of virtual images could likewise result in augmented ads that are similarly unrealistic. In the automotive realm, for example, a digital artist may leave out or oversimplify important features when rendering a product, in ways that mislead consumers into believing things about the product that are not accurate. By definition, digitally enhancing physical reality is a fundamental element of what virtual and augmented reality do; yet without careful attention to the design or products displayed in the virtual worlds, a company may find itself subject to a lawsuit.

3.  DISPUTES IN VIRTUAL AND AUGMENTED REALITY Continuing the discussion on how the law applies to virtual worlds, we observe that the law of tort may be relevant if a person using virtual reality technology is injured or property is damaged.16 To illustrate this idea, consider the possibility of “real-world” injuries or property damage which may occur when one plays a virtual reality game. To proactively counter claims of negligence and product liability tort actions, virtual reality manufacturers and “publishers” of the content displayed in virtual reality typically use “warning” stickers and disclaimer language in license agreements, to limit liability for 16   For example, in a study published in the British Medical Journal, BMJ 2014;349:g7267, researchers from the Netherlands outlined the findings of previously published reports involving injuries and other problems related to the original Nintendo system, Nintendo 64, and Wii gaming consoles. Nearly 1,200 articles involving mild to life-threatening injuries were identified. Reports of Nintendo controller injuries ranged from tendonitis, fractures, and neurological and psychological problems to surgical concerns. See www.aboutlawsuits.com/nintendo-injury-study-75923/#sthash. jDWNKPH0.dpuf. According to a case report published in the New England Journal of Medicine, many of the more severe injuries were related to use of the Wii Fit balance board, which rests two inches off the ground and requires users to quickly adjust their balance while standing on the board to play certain video games. Most commonly, users suffered what doctors called “Wii-knee,”  which is caused by frequent bending at the  knee while using the balance board. In some cases, such injuries have involved the kneecap becoming dislocated. See www.webmd.com/ fitness-exercise/news/20100203/when-wii-goes-wrong-video-game-injuries.

BARFIELD PRINT.indd 6

13/11/2018 11:52

The law of virtual reality and increasingly smart virtual avatars  7 manufacturers of virtual reality technology and for software designers of virtual and augmented reality games. Since the development of virtual reality games by companies such as Nintendo, various personal injury claims—such as for sprained wrists and dislocated shoulders arising from the use of the equipment to interact with the virtual world—have arisen. For example, Nintendo’s Wii Tennis and Wii Boxing were both the subject of civil action,17 regarding use of the controller’s safety strap.18 In Elvig v. Nintendo of America, Inc., a class action lawsuit which involved, among others, breach of implied warranty, product liability, and negligence, the court held that the danger involved in losing possession of the Wii controller while playing a game, namely that it could act as a “damaging missile,” were covered by the company’s repeated warnings in that regard.19 More specifically, on the product liability claim, Nintendo argued that it gave players adequate warning of the need to retain possession of the controller and advised them of the possibility that release of the controller during vigorous motion could result in breakage of the strap and lead to damage to persons or property. In agreement, the court noted that Nintendo advised players via a safety card included with the Wii system that “If you use excessive motion and let go of the Wii Remote, the wrist strap may break and you could lose control of the Wii Remote. This could injure people nearby or cause damage to other objects.”20 This, coupled with repeated instructions on the safety card that advised players “DO NOT LET GO OF THE REMOTE DURING GAME PLAY,” ensured that, if the player followed Nintendo’s instructions and heeded its warnings, the Wii system would not pose an unreasonable danger. On the implied warranty action, the court cited a lack of evidence to indicate just what the intended purpose of the strap was. To withstand summary judgment, the plaintiff needed to show more than an alternate plausible theory on the purpose of the strap; rather, the plaintiff needed evidence of the ordinary purpose of the strap and proof that it failed the ordinary purpose (the plaintiff failed to provide this). Based on the above reasoning the court granted summary judgment for the defendant on all claims.21 Additionally, game platforms for virtual reality games and service providers have sought to limit their liability for user generated content and in particular, user comments which could result in a cause of action for defamation. This they do by refraining from actively moderating games, to avoid serving as a publisher of content.22 Further, to avoid liability, content providers may establish takedown policies and procedures to quickly remove infringing content, and also to require users of virtual worlds to sign robust terms of use policies regulating how users interact with one another in the virtual world. Defamation may also be relevant for avatars communicating in virtual worlds. An interesting question is: How will (or should?) real-world defamation laws be used to protect the rights of avatars, for example, avatars that appear in the popular online game Second Life? In general, defamation laws are designed to protect a person from having their reputation tarnished in their community by another person making false claims about     19   20   21   22   17 18

British Journal of Medicine, ibid. Elvig v. Nintendo of America, Inc., 696 F.Supp.2d 1207 (2010). Id. Id. Id. See e.g., Defamation Act of 1996, Great Britain.

BARFIELD PRINT.indd 7

13/11/2018 11:52

8  Research handbook on the law of virtual and augmented reality them. Additionally, an action for defamation may occur when a communication is made about a person that harms their reputation in the eyes of at least a substantial minority in the community, or deters others from associating with that person. But what happens if an avatar makes defaming communications about a second avatar that lowers the second avatar’s reputation in a virtual world such as Second Life and as a result other avatars will not associate with them (commercially or socially)? What steps, if any, can the real person behind the defamed avatar take in order to rectify this situation? And with increasing intelligence and autonomy, would the avatar itself have any rights?23 An interesting case involving virtual images, Amaretto Ranch Breedables, LLC v. Ozimals, Inc., was litigated in a United States District Court and involved a DMCA takedown notice between companies that produced virtual animals in Second Life.24 The plaintiff brought forth several legal issues such as copyright, defamation, and trade libel claims which were based on one Second Life trading entity competing against another.25 The defamation action arose from comments made in-game with regard to a copyright dispute and potential damages it may have caused. For a takedown claim arising from activities in the virtual world, the DMCA permits content owners to prevent infringement by seeking a court order that requires an internet service provider to block or remove access to content that allegedly infringes copyrighted content. More specifically, the DMCA permits content owners to prevent infringement by seeking a court order that requires an internet service provider to block or remove access to content that allegedly infringes copyrighted content. In this case, Ozimals’ copyright claim alleged that Amaretto Ranch Breedables’ virtual horses infringed on their virtual bunnies and filed a takedown notice for their virtual horses and feed. Amaretto sued for a temporary restraining order and preliminary injunction, which was granted by the court and which barred Linden Research, the operator of Second Life, from removing their products. Had the DMCA order gone through, it would have cut off Amaretto’s source of income before it could challenge Ozimal’s copyright claim. Ultimately, the court granted Amaretto the temporary restraining order against Linden Research, stopping it from removing the horses until a preliminary injunction ruling was made. A leading jurisdiction for tort law disputes is California. Under California law,26 the elements of a defamation claim are: publication of a statement of fact that is false, that is unprivileged, and that has a natural tendency to injure or cause “special damage,” where the defendant’s fault in publishing the statement amounts to at least negligence. Publication could involve the output (or speech) of virtual avatars, may be written or oral, and involves communication to a third person (or avatar?) who understands the defamatory meaning of the statement and its application to the person to whom reference is made. Publication need not be to the “public” at large; communication to a single individual other than the plaintiff is sufficient. However, republication of a defamatory statement made by another is generally not protected. Further, under California law, a plaintiff need not show special damages—for example, damages to the plaintiff’s  Barfield, supra note 1.   Amaretto Ranch Breedables, LLC v. Ozimals, Inc., 29 (N.D. Cal. December 21, 2010). 25   Scheck, Justin, You Can Lead a Virtual Horse to Water, But You Might Get Sued Along the Way, Wall Street Journal (2011). 26   Cal. Civ. Code §§ 44, 45a, and 46. 23 24

BARFIELD PRINT.indd 8

13/11/2018 11:52

The law of virtual reality and increasingly smart virtual avatars  9 property, business, trade, profession, or occupation, including expenditures that resulted from the defamation—if the statement is defamation per se. A statement is defamation per se if it defames the plaintiff on its face, that is, without the need for extrinsic evidence to explain the statement’s defamatory nature.27 California courts recognize a number of privileges and defenses in the context of defamation actions, including the fair report privilege, the opinion and fair comment privileges, and substantial truth; all of these could be applicable to virtual worlds in a defamation action occurring with virtual avatars. As indicated above, another major area of law, the law of property, is also being challenged by the buying and selling of virtual images/property. For example, Linden Lab, the company behind Second Life, allows users to lease virtual land to other players, on which they can build and sell virtual items and services within the game. This raises the interesting question of whether the ownership of virtual goods should be treated as analogous to ownership of real property and goods in the real world as opposed to a licensed service provided by the operator of the virtual reality world and governed by the End User License Agreement (EULA). An example case is Bragg v. Linden Research, Inc.,28 in which players of Second Life were allowed to build objects, explore the world, interact with other players, and purchase land with real-world money. In 2003, Linden announced that it would recognize participants’ full intellectual property protection for the digital content they created or otherwise owned in Second Life, such as cars, homes, or slot machines. Players were also able to purchase “virtual land,” make improvements to that land, exclude other players from entering onto the land, rent the land, or sell the land to other players for a profit. The plaintiff, Bragg, had purchased numerous parcels of land in Second Life, including a parcel of virtual land named “Taesot.” However, Linden sent Bragg an email advising him that Taesot had been improperly purchased through an “exploit,” and as a result Linden took Taesot away. It then froze Bragg’s account, effectively confiscating the virtual property and currency that he maintained on his account with Second Life. When Bragg sued Linden, Linden argued that their agreement required mandatory arbitration, a clause which was central to the dispute. Before a person is permitted to participate in Second Life, they must accept the terms of service by clicking a button indicating acceptance. Bragg resisted enforcement of the terms of service arbitration provision on the basis that it was “both procedurally and substantively unconscionable.” In California law, the procedural component of an unconscionable contract can be satisfied by showing oppression through the existence of unequal bargaining positions, and surprise through hidden terms common in the context of adhesion contracts. The court stated that a contract is procedurally unconscionable if it is a contract of adhesion.29 A contract of adhesion is a “standardized contract, which, imposed and drafted by the party of superior bargaining strength, relegates to the subscribing party only the opportunity to adhere to the contract or reject it.” In this case, the Second Life terms of service comprised a standardized contract which only allowed the customer to agree to it or reject it.30 The court held: “When the weaker party is presented the clause and told to ‘take it or leave it’ without the opportunity for meaningful 27   See Cal. Civ. Code § 45a; Yow v. National Enquirer, Inc. 550 F.Supp.2d 1179, 1183 (E.D. Cal. 2008).  28   Bragg v. Linden Research, Inc., 487 F. Supp. 2d 593 (E.D. Penn. 2007). 29   Id. 30   Id.

BARFIELD PRINT.indd 9

13/11/2018 11:52

10  Research handbook on the law of virtual and augmented reality negotiation, oppression, and therefore procedural unconscionability, are present.”31 An arbitration agreement that is an essential part of a “take it or leave it” employment condition, without more, is procedurally unconscionable. In this case, there was no alternative to Second Life in the market at the time—so if Mr Bragg did not accept these terms he could not play Second Life or any other similar game or service, which created an unconscionable take-it-or-leave-it situation. Linden also had superior bargaining strength over Mr Bragg, and the contract was therefore one of adhesion. The plaintiff’s suit was ultimately settled before a final decision was reached; however, the District Court decided that the Second Life terms of service’s mandatory arbitration provision was unenforceable and, interestingly, that interaction with a person visiting a virtual world could be a factor to satisfy a state’s “minimum contacts” requirement for personal jurisdiction. If legal systems ultimately decide that virtual reality purchased and acquired content should be afforded similar status as real-world goods and property, this will have profound effects on the virtual reality business model and lead to further questions relating to common law contract law—and, in the US, the Uniform Commercial Code (UCC)— such as: Will users be able to buy and sell their virtual content in the real world? Does software used to create virtual worlds fall under the banner of goods or services? Are virtual images/objects used in virtual worlds goods or services? Will we see a move away from trying to deal with ownership and the rights around virtual content under the End User Terms and Conditions and a move toward discussing “leases” of virtual property, and therefore implicating the law of landlords and tenants? Furthermore, how will the theft of virtual goods and content by another user be dealt with? While courts in South Korea have, through the Game Industry Promotion Act, actively considered the question of the legal status of virtual reality goods,32 many jurisdictions have not addressed the issue of property rights in virtual and augmented reality. Surely this is an area requiring legislation, not to mention a body of case law to aid courts in deciding property disputes occurring in virtual and augmented reality worlds. As another important issue involving virtual and augmented reality worlds, consider the problem of determining jurisdiction for the actions occurring within virtual worlds especially when virtual avatars perform activities across state and international boundaries. In the US, the Federal Rules of Civil Procedure and states’ long-arm statutes apply. If a dispute arises in the virtual world, which real-world law and jurisdiction applies? The answer can be complex and may in part depend upon the nature of the dispute and the jurisdiction laws of a particular physical location. In virtual worlds, the service provider will typically seek to rely upon the laws and jurisdiction set out in its terms of service agreement. However, where a contract is made between a business and a consumer domiciled in the European Union (EU), e-commerce regulations provide that the business-to-consumer contract cannot deprive the consumer of the protection of the laws of the country in which he or she resides. This is fine in a commercial dispute, but what about situations occurring in virtual reality where a criminal offense has taken place—for

  Id.   Grubb, J., South Korea Blocks All Facebook Games as Part of a Government Crackdown, 2014, https://venturebeat.com/2014/09/02/south-korea-blocks-all-facebook-games-as-part-of-a-gove​ rnment-crackdown/. 31 32

BARFIELD PRINT.indd 10

13/11/2018 11:52

The law of virtual reality and increasingly smart virtual avatars  11 example, where the avatar of one player is alleged to have virtually assaulted or stolen from another player? Should this be decided by the criminal courts or is this a commercial dispute, and in either case, which law and jurisdiction applies? Do the related terms of service provide the framework from which to derive jurisdiction for multiuser programs and in determining which governmental regulations should be brought forth to interpret them? As mentioned above, in Bragg, the defendant filed a motion to remove himself from the litigation based on a lack of personal jurisdiction. Although much of the court’s discussion of minimum contacts is fairly boilerplate, the judge’s analysis included a unique twist when he recognized that the defendant’s avatar may have actually interacted with Bragg’s avatar within the virtual world: Once inside Second Life, participants could view virtual property, read additional materials about purchasing virtual property, interact with other avatars who owned virtual property, and, ultimately, purchase virtual property themselves.33 Significantly, participants could even interact with the defendant’s avatar in Second Life during town hall meetings. These are factors occurring in virtual worlds that courts may consider in determining jurisdiction. As another example of an emerging area of law where virtual environments have been implicated, consider the reconstruction of evidence of a crime scene. For example, in a criminal law case, the defendant was convicted and sentenced to 20 years’ confinement. As part of the evidence presented, a virtual reality recreation of the route driven to strike the victim was shown. The Court of Appeals of Texas upheld the allowance of this evidence, concluding that the probative value of the virtual reality crime scene recreation was not substantially outweighed by the danger of misleading the jury.34 Additionally, a topic for a law of virtual reality is whether “real-world” crimes such as those listed in the US Model Penal Code can be committed in virtual worlds. Given that there are few cases or legislation specifically addressing how virtual acts map to real-world criminal law, we must examine related cases and statutes for a sense of how courts might decide such issues. In the United Kingdom, for example, an individual hacked into games publisher Zynga’s social games platform and transferred 400 billion virtual poker chips into his account. Consider another example in which real-world crime bled into virtual worlds to the extent that the term “virtual mugging” was coined. It was argued that when some players of the virtual reality game Lineage II used bots to defeat other players’ characters and take their items, they were essentially mugging virtual avatars. Some countries have even introduced special police investigation units to cover “virtual crimes.” For instance, in the US the FBI’s White Collar Crime Division investigates fraud committed by business and government professionals, and particularly money laundering—a number and variety of methods are used by criminals to launder money, including the use of virtual currency. Similar agencies exist in other law enforcement bodies which focus on cybercrimes, such as Homeland Security’s Cyber Crimes Center, Europol’s European Cybercrime Centre, and England’s National Cyber Crime Unit. Undoubtedly, as virtual systems are increasingly used, examples of virtual crime will capture the attention of national and international enforcement agencies, with all of their established procedure and jurisdiction. Still, we expect that new legislation and international treaties will be

 Bragg, supra note 28.   Harris v. State, 152 S.W.3d 786 (Tex. App. 2004).

33 34

BARFIELD PRINT.indd 11

13/11/2018 11:52

12  Research handbook on the law of virtual and augmented reality needed to address the increasing incidences of virtual crime occurring across international boundaries. Further, intellectual property law will have a major role to play in virtual worlds, and particularly in rights for virtual avatars. For instance, virtual reality software (object and source code) such as software for other applications is primarily protected under copyright, and database rights may also be available. Historically, computer programs were not effectively protected by copyrights because computer programs were not viewed as a fixed, tangible object, and in the early days object code was viewed as a utilitarian good produced from source code rather than as a creative work. Due to lack of precedent, this outcome was reached while deciding how to handle computer program copyright. In 1974 the Commission on New Technological Uses of Copyrighted Works (CONTU) was established, which decided that computer programs, to the extent that they embody an author’s original creation, are proper copyright subject matter. In 1980, the United States Congress added the definition of “computer program” to 17 U.S.C. § 101 (the Copyright Act) and amended 17 U.S.C. § 117 to allow the owner of the program to make another copy or adaptation for use on a computer. In the US this legislation, plus court decisions such as Apple Computer, Inc. v. Franklin Computer Corp in 1983, clarified that the Copyright Act gave computer programs the copyright status of literary works.35 In response, many companies began to claim that they “licensed” but did not sell their products, in order to avoid transfer of rights to the enduser via the doctrine of first sale.36 These software license agreements are often labeled as EULAs. In 1998 the US Congress passed the Digital Millennium Copyright Act (DMCA) (discussed above in Amaretto Ranch Breedables), which criminalized evasion of copy protection (with certain exceptions) and destruction or mismanagement of copyright management information, but included a clause to exempt ISPs from liability of infringement if one of their subscribers infringed. In addition, the DMCA extended protection to those who copied a program for maintenance, repair, or backup as long as these copies were “destroyed in the event that continued possession of the computer program should cease to be rightful.”37 In particular, “owners of copies” may make additional copies for archival purposes, “as an essential step in the utilization of the computer program,” or for maintenance purposes. Furthermore, “owners of copies” have the right to resell their copies under the first sale doctrine and 17 U.S.C. § 109. These rights only apply to “owners of copies.” Most software vendors claim that their products are “licensed, not sold,” thus sidestepping 17 U.S.C. § 117 (which relates to limitations on exclusive rights and computer programs). American courts have taken varying approaches when confronted with these

  Apple Computer, Inc. v. Franklin Computer Corp., 714 F.2d 1240 (3d Cir. 1983).   See Step-Saver Data Systems, Inc. v. Wyse Technology, 939 F.2d 91 (1991 U.S. App. 16526). The first-sale doctrine is a legal concept playing an important role in US copyright and trademark law by limiting certain rights of a copyright or trademark owner. The doctrine enables the distribution chain of copyrighted products, library lending, giving, video rentals, and secondary markets for copyrighted works (for example, enabling individuals to sell their legally purchased books or CDs to others). In trademark law, this same doctrine enables reselling of trademarked products after the trademark holder put the products on the market. The doctrine is also referred to as the “right of first sale,” “first sale rule,” or “exhaustion rule.” 37   17 U.S.C. § 117 (1998). 35 36

BARFIELD PRINT.indd 12

13/11/2018 11:52

The law of virtual reality and increasingly smart virtual avatars  13 software license agreements. For example, in MAI Systems Corp. v. Peak Computer, Inc.38 and Microsoft v Harmony,39 different federal courts held that “licensed, not sold” language in a EULA was effective. However, other courts have held that no brightline rule distinguishes mere licenses from sales. So, given the differences in court decisions, how will virtual avatars be viewed? Will they be licensed, or the property of those who create them? Under some EULAs, the owner of the avatar retains copyright to the avatar. Of course, the question arises: As avatars become smarter and create other avatars, who then is the owner?

4.  VIRTUAL ENVIRONMENTS, GAMES, AND AVATARS One type of virtual environment that is accessed by millions of users, and that has generated significant interest from legal scholars, is the massively multiplayer online roleplaying game (MMORPG).40 Once a player enters a MMORPG, they engage in a variety of activities with other players who are accessing the game in the same way from all over the world. MMORPG developers are in charge of supervising the virtual world and guarantee the continuing interest of players by offering an updated set of tasks and activities to perform in the virtual environment.41 Many MMORPGs have been designed for profit: A player must either purchase the client software or pay a monthly fee in order to access the virtual world. An interesting feature of a MMORPG is that it allows its participants to design a virtual representation of their identity which is displayed in the online virtual environment. As noted throughout this chapter, such an identity is termed an avatar. Generally, the term “virtual avatar” (or avatar) is used to describe the simulation of a graphical form representing a particular person in a virtual environment. The most sophisticated avatars can become a sort of visual and cognitive prosthesis, representing an extension of self in the virtual world, or what the virtual environment visitor would like to be, or appear to be, in the virtual world. Virtual avatars may also represent the actions of a user, different aspects of a user’s persona, or the user’s social status in the virtual environment. A virtual avatar can take on almost any form, such as a realistic representation of the human that owns or created the avatar, another person’s identity (such as a living or deceased actor or historic figure), an animal, or even a mythical creature. Interestingly, the form the avatar takes can implicate several areas of law, such as copyright, trademark, and right of publicity.

  MAI Systems Corp. v. Peak Computer, Inc., 991 F.2d 511 (9th Cir. 1993).   Microsoft v. Harmony Computers & Electronics, 846 F. Supp. 208 (E.D.N.Y. 1994). 40   Caroline Bradely and A. Michael Froomkin, Virtual Worlds, Real Rules, 49 N.Y.L. SCH. L. REV. 103, 121 (2004–5). 41   Cory Ondrejka, Escaping the Gilded Cage: User Created Content and Building the Metaverse, 49 N.Y.L. SCH. L. REV. 81 (2004–5). See also Nicholas Yee, The Psychology of MMORPGs: Emotional Investment, Motivations, Relationship Formation, and Problematic Usage, in Avatars at Work and Play: Collaboration and Interaction in Shared Virtual Environments (R. Schroeder & A. Axelsson eds, London: Springer-Verlag 2005), available at http://www.nickyee.com/daedalus/ archives/02_04/Yee_Book_Chapter.pdf. 38 39

BARFIELD PRINT.indd 13

13/11/2018 11:52

14  Research handbook on the law of virtual and augmented reality How easy is it to create a virtual avatar? Commercial software has been designed to allow people to create their own interactive, emoting 3D avatar using photographs of their individual faces, and their own unique voices, as templates. Further, when a person chats in a 3D online world or plays an online computer game, they are operating a synthetic character or avatar. What makes for an interesting and effective avatar depends on the purpose for which the avatar is used. In the case of a virtual world where communication is important, facial features and expressiveness must be well supported. Additionally, in the case of action games, the physics of the virtual world and interaction within the world must be well supported. A recent development in virtual avatars is that they are getting smarter.42 Capable of performing a range of tasks, virtual avatars can write poetry, play chess, compose music, and portray a range of emotions and facial expressions. In electronic commerce, avatars are forming contracts;43 in the field of entertainment they are replacing actors;44 and in online games they are interacting with humans and other virtual avatars.45 In addition, in medicine, virtual avatars are helping to train medical students by playing the role of patient. For example, the Virtual Standardized Patient is an avatar that interacts with medical practitioners in much the same way as an actor would if hired to play the role of patient.46 The Virtual Standardized Patient uses natural language, emotion, behavior modeling, and composite facial expression and lipshape modeling to produce a natural patient–practitioner dialogue.47 When avatars interact with humans, facial expressions are key for communicating emotions in face-to-face conversation made simultaneously with speech. In some virtual avatar designs, collaborative virtual environments may force the user to explicitly set avatar emotions after they have entered text or voice input. However, some researchers are investigating a procedure based on an expert system that can be used to parse emotive expressions so that these emotions can be automatically displayed on the corresponding virtual avatar’s appearance.48 In many online games, a user must input avatar body language and facial expressions via key presses, which means it is almost impossible for users to chat and emote at the same time. To appear realistic, an avatar must react as humans do when communicating with each other, and facial expressions are a step toward designing “humanlike” avatars. 42   Norman I. Badler, Rama Bindiganavale, Juliet Bourne, Jan Allbeck, Jianping Shi & Martha Palmer, Real Time Virtual Humans, Center for Human Modeling and Simulation, Department of Computer and Information Science, University of Pennsylvania, available at www.cis.upenn. edu/~badler/bcs/Paper.htm. 43   See Jeff C. Dodd and James A. Hernandez, Contracting in Cyberspace, 1998 SMU Computer L. J. 1, 12 (1998). 44   Joseph J. Beard, Clones, Bones, and Twilight Zones: Protecting the Digital Persona of the Quick, the Dead and the Imaginary, 16 Berkeley Tech. L. J. 1165 (2001). 45   See generally Avatars in Networked Virtual Environments (Tolga K. Capin et al, eds 1999); see also generally Peter Plantec, Virtual Humans: A Build-It-Yourself Kit, Complete With Software and Step-by-Step Instructions (American Management Association; Bk&CD-Rom edition 2004). 46   Robert C. Hubal, Paul N. Kizakevich, Curry I. Giunn, Kevin D. Merino & Suzanne L. West, The Virtual Standardized Patient: Simulated Patient-Practitioner Dialogue for Patient Interview Training, 70 Stud Health Technol Inform. 133–8 (2000). 47   Id. 48   Jimena Olveres, Mark Billinghurst, Jesus Savage & Alistair Holden, Intelligent, Expressive Avatars, in Proceedings of the First Workshop on Embodied Conversational Characters (WECC 1998), Lake Tahoe, California, (October 12–15, 1998).

BARFIELD PRINT.indd 14

13/11/2018 11:52

The law of virtual reality and increasingly smart virtual avatars  15 Additionally, virtual environments can be designed for single inhabitants—such as a solo flight trainee—or for many, simultaneous participants. When a virtual environment supports multiple users, it can give rise to a virtual community. It has been estimated that some of the millions of people who visit virtual worlds spend more time in the virtual environment than the real world. These people are not just passively viewing the environment. They, or their virtual representatives, are interacting with other people or with virtual avatars of increasing intellectual capabilities. Further, people who spend significant amounts of time in virtual environments are doing more than playing video games; they are creating virtual worlds where they can assume identities, build wealth and social status, and generally participate in creating new worlds. An important issue in online roleplaying games is whether the licensor or participant owns the intellectual property (such as copyright to a virtual avatar) created while the game is being played.49 In an early version of Second Life, the terms of service agreement gave residents of the virtual world the right to retain full intellectual property protection for the digital content they create in the game, including avatar characters, clothing, scripts, textures, objects, and designs. Such rights have real-world consequences for the objects created in the virtual world. More recently, Second Life has indicated that users have “the non-exclusive, unrestricted . . . [full use of] . . . all or any portion of your User Content.”50 An example of a legal dispute involving virtual worlds is Eros LLC v. Doe,51 in which a copyright and trademark infringement action was pursued against the defendant. Eros created adult entertainment products in the virtual world of Second Life, which it sold to other Second Life users for Linden dollars (which can be converted into US dollars). Another Second Life user, with the avatar name “Volkov Catteneo,” allegedly copied Eros’s products and sold them to other Second Life users without authorization. The products at issue are beds that allow users of Second Life to have virtual sex in them.52 To protect their identity, Eros initiated a “John Doe” lawsuit for copyright and trademark infringement in federal court in Florida. The amended complaint alleged that the company Leatherwood acted in concert with unknown parties to make and sell unauthorized copies of Eros’s products. Eros reached a settlement with the defendant (known only as Second Life avatar Volkov Catteneo).53 The terms include an agreement that Leatherwood will not do any more copying; as noted, Eros had alleged that the defendant sold unauthorized copies of the company’s popular “SexGeb” avatar-animation products.54 In summary, what one can conclude from the above cases is that as people and virtual avatars spend significant amounts of time in virtual reality, more cases will be litigated across a broad range of legal topics, from intellectual property to criminal law, and from contracts to torts. Given the increased use of virtual avatars in virtual environments

49   Some foreign courts have begun to accept the notion of virtual property; for example, a Beijing court ordered the restitution of one player’s stolen virtual weapons. 50   See www.daz3d.com/forums/discussion/29511/ot-linden-lab-second-life-will-now-own-youruploaded-content. 51   Eros LLC v. Doe, United States District Court, Middle District of Florida (2007). 52   Id. 53   Id. 54   Id.

BARFIELD PRINT.indd 15

13/11/2018 11:52

16  Research handbook on the law of virtual and augmented reality for tasks such as psychotherapy, teaching, and electronic commerce, future causes of action could be directed at the virtual avatars themselves. If avatars continue to gain in intelligence and autonomy a host of compelling legal issues will be raised. For example, would intelligent avatars be able to bring forth claims involving their own civil liberties?55 And just what civil liberties would be awarded to intelligent virtual avatars? Professor George Wright has discussed the issue of equal protection under the law in the context of “enhanced humans,” concluding that “if there develops a typically unbridgeable gulf separating groups of contemporaries, we must adopt a substantially realistic understanding of equal protection that involves significant resource and opportunity transfers.”56 Wright’s comments were directed at the differences between enhanced and unenhanced humans; however, the possibility of “enhanced” intelligent avatars raises a host of new issues concerning equal protection under the law. In the context of humans, it may be possible to provide those requesting upgrades access to the appropriate technology. However, if intelligent avatars surpassed humans in intelligence, would technology be available to upgrade the humans to the level of the intelligent avatars? And if an intelligent avatar gained a level of intelligence such that they were superior to humans, would humans then be able to bring forth an equal protection claim against intelligent avatars? To best serve humanity’s interests, public policy may benefit by granting intelligent entities legal rights—if for no other reason than that they could then be regulated.

5.  TOWARD MORE INTELLIGENT VIRTUAL AVATARS The present format for the protection of virtual avatars’ rights is based on determining who their owner is and then analyzing that person’s rights with respect to the avatar or the avatar’s actions.57 In this model, the rights protected are those of the owner and not those of the virtual avatar (which lacks legal person status). However, as virtual avatars gain in intelligence and create works independent of human input, this analysis may become outdated. In this scenario, avatars themselves may need legal protection based on their own actions and actions directed against them. If avatars do continue to gain in ­intelligence and people spend more time in virtual worlds interacting with them, significant legal and policy issues will arise.58 For example, since many virtual worlds are created 55   George Wright, Personhood: 2.0: Enhanced and Unenhanced Persons and the Equal Protection of the Laws, 23 Quinnipiac L. Rev. 1047 (2005). A legal person, as opposed to a natural person, enjoys many of the rights and obligations of individual citizens, such as the ability to own property, sign binding contracts, and pay taxes; but they do not retain all the rights of a natural person, for example, they do not have the right to vote or hold public office. See generally Curtis M. Vazquez, Direct v. Indirect Obligations of Corporations Under International Law, 43 Colum. J. Transnat. L. Rev. 927, 944 (2005). A legal person (or artificial person), as opposed to a natural person, enjoys many of the rights and obligations of 657. 56  Wright, id at 1095. 57   See generally Woodrow Barfield, Issues of Law for Software Agents within Virtual Environments, 14 Presence: Teleoperators and Virtual Envs, 741–48 (2005). 58   Michael B. Sapherstein, The Implications of Virtual Reality Games for Tort Lawyers, B.C. Intell. Prop. & Tech. F. 112106 (1996) (discussing the tort consequences of using virtual reality

BARFIELD PRINT.indd 16

13/11/2018 11:52

The law of virtual reality and increasingly smart virtual avatars  17 by private companies for their subscribers and are thus controlled by the game’s creators, should the participants, the creators, or the virtual avatars (or some combination) set and control the permissible actions in the virtual environment?59 In contrast, should the users of the virtual environment set the rules of social interactions, the physical laws that govern the virtual world, or the laws and statutes that people and avatars live by? And as virtual avatars become more autonomous from human input and decisionmaking, and begin to discern their own algorithms, how should such entities be treated by the law?60 The field of artificial intelligence has provided the algorithms and techniques that have led to intelligent actions by virtual avatars. The software and algorithms that control virtual avatars, and artificial entities in general, are getting more sophisticated and “smarter.” As some commentators have argued, the smarter they get, the more the current law will be strained when deciding how to account for their actions.61 In general, advances in algorithms have resulted in levels of creativity exhibited by artificial entities that traditionally were considered only within the domain of humans. This raises several perplexing questions—for example, can a virtual avatar be an author or an inventor, own and sell intellectual property, or be liable for its actions?62 Many software programs which result in creative output use either knowledgebased systems, genetic algorithms, neural networks, or, more recently, machine learning techniques. Neural networks differ from traditional artificial intelligence applications because they do not require explicit symbolic representations to solve problems. Instead, they process and store information as patterns to represent information. Specifically, the knowledge contained within a neural network is represented by the connection strengths between processing elements in the network and the mutual reinforcement or inhibition of elements in the network by other elements. As noted previously, one area where neural networks have been used to create virtual avatars which display intelligent behavior is in the design of facial expressions. For intelligent avatars to be able to act as alter egos of their human owner, they may need to incorporate a high degree of similarity with their owner, including facial expressions and body posture. Neural networks can be trained to recognize and reproduce patterns such as those associated with facial expressions, and to produce such patterns based on external stimuli. Genetic algorithms are another technique used to create “intelligent acting” avatars in virtual reality. Generally, genetic algorithms are search procedures that use the principles of natural selection and genetics to solve problems. Genetic algorithms use evolutionary techniques based on optimization to develop a solution to a problem. The basic operagames such as reports that users may suffer from side effects including vertigo and dizziness after exposure to virtual environments). 59   Richard A. Bartle, Virtual Worldliness: What Imaginary Asks of the Real, 49 N.Y.L. Sch. L. Rev. 19 (2004–5) (the elements of an end-user license agreement could effectively limit the range of permissible actions allowed in a virtual environment). See generally Lawrence Lessig, The Limits in Open Code: Regulatory Standards and the Future of the Net, 14 Berkeley Tech. L. J. 759, 763 (1999). 60   See generally Curtis E.A. Karnow, Liability for Distributed Artificial Intelligences, 11 Berkeley Tech. L. J. 147 (1996). 61   Id. 62   See Pamela Samuelson, Allocating Ownership Rights in Computer-Generated Works, 47 Pitt. L. Rev. 1185 (1985–6) (providing a comprehensive overview of issues associated with whether a computer can be an author); see generally Karnow, supra note 60.

BARFIELD PRINT.indd 17

13/11/2018 11:52

18  Research handbook on the law of virtual and augmented reality tion of a genetic algorithm is fairly straightforward. First, a population of possible solutions to a problem is developed, then the better solutions are recombined with each other to form some new solutions. Finally, the new solutions are used to replace the poorer of the original solutions and the process is repeated. In addition to this, many avatars are designed to display appropriate social behavior in reaction to other avatars and people in a virtual environment. Genetic algorithms are useful for designing avatars that can display a range of social behaviors. The diversity of genetic customization is important in creating a unique avatar in a virtual world, and in being a part of a large, diverse community. To use a genetic algorithm to create various facial expressions, the design methodology of the avatar includes identifying variations in the parameters used in the computer code which control facial expressions, setting ranges for these parameters, and placing them into an array, which can be manipulated in a variety of ways. The array is called the genotype. Every unique avatar designed using genetic algorithms will have a different genotype. The gene ranges will provide an overall genetic space within which all possible avatars can exist. These genes will affect, for example, body shapes, colors, motions, facial proportions, and walking styles of an avatar. Another type of computing paradigm which has resulted in intelligent behavior for virtual avatars is an expert, or knowledge-based, system. Knowledge-based systems are those in which the computer algorithms are able to “learn” which solutions are retainable or usable by a series of comparisons with previously stated material. This type of programming is often referred to as an “expert system” because the system is based on imitating the methods of particular human practitioners, or experts, within a particular domain. As with neural networks, an expert system approach has been used to model facial expressions for virtual avatars. As these techniques mature, evolving ever more sophisticated avatars, the potential for legal issues will inevitably increase. The New Zealand company Soul Machine has perhaps created the most realistic avatar to date, Nadia, using advanced modeling techniques to interact with customers on an emotional level.63 Their design uses an intelligent model of how the human brain works and machine learning techniques to read, predict, and emulate human emotional responses. Further, the University of Southern California and Pinscreen have developed a method of generating photorealistic static avatars using a single input image and a “deep convolutional neural network.”64 Additionally, Loom.ai has developed a method to create realistically animated avatar faces from a single static image.65 Finally, Stanford’s Virtual Human Interaction Lab has been doing work for more than a decade utilizing virtual environments and avatars to study the science of human behavior. If the avatar’s face mimics that of a celebrity, would this be actionable under a right of publicity statute? Consider California’s common law and statutory right of publicity law. Generally speaking, the right of publicity in California protects against unauthorized uses of a person’s name or likeness for commercial and certain other exploitative 63   EllyStrang, Soul Machines Unveils Its FirstEmotionally Intelligent,LifelikeAvatar,idealog,http:// idealog.co.nz/tech/2017/02/soul-machines-unveils-its-first-emotionally-intelligent-lifelike-avatar. 64   S. Saito, L. Wei, L. Hu, et al, Photorealistic Facial Texture Inference using Deep Neural Networks, https://arxiv.org/pdf/1612.00523v1.pdf. 65   Ben Popper, This Tech Generates Realistic Avatars from a Single Selfie, The Verge, www. theverge.com/2016/12/13/13919766/loomai-vr-avatar-digital-character-3d-visual-effect.

BARFIELD PRINT.indd 18

13/11/2018 11:52

The law of virtual reality and increasingly smart virtual avatars  19 purposes. California has two systems of right of publicity law—a statute and a common law right—with some differences in coverage between the two. In California, the statutory right of publicity protects a person’s right in his or her name, voice, signature, photograph, and likeness.66 The term “voice” applies only to a person’s actual voice, not to imitations.67 However, as noted presently, the common law right of publicity might apply to voice imitators (and possibly an algorithm mimicking a well-known voice). The term “likeness” is the most difficult of the protected categories to precisely define. Courts have used the “readily identifiable” test to conclude that drawings, if sufficiently detailed, can constitute a “likeness.”68 In another case, the court ruled that a robot, if sufficiently detailed, could be a “likeness.”69 Less detailed robots (and virtual avatars?), though, may fall short of the “likeness” mark.70 Under California’s common law right of publicity, courts generally describe a four-step test, in which a plaintiff must allege: the defendant’s use of plaintiff’s “identity”; the appropriation of plaintiff’s name or likeness to defendant’s advantage, commercially or otherwise; lack of consent; and  resulting injury.71 Though the second prong of the standard four-step test mentions “name or likeness,” courts have held that the common law right is actually much broader.72 In California, courts have interpreted “identity” broadly, in such a way that it covers more uses than does the statutory right of publicity. For example, imitating someone’s voice is not a violation of the California statute, but it may violate the common law right.73 A robot can constitute a common law violation, even if not sufficiently detailed to violate the statute.74 Unlike the statute, the common law right is not explicitly limited to commercial uses of a plaintiff’s identity. However, the “less commercial” a use, the more that First Amendment concerns come into play. Purely commercial speech, such as advertising, does nothing more than “propose a commercial transaction”; if a defendant’s use falls outside the realm of the purely commercial, California’s common law right of publicity is less likely to apply. There are limits to both the common law and the statutory right of publicity. A right of publicity claim (either statutory or under the common law) fails if it is too similar to a copyright claim; in such a case, the state right of publicity law is preempted by federal copyright law. For example, in Laws v. Sony Music, Sony licensed one of the plaintiff’s songs and sampled it in a new recording.75 The plaintiff tried to bring a right of publicity claim, but the court ruled that Sony’s use of a licensed recording fell under copyright law, thus preempting the state claim. Generally speaking, if the allegedly infringing use of a 66  Wright, supra note 55. Professor George Wright has discussed the issue of equal protection under the law in the context of “enhanced humans,” concluding that “if there develops a typically unbridgeable gulf separating groups of contemporaries, we must adopt a substantially realistic understanding of equal protection that involves significant resource and opportunity transfers.” 67   See Midler v. Ford, 849 F.2d 460, 463 (9th Cir. 1988). 68   Newcombe v. Adolf Coors Co., 157 F.3d 686, 692–93 (9th Cir. 1998). 69   Wendt v. Host Intern., Inc., 125 F.3d 806, 810 (9th Cir. 1997). 70   White v. Samsung, 971 F.2d 1395, 1397 (9th Cir. 1992). 71   Id at 1397. 72   See Abdul-Jabbar v. General Motors, 85 F.3d 407, 413–14 (9th Cir. 1996). 73   See Waits v. Frito-Lay, 978 F.2d 1093, 1098–100 (9th Cir. 1992). 74  White, supra note 70. 75   Laws v. Sony Music, 294 F.Supp.2d 1160 (C.D. Cal. 2003).

BARFIELD PRINT.indd 19

13/11/2018 11:52

20  Research handbook on the law of virtual and augmented reality person’s identity primarily involves use of copyrighted work, there is a chance that the state law claim will be preempted. The First Amendment also limits the extent to which rights of publicity can limit speech about matters of public interest. As one court put it, “[u]nder the First Amendment, a cause of action for appropriation of another’s name and likeness may not be maintained against expressive works, whether factual or fictional.”76 As mentioned previously, the California statute contains exceptions for uses related to news, public affairs, sports, and politics and thus could relate to a virtual avatar serving as a news anchor or other TV personality. Courts often focus on this statutory safe harbor, rather than the First Amendment directly, when confronting statutory right of publicity claims.77 The First Amendment is more often directly relevant in common law right of publicity cases, since there is no statutory safe harbor. But since cases often involve both common law and statutory claims, the First Amendment analyses often cover both the statute and the common law.

6.  VIRTUAL AVATARS AND WORKS OF AUTHORSHIP As indicated earlier, computers using methods of artificial intelligence have been programmed to compose music, write poetry, and write parts of a book. These are all areas deemed to reflect a high level of human creativity, and which result in copyrightable works of authorship. It has been argued that, through a process of copyright reform, the requirement of creativity should be modified to include a contextual and cultural analysis.78 Under this system, it may be easier for artificially intelligent generated works to pass the requirement of creativity. Once virtual avatars create works of authorship, especially if they do so independent from human input, traditional copyright notions of authorship and originality will need to be addressed. This raises the question of whether the copyright law as currently enacted can adequately address issues of authorship in a world of increasingly intelligent artificial entities. Generally, if a work is completely attributable to an intelligent avatar, then that work will be outside the ambit of federal copyright law.79 In contrast, if a human author can be associated with the work of an intelligent avatar, copyright law as it currently exists may adequately account for the output generated by the avatar. Under the Copyright Act, the author of a work is the initial owner of the copyright in it, and may exploit the work herself or transfer some or all of her rights in that work to others.80 The author is generally the person who conceives of the copyrightable expression and fixes it, or causes it to be fixed, in a tangible form. The Ninth Circuit, in MAI Systems Corp. v. Peak Computer, Inc., held that the loading of software into a computer’s random

  Daly v. Viacom, 238 F.Supp. 2d. 1118, 1123 (N.D. Cal. 2002).   See, e.g., Gionfriddo v. Major League Baseball, 94 Cal. App. 4th 400, 415–17 (Cal. Ct. App. 2001). 78   Julie Cohen, Creativity and Culture in Copyright Theory, 40 U.C. Davis L. Rev. 1151–205 (2007), http://scholarship.law.georgetown.edu/cgi/viewcontent.cgi?article=1062&context=facpub. 79   See generally Timothy L. Butler, Can a Computer Be an Author? Copyright Aspects of Artificial Intelligence, 4 Hastings Comm. & Ent. L. J. 707 (1982). 80   17 U.S.C. § 102(a) (2005). 76 77

BARFIELD PRINT.indd 20

13/11/2018 11:52

The law of virtual reality and increasingly smart virtual avatars  21 access memory was sufficiently permanent for it to be deemed fixed.81 Given this decision, an avatar that creates a work may be thought to have fixed it at the moment of creation.82 Therefore, in a copyright dispute involving a virtual avatar the issue for a court will not be whether the work is fixed, but whether the virtual avatar or another party “conceived” of the work and can serve as the author. In some cases an avatar’s owner or programmer may be so far removed from the avatar’s output that they may not have any knowledge that the output exists, or even recognize that it resulted from their original input. In such a case, should this person be considered an author? Given the increased complexity of virtual avatars, it is pertinent to ask: How would this decision serve the policy of encouraging authors to create? Are there any existing doctrines under copyright law where a copyright can vest in an entity that did not conceive the work or fix it in a tangible medium of expression? The answer is yes; for example, the “work made for hire” doctrine is an important exception to the rule that the party who conceived of the idea is the “legal” author of a work. When a work is made for hire, within the meaning of the Copyright Act, the employer or commissioning party who pays for the creation of the work is deemed the author, rather than the employee who may actually have conceived of the work and fixed the expression.83 One possible way to solve the problem of ownership of the intellectual property created by intelligent avatars is to always deem them as works for hire, in which case the employer or commissioning party would be the author under the Copyright Act. However, can a virtual avatar serve as an employee? Conversely, could a programmer be considered the employer of an intelligent avatar? If yes, then why not assume that as an employee the intelligent avatar would have rights, either contractual or under the Copyright Act, to the intellectual property they created? Further complicating the issue, authorship, if indeed granted to an intelligent avatar, need not indicate copyright ownership.84 Other legal theories may also be useful for thinking about the rights of virtual avatars. For example, could the intellectual property created by avatars be considered a joint work between the avatar, programmer, or avatar owner?85 The US Copyright Act defines a joint work as “a work prepared by two or more authors with the intention that their contributions be merged into inseparable or interdependent parts of a unitary whole.”86 The programmer’s contribution to the joint work would be the algorithms to direct the avatar’s behavior and the programming required to create the avatar’s appearance. The owner’s contribution would be the input directing the avatar’s output. The avatar’s contribution would vary from significant to less meaningful depending on the amount of input supplied by the programmer or owner and the algorithms used to design the avatars.

 MAI, supra note 38.   Id. at 519. 83   See generally Darin Glasser, Copyrights on Computer-Generated Works: Whom, If Anyone, Do We Reward? 2001 Duke L. & Tech. Rev. 24 (2001). 84   Jane Ginsburg, The Concept of Authorship in Comparative Copyright Law, 53 Depaul L. Rev. 1063 (2003). 85   See generally Tarcisio Queiroz Cerqueira, Some Common and Civil Thoughts on Computer Generated Works, https://libros-revistas-derecho.vlex.es/vid/some-thoughts-computer-generatedworks-107749; see 17 U.S.C. § 101 (1976) (defining a “joint work”). 86   17 U.S.C. § 101 (1976) (defining a “joint work”). 81 82

BARFIELD PRINT.indd 21

13/11/2018 11:52

22  Research handbook on the law of virtual and augmented reality For example, if using techniques such as neural nets, genetic algorithms, or deep learning, the avatar could make significant contributions to a joint work. For a joint work under the Copyright Act, the authors are considered coowners of a single copy of the work. Thus, if a joint work was found, the programmer and avatar would each own an undivided interest in the copyright. But what if the avatar is learning within the virtual environment, and creates an output completely independent of the programmer’s original effort? Should the court then view the avatar’s output as an original work of authorship, or as a derivative of the programmer’s original input? If so, who would the court consider to be the author of the avatar’s output? Would the court deem the work original if created by an avatar, and thus award a copyright to the avatar? Under the US Copyright Act, copyright subsists “in original works of authorship fixed in any tangible medium of expression, now known or later developed, from which they can be perceived, reproduced or otherwise communicated, either directly or with the aid of a machine or device.”87 For a work to be original, the author must have engaged in some intellectual endeavor of their own and not just have copied from a preexisting source, and the work must exhibit a minimal amount of creativity.88 In the context of intelligent avatars, an important issue is whether the “works” of an avatar can be considered original. If avatars create original works of authorship eligible for copyright protection, who will the court determine to be the author of such works: the original programmer(s), the employer of the programmer, the avatar’s owner, the avatar—or, as discussed above, will the work be considered a joint work under the copyright law with multiple owners?89 The issue of whether computer-generated output can be eligible for copyright protection has received some attention in the past, with some commentators concluding that a computer can be an author under the Copyright Act,90 and some suggesting extending copyright protection though not necessarily authorship.91 For example, Harvard law professor Arthur Miller argues against copyright protection for artificially intelligent entities since they need no incentive to produce a work of authorship.92 The process of creating a work involving a virtual avatar involves the effort of a programmer to create the avatar, the software and algorithms used to create the avatar’s appearance and behavior, and a computer (or nowadays the cloud) used to store the code used to design the avatar. The software can consist of rules that allow little or no autonomous actions by the avatar, or can consist of neural nets or genetic algorithms (or deep learning techniques) which allow the avatar to learn and act in significantly different ways than the original set of parameters used to design the   17 U.S.C. § 102(a) (subject matter of copyright).   Id. 89   See generally Tal Vigderson, Hamlet II: The Sequel? The Rights of Authors vs. ComputerGenerated “Read-Alike” Works, 28 Loy. L. A. L. Rev. 401 (1994) (discussing whether a romance novel written by an AI that was programmed to mimic author Jacqueline Susann might inappropriately copy Susann’s style). 90   See generally Karl F. Milde, Jr., Can a Computer Be an “Author” or an “Inventor?” 51 J. Pat. Off. Soc’y 378 (1969). 91   See generally Arthur R. Miller, Copyright Protection for Computer Programs, Databases, and Computer-Generated Works: Is Anything New Since CONTU? 106 Harv. L. Rev. 977 (1993) (concluding that AIs should not be authors because computers need no incentive to produce their output). 92   Id. 87 88

BARFIELD PRINT.indd 22

13/11/2018 11:52

The law of virtual reality and increasingly smart virtual avatars  23 avatar. In order to determine whether an avatar can be an author and receive copyright protections for its works, the interests of the programmer, employer, and avatar will need to be addressed. Granting authorship rights to an intelligent avatar will be difficult under the current copyright law.93 One reason for this is enforceability of the rights enumerated under the Copyright Act. Would an avatar be capable of enforcing such rights, or have standing to initiate an action for copyright infringement? Under current law, the answer is no. Further, awarding copyright protection to an avatar would imply that the avatar can have ideas that led to original works of authorship.94 What separates avatars that act with intelligence from avatars which are designed to perform a limited set of actions strictly under human control is the ability of the “intelligent avatar” to apply existing knowledge to a new set of facts or problems. Here the relevant inquiry into whether the avatar’s actions translate into an original work of authorship is whether the avatar is simply reinterpreting another author’s work, and whether the avatar’s output is completely dependent on the programmer’s instructions. The standard for what constitutes an original work under the Copyright Act has been decided by the US Supreme Court in the seminal case Feist Publ’ns, Inc. v. Rural Tel. Serv. Co., Inc.95 Discussing the requirement for originality, the Supreme Court found that telephone white page listings did not satisfy the originality requirement because they lacked minimal creativity.96 The Court noted that the author’s “selection and arrangement of the facts could not be so mechanical or routine as to require no creativity whatsoever.”97 As the Court discussed, “[o]riginal, as the term is used in copyright, means only that the work was independently created by the author, and that it possesses some minimal degree of creativity.”98 Under this analysis, a court may determine that an avatar using algorithms is simply performing in a mechanical or routine manner. In this case, the avatar could not receive copyright protection for its work, due to lack of creativity. However, consider an avatar with the capability to learn and respond to events in the virtual environment. In this case, the problem solving would be far from mechanical or routine. Even so, before copyright protection is awarded to the output of an avatar, the work must be deemed original. This does not, in itself, seem to be an obstacle for an intelligent avatar. However, the avatar would have to be deemed an author. As discussed throughout this section, this is a more difficult bar to overcome, given the current copyright statute. Interestingly, there is some precedent for the assertion that an “author” need not be a human being (see the case discussed in the following paragraph). As noted previously, under the work for hire doctrine a corporation may be deemed the author of a work, although this conclusion seems to conflict with some case law. In an interesting and somewhat odd case, the issue of whether a nonhuman can be an author was addressed by the Ninth Circuit in Urantia Found. v. Maaherra.99 This case  Milde, supra note 90.   See Butler, supra note 79, at 726–33. 95   Feist Publ’ns, Inc. v. Rural Tel. Serv. Co., Inc., 499 U.S. 340 (1991). 96   Id at 362. 97   Id. 98   Id at 345. 99   Urantia Found. v. Maaherra, 114 F.3d 955 (9th Cir. 1997). 93 94

BARFIELD PRINT.indd 23

13/11/2018 11:52

24  Research handbook on the law of virtual and augmented reality involved a questionable claim that a superior being authored a particular work, but for our purposes here, the analysis the court used to decide the case offers an interesting insight into how the law might view authorship rights for intelligent avatars. The case involved a copyright dispute between parties who believed the copyrighted work, the “Urantia Book,” was authored by celestial beings and only transcribed, compiled, and collected by “mere mortals.”100 The plaintiff, Urantia Foundation, claimed that the defendant, Maaherra, infringed the Foundation’s copyright when she distributed a computerized version of the Urantia book on disk.101 Maaherra conceded copying, so the issue before the court was whether the Foundation owned a valid copyright in the book. Complicating matters, both parties believed that the words in the book were “authored” by nonhuman spiritual beings described in terms such as the “Divine Counselor,” the “Chief of the Corps of Superuniverse Personalities,” and the “Chief of the Archangels of Nebadon.”102 These spiritual entities were claimed to have delivered the teachings that were eventually assembled in the book, through a patient of a Chicago psychiatrist.103 A threshold issue in this case was whether the work, because it was claimed to embody the words of celestial beings rather than human beings, was copyrightable at all.104 In Feist Publications, Inc. v. Rural Television Service Co., Inc., the court—in discussing a threshold requirement for copyright—said, “To qualify for copyright protection, a work must be original to the author.”105 The core statute from the Copyright Act provides: “[C]opyright protection subsists . . . in original works of authorship fixed in any tangible medium of expression . . . from which they can be perceived, reproduced, or otherwise communicated, either directly or with the aid of a machine or device.”106 As the court reasoned, “[o]riginal, as the term is used in copyright, means only that the work was independently created by the author (as opposed to copied from other works), and that it possesses at least some minimal degree of creativity.”107 Maaherra claimed that there can be no valid copyright in the book because it lacked the requisite ingredient of human creativity, and that therefore the book was not a “work of authorship” within the meaning of the Copyright Act.108 Interestingly, nowhere in the US copyright laws is there an express requirement for “human” authorship, and considerable controversy has arisen in recent years over   Id at 956.   Id at 958. 102   Id at 956. 103   Id. 104  See Oliver v. Saint Germain Found., 41 F.Supp. 296 (S.D.Cal. 1941). In Oliver, the plaintiff’s religious text proclaimed that the facts contained in the text had come straight from a spirit, and that the spirit was the author of the history in the text. The plaintiff (unsuccessfully) claimed copyright protection in the divine revelations themselves, and in the methods of spiritual communication, rather than in the plaintiff’s specific selection or arrangement of these divine revelations. The defendant in Oliver had not copied that arrangement and selection, but simply had written another text using the same divine “facts.” The court in Oliver made it clear that, had the claim been that the selection and arrangement of the divine revelations had been infringed, the plaintiff’s copyright infringement claim might have had merit. 105  Feist, supra note 95. 106   17 U.S.C. § 102(a) (1976). 107   Feist, 499 U.S. at Clevenger v. Baker Voorhis & Co., 168 N.E.2d. 643 (N.Y. 1960). 108  Urantia, supra note 99 at 958. 100 101

BARFIELD PRINT.indd 24

13/11/2018 11:52

The law of virtual reality and increasingly smart virtual avatars  25 the copyrightability of computer-generated works. The Urantia court argued that the copyright law does not protect the “creations of divine beings,” but that the copyright laws protect some element of human creativity.109 The court stated: “At the very least, for a worldly entity to be guilty of infringing a copyright, that entity must have copied something created by another worldly entity.”110 For copyright purposes, the Urantia court reasoned, a work is copyrightable if copyright is claimed by the first human beings who compiled, selected, coordinated, and arranged the Urantia teachings, “in such a way that the resulting work as a whole constitutes an original work of authorship.”111 The court said that the party who was responsible for the creation of a tangible literary form that could be read by others could claim copyright for themselves as “authors.”112 They were responsible for the religious revelations appearing “in such a way as to render the work as a whole original.”113 Thus, notwithstanding the Urantia Book’s claimed nonhuman origin, the papers in the form in which they were originally organized and compiled by the members of the Contact Commission were at least partially the product of human creativity.114 The court reasoned that the papers did not belong to that “narrow category of works in which the creative spark is utterly lacking or so trivial as to be virtually nonexistent.”115 From the Ninth Circuit’s analysis in Urantia, one can summarize the decision as calling for a human author to find a copyrightable work even if the author did not conceive of the original work. However, the issue may not be settled as under § 201(b) of the copyright statute: If the work is a work for hire, a nonhuman entity such as a corporation can be deemed the author of a work. 17 U.S. Code § 201(b) (1976): In the case of a work made for hire, the employer of other person for whom the work was prepared is considered the author for purposes of this title, and unless the parties have expressly agreed otherwise in a written instrument signed by them, owns all of the rights comprised in the copyright.

This apparent conflict in the law will be even further revealed as artificial entities gain more intelligence, self-program, and make decisions independent from any human. Could an avatar be registered as a corporation, and thus be deemed the author of a work for hire, such as the work of another avatar? Some commentators suggest a “grantable” framework for personhood by which nonhuman entities could assume legal personhood status by the grant of someone already held as such.116 Under such a framework, ­personhood   Id at 958.   Id at 958. It would be interesting to consider whether an intelligent avatar would constitute a “worldly entity” under the court’s reasoning. 111   Id at 958; see also 17 U.S.C. § 101 (1976) (defining a “compilation”); see 17 U.S.C. § 103 (1976) (providing that compilations are copyrightable). Under this logic, the user of the avatar would be deemed the author. 112  Urantia, supra note 99 at 960. 113   Id at 958. 114  Feist, supra note 95; see also Urantia, supra note 99 at 958. 115  Urantia, id at 958. 116   Shawn Bayern, The Implications of Modern Business-Entity Law for the Regulations of Autonomous Systems, 19 Stan. Tech. L. Rev. 93 (2015) available online at: https://journals.law.­stan​ ford.edu/sites/default/files/stanford-technology-law-review/print/2015/10/19-1-4-bayern-final_0.pdf. 109 110

BARFIELD PRINT.indd 25

13/11/2018 11:52

26  Research handbook on the law of virtual and augmented reality status could easily be given to an avatar or artificially intelligent system insofar as it is working for a (human) person. One obstacle to gaining copyright protection is determining whether the avatar is self-aware that it created the work. If the avatar is not self-aware, it can be argued that its output is merely a digital reinterpretation of what it has been programmed to do. Thus, it is not exhibiting any level of creativity required for copyright protection. In this case, if copying were found, the avatar’s owner—and not the avatar—would be liable for copyright infringement.117 Under copyright law, in order for a human author to be liable for copyright infringement, a significant amount of copying would have to be found.118 In the future it may be argued that a virtual avatar that is self-aware (if possible) and producing creative works of authorship is no different than a human author—liable for copyright infringement if a significant amount of copying is found, or on the other hand producing a copyrightable work if the requisite level of creativity exists in the work, and no copying occurred.119 One issue that has impacted the debate over whether the output of virtual avatars should be eligible for copyright protection is the lack of humanlike performance by avatars.120 That is, in many cases, virtual avatars and software agents have performed tasks within a narrow range as defined by a human owner, without showing any creativity beyond the original parameters of the software used to create the avatar. However, recent advances in neural networks have led to works that are different in nature from conventional computer-generated works. The human owner of a neural network can be quite removed from the authorship process and output of the neural network.121 Procedures used by neural nets mimic human brain processes to some extent, and are relevant for the issue of whether the avatar is aware of its own creations. One commentator has argued that the issue of copyrighting neural network weights confronts the issue of intelligent entity authorship head-on.122 On this point, the Copyright Office has already registered a set of neural network weights. Note that avatars may be designed using neural networks and that the structure of the neural network may change as the avatar learns. Thus, one can wonder whether the output of an intelligent avatar operating using a neural network and other artificially intelligent techniques would be eligible for copyright protection, since the weights assigned to the neural networks can be registered. That is, neural network architectures embodied in conventional software are copyrightable, just as are other forms of software. Interconnection weights derived by training a neural network represent a new and valuable form of intellectual property, and courts are typically inclined to protect economic rights. Therefore, copyright law seems to offer one possible means to protect neural network weights and therefore the output of virtual avatars.123 In the area of creative writing, according to one commentator, “computer technology is advancing to the point where a computer may soon be able to generate works in the

  See generally Karnow, supra note 60, at 181–3.   See, e.g., Midler v. Ford Motor Co., 849 F.2d 460, 462 (9th Cir. 1988). 119   See generally Cerqueira, supra note 85. 120   See generally Barfield, supra note 57. 121   See generally Glasser, supra note 83. 122   Donald L. Wenskay, Neural Networks: A Prescription for Effective Protection, 8 Computer Lawyer 12 (1991). 123   Id. 117 118

BARFIELD PRINT.indd 26

13/11/2018 11:52

The law of virtual reality and increasingly smart virtual avatars  27 style of any author that it is programmed to duplicate.”124 In one example, a program was written to write in the style of bestselling author Jacqueline Susann. The result was a published book, Just This Once. To create this work, the programmer used two of Ms Susann’s novels, Valley of the Dolls and Once Is Not Enough, to extract rules which represented the author’s style.125 In order to duplicate Ms Susann’s style, the programmer wrote thousands of computer-coded rules relating to how characters interacted, all based on Ms Susann’s works. The rules, numbering in the thousands, were input into a computer to produce the tone and plot of the book.126 It has been argued that current copyright law is not equipped to deal with the potential legal ramifications of such computer-generated works.127 Copyright law protects the expression of an idea, but not the idea itself.128 And protection extends to works fixed in a tangible medium of expression129—it does not extend to procedures, processes, systems, methods of operation, concepts, principles, or discoveries.130 Once an avatar gains sufficient intelligence to make a claim for copyright protection, it must use more creativity in producing an output than a standard procedure or method. If the writing style of an author is characterized as a system or method of operation, then it may not be protectable.131 To determine if a writer’s style can be protected, it must first be defined. In copyright terms, this is referred to as “dissection.”132 In the above example, the programmer admitted using Susann’s style, reducing her style to thousands of rules equaling thousands of lines of computer code. Most human authors create works by improving on another’s style, and generally such improvements are copyrightable. However, as noted by one commentator, “when a computer is programmed to specifically imitate an author’s style, the human interpretive element is removed.”133 If we assume an avatar with artificial intelligence has developed to the point where it can interpret an author’s style and then create a new work based on that style, and if we also assume that no human was involved in the work being written, still something worthy of copyright protection has been created, but who should receive a copyright for this work? If an artificially intelligent entity cannot be the author of a work, the choice for authorship is either the programmer or user of the software. If the user simply turns on  Vigderson, supra note 89, at 402.   John Boudreau, A Romance Novel with Byte; Author Teams Up with Computer to Write Book in Steamy Style of Jacqueline Susann, L.A. Times at E6 (Aug. 11, 1993). 126  Vigderson, supra note 89 (French identified 200 idiosyncrasies in Susann’s writing. These idiosyncrasies related to language, character, and action. The rules French programmed were designed to teach the 200 idiosyncrasies to the computer). 127   See generally Glasser, supra note 83. 128   17 U.S.C. § 102(a), (b) (1976). 129   17 U.S.C. § 102(a) (1976). 130   17 U.S.C. § 102(b) (1976). 131  Vigderson, supra note 89. 132   See Computer Assocs. Int’l v. Altai, Inc., 982 F.2d 693 (2d Cir. 1992) (Altai enunciated the abstraction test from Nichols v. Universal Pictures Corp., 45 F.2d 119 (2d Cir. 1930), cert. denied 282 U.S. 902 (1931) (upon any work a great number of patterns of increasing generality will fit equally well, as more and more of the incident is left out). See also Brown Bag Software v. Symantec Corp., 960 F.2d 1465, 1475 (9th Cir. 1992) (endorsing “analytic dissection” of computer programs in order to isolate protectable expression), cert. denied, 113 S. Ct. 198 (1992). 133  Vigderson, supra note 89, at 406. Perhaps the human interpretative elements can be found in the software? 124 125

BARFIELD PRINT.indd 27

13/11/2018 11:52

28  Research handbook on the law of virtual and augmented reality the computer and runs the program, no requisite level of creativity for copyright is shown. As for the programmer’s rights, the right to copyright a program is clear, but what about the programmer’s rights to the output of the program? If the program is not run by the programmer but by another person, or by the avatar itself, the programmer would not be the person who “fixed” the work or embodied it in a tangible medium of expression, both requirements for copyright. Therefore, when an intelligent avatar creates a work on its own, determining who the author should be is problematic. Further, in discussing who should be the author of a work generated by an intelligent avatar, the issue of whether the avatar is creating a derivative work in copying the style of a human author should be considered. The Copyright Act defines a derivative work as “a work based upon one or more preexisting works.”134 If Just This Once, a computergenerated work, is viewed as a derivative work,135 then it could be covered under an expansive interpretation of copyright law. However, if an author recognized that his writing style had been copied by a virtual avatar but the words had changed such that no case for copying could be made, then the author would have no cause of action for copyright infringement because under a traditional infringement analysis there would be no substantial similarity. Copyright expert Melville Nimmer defines authorship as “a sine qua non for any claim of copyright . . . the person claiming copyright must either himself be the author, or he must have succeeded to the rights of the author.”136 The Ninth Circuit expressed a narrow interpretation of a derivative work in Litchfield v. Spielberg.137 In Spielberg, the plaintiffs argued that substantial similarity was not a requirement to find that an infringing work was derivative. The Spielberg court soundly rejected this argument, stating that substantial similarity was necessary. It seems reasonable that the “substantial similarity” standard could also be used to analyze the work of virtual avatars.

7.  DESIGN AND USE OF VIRTUAL AVATARS This section presents several intellectual property schemes that can be used to protect the rights of the virtual avatar’s owner and, in the future, with legal rights for virtual avatars, the rights of the intelligent avatar itself. As a basic principle, one needs to consider that a virtual avatar is more than the graphical image that appears in a virtual environment; an avatar also includes the software and algorithms used to design the avatar. We first note that under the Copyright Act, the visual image of the avatar appearing in the virtual environment can receive protection as a pictorial character.138 However, characters may

  17 U.S.C. § 101 (1976).   A “derivative work” is a work based upon one or more preexisting works, such as a translation, musical arrangement, dramatization, fictionalization, motion picture version, sound recording, art reproduction, abridgment, condensation, or any other form in which a work may be recast, transformed, or adapted. A work consisting of editorial revisions, annotations, elaborations, or other modifications which as a whole, represent an original work of authorship, is a “derivative work.” 17 U.S.C. § 101 (1976). 136   Melville B. Nimmer and David Nimmer, Nimmer on Copyright § 5.01 [A], at 5–3 (1993). 137   Litchfield v. Spielberg, 736 F.2d 1352 (9th Cir. 1984), cert. denied 470 U.S. 1052 (1988). 138   17 U.S.C. § 102 (a)(5) (1976). 134 135

BARFIELD PRINT.indd 28

13/11/2018 11:52

The law of virtual reality and increasingly smart virtual avatars  29 also be created with words, in which case they receive protection under the Copyright Act as a literary work.139 Under the copyright statute, the protection of literary characters normally is distinguished from the protection of pictorial characters.140 Due to the unique nature of virtual avatars, existing in the form of software and in the form of an image appearing in a virtual environment, we surmise an avatar may be eligible for dual protection as a pictorial character and as a literary character. The mode of thinking about copyright protection for virtual avatars as literary characters is a less common one. Support for the argument that an avatar could be protected as a literary work is provided by Universal City Studios v. Reimerdes, where the court held that code is eligible to receive First Amendment protection as speech.141 The court held that code is a means of expressing ideas, and thus “the First Amendment must be considered before its dissemination may be prohibited or regulated.”142 If software is used to describe the visual appearance of a virtual avatar, how it reacts in a virtual environment, even its possible range of speech, then the question becomes whether software may be protectable as speech under the First Amendment. Based on the decision in Universal City Studios,143 one could argue that an avatar could receive protection under the Copyright Act as a literary character given that the code used to design the avatar could be protectable as speech. Also supporting this argument are cases involving cartoon, movie, or television characters; the Ninth Circuit has been willing to find copyright protection when the character at issue has a visual representation, as well as a personality described by a word or character line.144 Therefore, the more the virtual avatar displays a unique character, the more likely the court is to find the avatar to be more than an idea, but rather expression deserving of copyright protection.145 That code may be protected under the First Amendment as speech has significance for the rights of virtual avatars should they continue to get smarter. In Universal City Studios, the court concluded that communications do not lose constitutional status as speech simply because they are expressed in the language of computer code.146 This conclusion raises the question of whether the software used to design the avatar itself is protected under the First Amendment. Although the Universal City Studios case did not deal with the issue of whether the First Amendment right applied to virtual avatars, the case does provide insight as to what rights may someday be awarded intelligent virtual avatars, even suggesting that they may even receive constitutional rights.147

  17 U.S.C. § 102 (a)(1) (1976).   See generally Walt Disney Prod. v. Air Pirates, 581 F.2d 751 (9th Cir. 1978). 141   Universal City Studios v. Reimerdes, 111 F.Supp.2d 294, 327 (S.D.N.Y. 2000). 142   Id. See generally Junger v. Daley, 209 F.3d 481, 485 (6th Cir. 2000); Bernstein v. U.S. Dept. of State, 922 F.Supp 1426, 1436 (N.D.Cal. 1996) (First Amendment extends to source code); see Karn v. U.S. Dept. of State, 925 F.Supp. 1, 10 (D.D.C. 1996) (assuming First Amendment protection extends to source code). 143   University City Studios, 111 F.Supp.2d 294. 144   See Gardner v. Nike, Inc., 279 F.3d 774 (9th Cir. 2002.); Wrench LLC v. Taco Bell Corp., 256 F.3d 446 (6th Cir. 2001). 145   See Seals-McClellan v. Dreamworks, Inc., 120 Fed.Appx. 3 (9th Cir. 2004); Murray Hill Publ’ns, Inc. v. Twentieth Century, 361 F.3d 312 (6th Cir. 2004). 146   University City Studios, 111 F.Supp.2d 327. 147   Id. 139 140

BARFIELD PRINT.indd 29

13/11/2018 11:52

30  Research handbook on the law of virtual and augmented reality The increased complexity of visual images has led one commentator to conclude that the situation existing in many courts has resulted in the convergence of distinct bodies of law, such as copyright, trademark, and unfair competition, into a new body of law formulated solely to protect characters.148 According to this commentator, the interplay of many factors has resulted in this convergence of the law.149 These factors include: (1) the profits that can be made from the commercialization of characters, such as virtual avatars who are able to take on a life of their own in settings that differ from those in which the avatar was originally designed to inhabit;150 (2) the ability of avatars to function as entertainment products that are recognized under federal, state, and common law because they “suggest, if not clearly indicate, origin” of the products  or  services  with which the avatar is associated; and (3) the quality that a virtual avatar through extended use, can lead the public to relate to the character as being human.151 In the context of virtual avatars, the copyrightable expression of a character is much more than just the character’s physical appearance. It includes the specific name and character traits of that character as well. In Warner Bros. Inc. v. American Broadcasting Companies, Inc., the court noted that, in determining whether a character in a second work infringed a cartoon character, “courts have generally considered not only the visual resemblance but also the totality of the characters’ attributes and traits.”152 A similar result was shown in Detective Comics, Inc. v. Bruns Publications.153 Here, the court found that the character “Superman” was infringed in a competing comic book publication featuring the character “Wonderman.” The court held that the infringing work “appropriated the pictorial and literary details embodied in” the copyrights protecting Superman.154 To summarize the above courts’ decisions in the context of virtual avatars, a copyright infringement action will involve more than just a showing of the physical similarity between two virtual avatars; the court will also consider the range of behaviors exhibited by the avatars, and even the avatar’s digital speech.155 One of the more difficult problems of applying copyright law analysis and protection to virtual avatars will be to ascertain how such protection will be extended to protect a particular virtual avatar after that avatar has taken on a life of its own, so to speak, and no longer exists in the original context in which it first appeared. Will copyright protection be lost if the virtual avatar’s appearance has changed? For virtual avatars designed using genetic algorithms, once they have mutated their appearance and behavior, will they still be eligible for copyright protection? In order to ascertain whether a virtual avatar might be entitled to copyright protection, the courts likely will follow the

148   Michael Todd Helfand, When Mickey Mouse Is as Strong as Superman: The Convergence of Intellectual Property Laws to Protect Fictional Literary and Pictorial Characters, 44 Stan L. Rev. 623, 641 (1992). 149   Id at 628. 150   Id. 151   See generally Helfand, supra note 148 at 628. 152   Warner Bros. Inc. v. Am. Broadcasting Cos, Inc., 720 F.2d 231, 241 (2d Cir. 1983). 153   Detective Comics, Inc. v. Bruns Publ’ns, 111 F.2d. 432 (2d Cir. 1940). 154   Id. at 433. 155   See generally Midler v. Ford Motor Co., 849 F.2d 460 (9th Cir. 1988).

BARFIELD PRINT.indd 30

13/11/2018 11:52

The law of virtual reality and increasingly smart virtual avatars  31 “character delineation” test used to analyze the copyrightability of graphical images.156 Under this test, the critical issue is whether the avatar is sufficiently and distinctively delineated so that it warrants protection.157 Because copyright law does not protect ideas from infringement but instead only protects the expression of those ideas, courts considering copyright do not protect character types.158 Therefore, while a court would likely not extend copyright protection to a virtual avatar possessing superpowers, the courts will likely extend copyright protection to a specifically delineated “superpowered” virtual avatar, without bestowing a monopoly on the mere character of a “super avatar.” Based on this conclusion, a good way to protect a virtual avatar under copyright law will be to ensure that the avatar’s appearance and personality are specific and unique. Past characters that have received copyright protection have displayed consistent, widely identifiable, traits.159

8. PROTECTION OF VIRTUAL AVATARS UNDER TRADEMARK AND UNFAIR COMPETITION LAW Other avenues for the protection of the rights of the owner of the virtual avatar, and someday the rights of the virtual avatar itself, may include trademark and unfair competition law. Federal, state, and common law protection will protect the avatar from being used by another party without authorization when the avatar functions as a form of identification and is recognized by the public as paired to a product.160 This protection could prevent the duplication of the trademark owner’s avatar, or the imitation of that avatar where the likely result would be to cause public confusion, mistake, or deception with regard to the source of the products or services that carry the likeness of the avatar.161 Trademark law will not permit a graphic character to be trademarked solely for its own protection. However, it does permit the character’s name and likeness to be trademarked when the function of that trademark is to indicate the source of the products and services bearing that mark.162 As may be expected, there will be advantages and disadvantages to protecting a virtual avatar as a trademark. On the positive side, to obtain a trademark, the avatar will not

156   Metro-Goldwyn-Mayer, Inc. v. American Honda Motor Co., Inc., 900 F.Supp. 1287 (C.D.Cal. 1995); Anderson v. Stallone, No. 87-0592 WDK, 1989 U.S. Dist. LEXIS 11109 (C.D.Cal. Apr. 25, 1989). 157   See generally Metro-Goldwyn, 900 F.Supp. at 1296. 158   See 17 U.S.C. § 102(b) (1976). 159   See, e.g., Toho Co., Ltd. v. William Morrow and Co., Inc., 33 F.Supp.2d 1206, 1215 (C.D.Cal. 1998) (Godzilla); Metro-Goldwyn-Mayer, 900 F.Supp. at 1297 (James Bond); Anderson, 1989 U.S. Dist. LEXIS 11109 (Rocky Balboa). 160   See generally Restatement (Third) of Unfair Competition, §§ 13. 16. 17. 161   See generally Kellogg Co. v. Exxon Corp., 209 F.3d 562 (6th Cir. 200) overruled on other grounds by Moseley v. Secret Catalogue, Inc. 537 U.S. 418 (2003) 162   Mark Fowler, Copyright in Fictional Characters: Can I Have Don Draper Make a Cameo Appearance in My Novel? 2011, www.rightsofwriters.com/2011/04/copyright-in-fictionalcharacters-can-i.html (last visited August 29, 2017); see Two Pesos, Inc. v. Taco Cabana, Inc., 505 U.S. 763, 769 (1992).

BARFIELD PRINT.indd 31

13/11/2018 11:52

32  Research handbook on the law of virtual and augmented reality have to include the originality attributes that are required under copyright law.163 In addition, in order to prove trademark infringement, the trademark owner will not need to prove that the infringer had access to the avatar, as is required under copyright law,164 but only that the mark was used by a party other than the owner of the mark without permission. Finally, the longer term of protection—potentially perpetual just as long as the registration requirements are fulfilled and the mark is not abandoned or does not lose its status as a trademark—can be valuable and profitable.165 This is especially true for successful and highly marketable graphic characters, such as many of the Disney and Warner Brothers characters. On the negative side, federal trademark protection for an avatar can be costly.166 This will be especially true if the avatar is extensively used or licensed for use in multiple media formats, or in merchandising programs for many different categories of products and/or services. In addition, because trademark protection is territorial, the avatar serving as a mark may need to be registered in countries other than the United States to provide the maximum degree of protection.167 Because neural nets and genetic algorithms allow an avatar to learn and change their appearance, and any changes in the appearance of the avatar could destroy the original trademark protection, additional trademark registrations (incurring costs) may be necessary to ensure that the current appearance of the avatar remains protected. Another legal theory which may be used to protect an avatar is unfair competition law.168 Unfair competition laws involve a variety of different causes of action that primarily fall into three categories: (1) misrepresentation; (2) sponsorship; and (3) misappropriation.169 Misrepresentation occurs when a party represents that a particular character is associated with their product or service, when in reality it is not.170 Sponsorship occurs when a party indicates that a particular character has endorsed its product or service when it has not.171 Misappropriation, which may be most relevant for the protection of virtual avatars, may occur when a party steals another’s avatar in order to associate it with their product or service.172 Therefore, when one brings an unfair competition action, the injured party is

163   The originality requirement for copyright is expressed in Feist, 499 U.S. at 346; 17 U.S.C. § 101; and U.S. Const. art. I, § 8, cl. 8. 164   See generally U.S. Copyright Office, Copyright basics (Circular 1) available at www.­copy​ right.gov/circs/circ01.pdf. 165   See generally Graeme B. Dinwoodie and Mark D. Janis, Trademarks and Unfair Competition: Law and Policy, 2nd edition (Aspen Casebook Series 2014). 166   Id. 167   If an avatar gains in intelligence, could it then serve as a trademark? The subject matter of trademark covers “any word, name, symbol, or device.” 15 U.S.C. § 1127 (200). Could an avatar that produced its own output serve as either a symbol or device? It seems that an avatar that gained legal rights would not be an appropriate subject for trademark law. 168   See Nike, Inc. v. Kasky, 539 U.S. 654 (2003) 169   See Dinwoodie and Janis, supra note 165. 170   See Mary LaFrance, When You Wish upon Dastar: Creative Provenance and the Lanham Act, 23 Cardozo Arts & Ent. L. J. 197 (2005). 171   See Joseph R. Dreitler, The Tiger Woods Case—Has the Sixth Circuit Abandoned Trademark Law, ETW Corp. v. Jireh Publishing, Inc., 38 Akron L. Rev. 337 (2005). 172   See Peter S. Menell, Regulating “Spyware”: The Limitation of State “Laboratories” and the Case for Federal Preemption of State Unfair Competition Laws, 20 Berkeley Tech. L. J. 1363 (2005). For an example of a real word case dealing with images, see Kellogg, 209 F.3d 562.

BARFIELD PRINT.indd 32

13/11/2018 11:52

The law of virtual reality and increasingly smart virtual avatars  33 claiming that their character has been wrongly associated with another party’s product, service, person, company, or idea.173 If such misuse of a graphic character occurs, and it is determined under the reasonable person standard that the graphic character had been misrepresented,174 used falsely as a sponsor, or misappropriated, then the party engaged in such misuse could be found liable for trademark infringement.175 Most courts have recognized trademark protection for graphic characters and have found trademark infringement liability under both trademark and unfair competition law.176 Therefore, if avatars are used for commercial purposes, in addition to copyright protection, other claims to protect avatars can be brought, including right of publicity and trademark or unfair competition. An example of case law in this area is Walt Disney Productions v. Air Pirates,177 where the court appeared to commingle copyright and trademark infringement criteria by stating that the Disney characters used by the defendants had “achieved a high degree of ‘recognition’ and ‘identification’” and that these elements helped make the characters protectable under copyright law.178

9. THE DOCTRINE OF MORAL RIGHTS FOR VIRTUAL AVATARS The doctrine of moral rights refers to rights regarding the personality of the artist and to the preservation of the integrity of an artist’s intellectual creations.179 The Visual Artists Rights Act of 1990 (VARA)180 provides moral rights protection for artists and protects the personal interests in their work, even after the copyright is transferred to a third party purchaser.181 VARA was the result of efforts on the part of moral rights advocates to overcome Congress’ failure to adopt the moral rights provision of the Berne Convention. The legislation protects works of visual art and gives the artist two kinds of moral rights—the right of attribution and the right of integrity.182 The right of attribution allows the artist to claim authorship of a work and prevent the use of her name as the author of any work which she did not create.183 Presently, no intelligent avatar is awarded attribution rights for its output, but in the future this might be a necessary outcome given the avatar’s

  See Mitsubishi Motors Corp. v. Soler Chrysler-Plymouth, Inc., 473 U.S. 614 (1985).   See generally Joseph Gibbons Llewellyn, Semiotics for the Scandalous and the Immoral and the Disparaging: Section 2(A) Trademark Law after Lawrence v. Texas, 9 Marq. Intell. Prop. L. Rev. 187 (2005). 175   See generally Freeman v. Time, Inc., 68 F.3d 285, 289 (9th Cir, 1995). 176   Fisher v. Star Co., 132 N.E. 133 (1921), cert. denied, 257 U.S. 654 (1921). 177   Walt Disney Prod. v. Air Pirates, 581 F.2d 751 (1978). 178  Helfand, supra note 148, at 643–5. 179   See Ronald B. Standler, Moral Rights of Authors in the USA (1988), www.rbs2.com/moral.pdf. 180   17 U.S.C. §§ 101, 106A, 113, 301, 501(a) (1990). 181   William A. Tanenbaum and Jeffrey M. Butler, The Impact of the Visual Artists Rights Act, 9 N.Y. L. J., 1 (1993) (the moral rights provided in VARA are independent of the usual copyright and are retained by the artist, even if the economic copyrights are sold or assigned). 182   “The right of attribution [is] known as the right of paternity in European practice.” Tanenbaum and Butler, id at 11, col. 1. 183   17 U.S.C. § 106A(a)(1) (1990). 173 174

BARFIELD PRINT.indd 33

13/11/2018 11:52

34  Research handbook on the law of virtual and augmented reality ability to create unique and creative works beyond the original programming. The right of attribution allows the artist the right to prevent the use of her name in connection with a mutilated, distorted, or otherwise modified work, if that alteration would be “prejudicial to . . . her honor or reputation.”184 Likewise, the right of integrity gives an artist the right to prevent intentional mutilations, distortions, and other modifications of a work which would be prejudicial to her honor or reputation.185 The rights granted under VARA may not be transferred, but may be waived by the artist.186 VARA’s passage was a big step toward recognizing moral rights in the United States. However, the enacted version does not protect motion pictures, even though the original version did provide such protection.187 Without the protection that VARA provides other artists, film directors can have altered works attributed to them. One difference, however, between works protected by VARA and motion pictures is that when films are colorized or otherwise altered, the original generally still exists.188 When a “painting or sculpture is altered, the original work is changed forever.”189 Virtual avatars seem to fit better into the film category, since the concept of an “original” is difficult to apply to virtual avatars given that they exist as bits. If courts follow this reasoning, a virtual avatar would not be protected under VARA. The moral rights doctrine is included in the copyright laws of many European countries, as well as the laws of countries subscribing to the Berne Convention.190 Given that virtual avatars reside in a virtual environment which is most likely accessible on the internet, the moral rights doctrine as applied in Europe could be relevant for the protection of avatars created in the United States. Article 6bis of the Berne Convention requires that countries that are members recognize, independently of the author’s economic rights, that “the author shall have the right to claim authorship of the work”—the right of paternity—and “to object to any distortion, mutilation or other modification of, or other derogatory action in relation to the said work, which would be prejudicial to his honor or reputation”—the right of integrity.191 The scope of moral rights protection varies among countries that recognize these rights. However, the doctrine encompasses three major elements: (1) the right of disclosure; (2) the right of paternity; and (3) the right of integrity.192   17 U.S.C. § 106A(a)(2) (1990).   17 U.S.C. § 106A(a)(3) (1990). 186   17 U.S.C. § 106A(e)(1) (1990). 187   Timothy M. Casey, The Visual Artists Rights Act, 14 Hastings Comm. & Ent. L. J. 85, 98 (1991). 188   17 U.S.C. § 101 fully defines “work of visual art” as “(1) a painting, drawing, print, or sculpture, existing in a single copy, in a limited edition of 200 copies or fewer that are signed and consecutively numbered by the author, or, in the case of a sculpture, in multiple cast, carved, or fabricated sculptures of 200 or fewer that are consecutively numbered by the author and bear the signature or other identifying mark of the author; or (2) a still photographic image produced for exhibition purposes only, existing in a single copy that is signed by the author, or in a limited edition of 200 copies or fewer that are signed and consecutively numbered by the author.” 17 U.S.C. § 101 (1990). 189  Casey, supra note 187, at 99. 190   See Berne Convention Implementation Act of 1988, Pub. L. No. 100-568, 102 Stat. 2853 (1988). 191   Berne Convention for the Protection of Literary and Artistic Works, ar 6bis (Sept 9, 1886; revised July 24 1974 and amended 1979; entered into force for the U.S. Mar. 1, 1989 (Sen. Treaty Doc. 99-127)) U.S.T. Lexis 160 or 1 B.D.I.E.L. 715. 192   David J. Kohs, Paint Your Wagon—Please! Colorization, Copyright, and the Search for Moral Rights, 40 Fed. Comm. L. J. 1, 4 (1988). 184 185

BARFIELD PRINT.indd 34

13/11/2018 11:52

The law of virtual reality and increasingly smart virtual avatars  35 Under the right of disclosure, the creator has the privilege of determining when to release his work. The basis of this right is the theory that the creator is the sole judge of when a work is first ready for public dissemination.193 The second element of the doctrine of moral rights under the Berne Convention is the right of paternity, which entitles the author to have his name and authorship recognized.194 This right allows the creator to present himself to the public as the creator of a work.195 Furthermore, the right of paternity permits the author to require others to acknowledge his authorship and prevent others from attributing works to him which he did not originate.196 The third element, the right of integrity, is the right most pertinent to virtual avatars. The right of integrity enables the creator to prevent any distortion of or modification to his work if the alteration would constitute a misrepresentation of his artistic expression.197 This right, like the other moral rights, is held by the creator and is independent of any economic rights that he may or may not have in the work. The United States enacted the Berne Convention Act in 1988. However, the implementing legislation indicated that the law in the United States as it existed on the date of enactment satisfied the United States’ obligations under Article 6bis of the Berne Convention and that no further rights were to be recognized for that purpose.198 Thus, the Implementation Act did not change the pre-Berne Convention precedents or balance of rights between authors and proprietors. As Section 3(b) of the Berne Convention Implementation Act stated, no change in American law regarding the right of paternity or the right of integrity occurred as the result of implementation of this new legislation. Accordingly, the legal theories previously used to protect an author’s moral rights, based upon provisions of the Lanham Act and common law principles, remain the law in the United States. Another legal theory used by the courts to protect the integrity of a work prior to the United States’ ratification of the Berne Convention is embodied in the law of defamation. As stated earlier, an action for defamation protects an individual from harm to his reputation or his standing in the community.199 Given the ability of virtual avatars to take on the look of another person, this tort may still serve people who have been harmed by a “lookalike” avatar, especially if it portrays them in a false light. In Clevenger v. Baker Voorhis & Co., a publisher revised an edition of a well-known attorney’s law book.200 By including the author’s name on the title page, the revision implicitly misrepresented that the author himself, rather than the publisher, had written the revision, which contained many errors.201 Because publishing in the name of a   Id at 12.   Berne Convention, supra note 191. 195   Id. 196   Id. 197  Kohs, supra note 192, at 12. 198   Berne Convention Implementation Act, supra note 191 (amending title 17 of the United States Code to make the changes in the United States copyright law that are necessary for the United States to adhere to the Berne Convention). 199   See Furine Blaise, Game Over: Issues Arising When Copyrighted Work is Licensed to Video Game Manufacturers, 15 Alb. L. J. Sci. & Tech. 517 (2005). 200   Clevenger v. Baker Voorhis & Co., 168 N.E.2d 643 (N.Y. 1960). 201   Id at 644. 193 194

BARFIELD PRINT.indd 35

13/11/2018 11:52

36  Research handbook on the law of virtual and augmented reality ­ ell-known author of a literary work tended to injure his position in the legal comw munity, the court held that the plaintiff had a cause of action against the publisher based upon defamation.202 Similarly, in Ben-Oliel v. Press Publishing Co., the Court of Appeals of New York held that attribution of an inaccurate newspaper article on the social customs of Palestine and Mosaic symbolism to a well-known authority, who did not in fact write it, constituted libel.203 As the foregoing discussion demonstrates, both Section 43(a) of the Lanham Act and the law of defamation are used to preserve the integrity of an author’s work. Both theories may aid a party who alleges that they have been harmed by an avatar. Could such theories also be used by intelligent virtual avatars to protect the integrity of their image and output? Section 43(a) of the Lanham Act provides two general theories of liability:204 (1) false representations regarding the origin, endorsement or association of goods or services through the wrongful use of another’s distinctive mark, name, trade dress, or other device (“false endorsement” or “false association”), and (2) false representations in advertising concerning the quality of services or goods (“false advertising”).205

To prevail under Section 43(a) of the Lanham Act, a plaintiff must show that it has “a valid, protectible trademark and that the defendant’s use of a colorable imitation of the trademark is likely to cause confusion among consumers.”206

  Id at 645–6.   Ben-Oliel v. Press Publ’g Co., 167 N.E. 432 (Ct. App. N.Y. 1929); see also Am. Law Book Co. v. Chamberlayne, 165 F. 313 (2d Cir. 1908) (acknowledging possibility of recovering damages for libel resulting from publication of mutilated or altered form of author’s work). 204   The text of Section 43(a) of the Lanham Act provides that: (1) Any person who, on or in connection with any goods or services, or any container for goods, uses in commerce any word, term, name, symbol, or device, or any combination thereof, or any false designation of origin, false or misleading description of fact, or false or misleading representation of fact, which—(A) is likely to cause confusion, or to cause mistake, or to deceive as to the affiliation, connection, or association of such person with another person, or as to the origin, sponsorship, or approval of his or her goods, services, or commercial activities by another person, or (B) in commercial advertising or promotion, misrepresents the nature, characteristics, qualities, or geographic origin of his or her or another person’s goods services, or commercial activities, shall be liable in a civil action by any person who believes that he or she is or is likely to be damaged by such act. (2) As used in this subsection, the term “any person” includes any State, instrumentality of a State or employee of a State or instrumentality of a State acting in his or her official capacity. Any State, and any such instrumentality, officer, or employee, shall be subject to the provisions of this Act in the same manner and to the same extent as any nongovernmental entity. (3) In a civil action for trade dress infringement under this Act for trade dress not registered on the principal register, the person who asserts trade dress protection has the burden of proving that the matter sought to be protected is not functional. 202 203

  See L.S. Heath & Son, Inc. v. AT&T Infor. Sys., Inc., 9 F.3d 561, 575 (7th Cir. 1993).   Washington Speakers Bureau, Inc. v. Leading Authorities, Inc., 33 F.Supp.2d 488, 493 (E.D. Va. 1999). 205 206

BARFIELD PRINT.indd 36

13/11/2018 11:52

The law of virtual reality and increasingly smart virtual avatars  37

10.  CONCLUSIONS: TOWARD AVATAR PERSONHOOD There are three notable trends in virtual avatars: (1) they are getting smarter; (2) their physical appearance is becoming more photorealistic and humanlike; and (3) their behavior is becoming more sophisticated. With regard to the three points above, imagine one day that a virtual avatar claims that it is a person,207 and that it is therefore entitled to certain constitutional rights. Should the law grant constitutional rights to intelligent avatars that have intellectual capabilities like those of humans? The answer may turn out to vary with the nature of the constitutional right considered and our understanding of the underlying justification for the right. For example, Professors Samuelson and Miller, as well as other legal scholars, have noted that a rationale for copyright is to provide an incentive for authors to create copyrightable works.208 As they argue, since “software and machines” currently need no such incentive to create works, there can be no copyright awarded to such entities. The lack of incentive shown by a virtual avatar (at least current versions) when producing an output under the direction of a human could conceivably be overcome when avatars evolve that self-program and pursue objectives of their own. The economic justification for producing an output when a human author is involved may also provide an incentive for virtual avatars. An intelligent avatar that evolved to the point at which it could claim to be conscious and deserving of legal rights may need to purchase computing resources and memory, or even require the assistance of humans or other artificially intelligent entities for the performance of tasks. In addition, other incentives to motivate virtual avatars to produce outputs useful to society may become clear as virtual avatars evolve and interact further with humans and other avatars. Imagine also that an intelligent avatar claims that it cannot be owned and has been forced into involuntary servitude. A lawyer takes its case and files a civil rights action on its behalf against its owner. How should the legal system deal with such a claim? Would the intelligent avatar have standing to pursue such an action?209 And with regard to intellectual property rights, what if an intelligent avatar creates a work completely independent from a human’s input that meets the requirements for copyright? Would the court then award the avatar a copyright for the work? The current answer is surely no, but why not? The work could clearly pass the copyright hurdles of an original work fixed in a tangible medium of expression. To some, the argument of antagonists of awarding a copyright to an artificial entity comes down to a requirement that a human being be the author of a copyrightable work. For this reason, the issue of personhood for nonhuman entities becomes an important topic when discussing legal rights for intelligent avatars. Before exploring the issue

207   Hans Moravec, Mind Children: The Future of Robot and Human Intelligence 5968 (Harvard Univ. Press 1988) (estimating that it would take roughly ten trillion calculations per second to equal the speed of the human brain and that computers will reach this speed around 2020). 208  Samuelson, supra note 62. 209   See generally Tom Regan, The Case for Animal Rights (University of California Press 1983); Christopher D. Stone, Should Trees Have Standing? Toward Legal Rights for Natural Objects, 45 S. Calif. L. Rev. 450, 458ff (1972); See generally Joseph Mendelson III, Should Animals Have Standing? A Review of Standing Under The Animal Welfare Act, 24 B. C. Envtl. Aff. L. Rev. 795 (1997).

BARFIELD PRINT.indd 37

13/11/2018 11:52

38  Research handbook on the law of virtual and augmented reality of personhood for artificially intelligent entities in greater detail, it should be noted that granting legal recognition to nonhuman entities may not pose an insurmountable problem doctrinally. It is already done for corporations.210 In terms of policy considerations, Professor Samuelson of Berkeley has previously argued that the ownership allocation between humans and software should not only make sense, but also reflect the realities of the world.211 Those realities, in regard to intelligent systems, have changed dramatically since antagonists argued against the idea of copyright protection for artificially intelligent entities in the 1980s and early 1990s. Given the advances in autonomous machines, smart computer vision systems, and self-programming neural nets, Samuelson’s past statement is even more relevant for these times than it was when first made.212 Judge Curtis Karnow introduced the term “electronic person,” or “epers,” when discussing the issue of legal rights for “agents” or “avatars” existing within cyberspace. 213 Taking a liberal view on legal rights for software agents, Karnow argued that epers should be allowed to own physical property, maintain bank accounts, enter into contracts, and be recognized as authors of expression, subject to constitutional protection.214 Karnow,215 as well as Solum,216 have previously addressed the issue of personhood for artificially intelligent entities. According to Solum, “the question whether an entity should be considered a legal person is reducible to other questions about whether or not the entity can and should be made the subject of a set of legal rights and duties.”217 For example, “the particular bundle of rights and duties that accompanies legal personhood varies with the nature of the entity.”218 In this context, both corporations and natural persons are considered legal persons, but they have different sets of legal rights and duties.219 Intuitively, when one uses the term “person” she means to refer to a human being as opposed to a virtual avatar controlled by software.220 However, based on legal principles, the definition of a person is not as straightforward as one might expect. Black’s Law Dictionary defines a person as “[a]n entity (such as a corporation) that is recognized by law as having the rights and duties of a human being.”221 Furthermore, an artificial

  See Santa Clara Cty. V. S. Pac. Ry., 118 U.S. 394 (1886).  Samuelson, supra note 62, at 1192. 212   One could argue that an intelligent avatar and the programmer could share rights to any intellectual property created by the avatar, since the programmer wrote the initial software to create the avatar. However, if the avatar were to become truly autonomous and create works independent from the initial programming, would granting the programmer rights to the avatar’s property then be similar to the idea of granting property rights to one’s parents once the child reached adulthood? 213  Karnow, supra note 60, at 128. 214   Id at 128. 215   See generally Curtis E. A. Karnow, The Encrypted Self: Fleshing Out the Rights of Electronic Personalities, in Future Codes: Essays in Advanced Computer Technology and the Law 117–36 (Curtis E.A. Karnow ed., Artech House Publisher 1997). 216   See Lawrence B. Solum, Legal Personhood for Artificial Intelligences, 70 N. C. L. Rev. 1231 (1992). 217   Id at 1239. 218   Id. 219   Id. See generally Jonathan Chaplin, Political Perspective: Toward A Social Pluralist Theory of Institutional Rights, 3 Ave Maria L. Rev. 147 (2005). See supra note 34. 220   See generally Barfield, supra note 57. 221   Black’s Law Dictionary 1162 (7th ed. 1999). 210 211

BARFIELD PRINT.indd 38

13/11/2018 11:52

The law of virtual reality and increasingly smart virtual avatars  39 person is defined as “[a]n entity, such as a corporation, created by law and given certain legal rights and duties of a human being; real or imaginary, who for purposes of legal reasoning is treated more or less as a human being (also termed a legal person).”222 Based on the latter definition, an intelligent avatar could be regarded as an artificial person and awarded some legal rights. While every human being, regardless of intellectual capabilities, is considered to be a “legal person,” not all persons are considered human beings.223 Indeed, under common law, corporations are regarded as “persons” with full rights to sue, be sued, hold property, and so on. However, as noted by Solum,224 corporations have [human] boards of directors which exert control over the corporation; in contrast, avatars in some domains already perform complex tasks without human supervision. Since corporations have the status of a person for some legal purposes, we can ask whether this legal principle should be considered as precedent for the issue of legal personhood for avatars. There are several reasons why legal personhood is denied to current implementations of avatars. One is the lack of a full repertoire of intellectual abilities similar to those of humans: To be granted legal personhood, it will not be enough for avatars to be idiot savants, experts in a narrow field of knowledge or conduct (such as making theater reservations or playing chess). Instead, avatars will have to exhibit a broad range of intellectual abilities before they begin to approach humanlike cognitive and perceptual capabilities, and thus warrant consideration of their status vis-à-vis legal personhood. Another reason why legal personhood is denied to current versions of avatars is the lack of self-awareness in such systems. Without self-awareness, an avatar is not only denied legal personhood, but also denied the characteristic of being alive. In fact, when the crucial aspects of personhood are irretrievably lost, it is generally assumed that an individual has died, that is, is no longer a person.225 Finally, another reason why avatars are denied legal personhood is based on legal precedent: No such entity has ever approached human levels of intelligence or self-awareness. Thus, the issue of legal personhood for such systems has not been considered.226 The debate on legal personhood for virtual avatars can benefit from a consideration of the legal status of humans and great apes, two species which clearly differ in levels of intelligence, although great apes are certainly intelligent creatures, have complex social structures, and are genetically similar to humans.227 We deny legal personhood to great apes not only because they are not human beings, but also because they have a significantly lower level of intelligence than the “normal” human and it is unclear as to whether they exhibit self-awareness. Although some apes may have the capability to learn language, as evidenced through signing at the level of a three- to four-year-old child, they

  Id at 1162.  Barfield, supra note 57. 224  Solum, supra note 216, at 1239. 225   Steven Goldberg, The Changing Face of Death: Computers, Consciousness and Nancy Cruzan, 43 Stan L. Rev. 659 (1991). 226   See generally Barfield, supra note 57; see generally Martine Rothblatt, Bioethics: Should We Stop a Company From Unplugging an Intelligent Machine? http://www.Kurzweil ai.net/meme/frame. html?m=4. 227   See The Great Ape Legal Project, www.aldf.org/article.asp?cid=20; see Jens David Ohlin, Is the Concept of the Person Necessary for Human Rights? 105 Colum L. Rev. 209 (2005). 222 223

BARFIELD PRINT.indd 39

13/11/2018 11:52

40  Research handbook on the law of virtual and augmented reality are not provided legal personhood. In contrast, people with severe cognitive deficits are provided the legal protection of personhood, regardless of their intellectual capabilities, although the state may assume some responsibility toward their upkeep. So, if humans with cognitive deficits are awarded the status of legal person, why then not consider such rights appropriate for intelligent avatars that may at the least be equally smart? We can also consider the legal status of children in current society as legal precedent for the treatment of intelligent avatars.228 Under the law, children share several attributes of personhood with adults, but their immaturity disables them from receiving all the legal rights of an adult. Until fully possessed of mature reason and adult perspective, children are not legally allowed to assume either the prerogatives or burdens of full legal personhood. However, upon the age of majority, the law fully invests its citizens with constitutional rights, giving them both legal prerogatives and burdens. Before the age of majority, the law seems to manifest a gradual investment in children of legal personhood, roughly corresponding to their gradual attainment of adulthood. Until the age of majority, however, the law views children as lacking in at least some essential attributes of adulthood necessary to their exercise of legal rights and assumption of legal burdens. Arguably, we exclude children from legal standing and personhood for their own protection, providing other remedies for their claims. Indeed, the law assigns children’s claims to parents and the state, assuming one or the other party will best represent children’s interests. Children cannot, the reasoning follows, know or do what is best for them. In the context of intelligent avatars, would it be prudent to treat such entities from a similar legal perspective as minors, affording them some legal rights, but not those of a mature adult? What the above examples seem to suggest is that granting significant rights to virtual avatars based solely on intellectual capability, is ripe with contradictions.229 With the exception of corporations, the essential aspect of an entity that seems to lead to legal rights is self-awareness and humanlike intelligence. For the time being, virtual avatars will be regarded as computer programs consisting of datasets and algorithms, along with a visual representation. As such, they may receive the legal protection that is awarded software, and the protection awarded images from copyright and trademark law.230 However, unlike standard software programs, intelligent avatars may deviate from their original programming until they are no longer recognizable to the original programmer(s). Avatars may run on a single computer or local cluster, or in a distributed fashion across a public network. They may be designed using “classical,” or deterministic, programming algorithms, in which case they should be able to summarize or “explain” their thought process, which could then be evaluated using step by step logic. More likely, however, intelligent avatars will have a substantial “neural network” or deep

228   See generally Ralph C. Brashier, Children and Inheritance in the Nontraditional Family, 93 Utah L. Rev. 983 (1996); Children’s Rights: An Overview, available at http://www.law.cornell.edu/ topics/childrens_rights.htlml; Children’s Rights, available at http://hrworg/children/child-legal.htm (last visited Nov. 6, 2005). 229  Barfield, supra note 57. 230   See Data Cash Sys. Inc. v. JS&A Group, Inc. 480 F.Supp. 1063 (N.D.Ill. 1979) (dealing with the copyrightability of computer programs); see generally Pamela Samuelson, CONTU Revisited: The Case Against Copyright Protection for Computer Programs in Machine Readable Form, 1984 Duke L. J. 663 (1984).

BARFIELD PRINT.indd 40

13/11/2018 11:52

The law of virtual reality and increasingly smart virtual avatars  41 learning component so their internal state may consist of a large number of unlabeled weight values or sophisticated algorithms. In this case, they may output an answer without being able to “explain” it. Intelligent avatars may have a reflective capability that can at least partly describe and summarize the weights used to reach a given conclusion. According to one commentator, one might expect avatars to become strong believers in intellectual property law (copyrights, patents, trade secrets, etc.), to prevent their code and data from being stolen and copied, thus dramatically lowering their potential wages due to competition with clones of themselves.231 Since all machines have owners who pay their rent, power, and network connection charges, under the current law we can always look to the owner, whether a human or a corporation, and hold them responsible for the acts of the avatar, while assuming that the avatar merely acts as their agent. Under this view, the avatars, no matter how smart or decentralized, is just an item of personal property. If the avatar enters into a contract, that agreement binds the owner (subject to the usual rules of contract formation) and not the avatar, and if the avatar commits a tort, its owner is liable to pay compensation for any damages.232 In conclusion, a major event in US corporate law was the landmark Supreme Court decision to treat corporations as “persons” entitled to the equal protection of the laws under the 14th Amendment. Will there also be a similar landmark case for virtual avatars, or, as necessity dictates, will rights for avatars appear slowly, without any particular landmark decision paving the way for their emancipation? Many questions remain unanswered, as there is literally no case law on the rights of artificially intelligent entities in general, and intelligent avatars specifically. However, given the increasing intelligence of avatars, significant legal disputes involving their actions very likely will arise in the future. This chapter has provided a framework in which to consider how future litigation may develop, and potential causes of action which may be raised, especially in the context of virtual avatars.

SOURCES Cases Abdul-Jabbar v. General Motors, 85 F.3d 407, 413–14 (9th Cir. 1996). Amaretto Ranch Breedables, LLC v. Ozimals, Inc., 29 (N.D.Cal. Dec. 21, 2010). Am. Law Book Co. v. Chamberlayne, 165 F. 313 (2nd Cir. 1908). Anderson v. Stallone, No 87-0592 WDK, 1989 U.S. Dist. LEXIS 11109 (C.D.Cal. Apr. 25, 1989). Apple Computer, Inc. v. Franklin Computer Corp., 714 F.2d 1240 (3d Cir. 1983). Ben-Oliel v. Press Publ’g Co., 167 N.E. 432 (Ct. App. N.Y. 1929). Bernstein v. U.S. Dept. of State, 922 F. Supp 1426, 1436 (N.D.Cal. 1996). Bragg v. Linden Research, Inc., 487 F. Supp. 2d 593 (E.D. Penn. 2007). Brown Bag Software v. Symantec Corp., 960 F.2d 1465, 1475 (9th Cir. 1992). Clevenger v. Baker Voorhis & Co., 168 N.E.2d. 643 (N.Y. 1960).

231   See Karnow, supra note 215 at 128 (including a discussion of the rights of electronic persons or “epers”). 232   See generally Karnow, supra note 60 (discussing the difficulty of finding a responsible party given a distributed computing system).

BARFIELD PRINT.indd 41

13/11/2018 11:52

42  Research handbook on the law of virtual and augmented reality Computer Assocs. Int’l v. Altai, Inc., 982 F.2d 693 (2d Cir. 1992). Daly v. Viacom, 238 F. Supp. 2d. 1118, 1123 (N.D.Cal. 2002). Detective Comics, Inc. v. Bruns Publ’ns, 111 F.2d. 432 (2d Cir. 1940). Eros LLC v. Doe, United States District Court, Middle District of Florida (2007). Evans v. Linden Research, Inc., 763 F. Supp. 2d 735 (E.D. Pa. 2011). Elvig v. Nintendo of America, Inc., 696 F.Supp. 2d 1207 (2010). Feist Publ’ns, Inc. v. Rural Tel. Serv. Co., Inc., 499 U.S. 340 (1991). Fisher v. Star Co., 132 N.E. 133 (1921), cert. denied, 257 U.S. 654 (1921). Freeman v. Time, Inc., 68 F.3d 285, 289 (9th Cir. 1995). Gardner v. Nike, Inc., 279 F.3d 774 (9th Cir. 2002). Gionfriddo v. Major League Baseball, 94 Cal. App. 4th 400, 415–17 (Cal. Ct. App. 2001). Harris v. State, 152 S.W.3d 786 (Tex. App. 2004). Junger v. Daley, 209 F.3d 481, 485 (6th Cir. 2000). Karn v. U.S. Dept. of State, 925 F.Supp. 1, 10 (D.D.C. 1996). Kellogg Co. v. Exxon Corp., 209 F.3d 562 (6th Cir. 2000). Laws v. Sony Music, 294 F.Supp.2d 1160 (C.D.Cal. 2003). Litchfield v. Spielberg, 736 F.2d 1352 (9th Cir. 1984). L.S. Heath & Son, Inc. v. AT&T Infor. Sys., Inc., 9 F.3d 561, 575 (7th Cir. 1993). MAI Systems Corp. v. Peak Computer, Inc., 991 F.2d 511 (9th Cir. 1993). Metro-Goldwyn-Mayer, Inc. v. American Honda Motor Co., Inc., 900 F.Supp. 1287 (C.D.Cal. 1995). Microsoft v. Harmony Computers & Electronics, 846 F. Supp. 208 (E.D.N.Y. 1994). Midler v. Ford Motor Co., 849 F.2d 460 (9th Cir. 1988). Moseley v. V Secret Catalogue, Inc. 537 U.S. 418 (2003). Murray Hill Publ’ns, Inc. v. Twentieth Century, 361 F.3d 312 (6th Cir. 2004). Newcombe v. Adolf Coors Co., 157 F.3d 686, 692º3 (9th Cir. 1998). Nichols v. Universal Pictures Corp., 45 F.2d 119 (2d Cir. 1930). Nike, Inc. v. Kasky, 539 U.S. 654 (2003). Oliver v. Saint Germain Found., 41 F.Supp. 296 (S.D.Cal. 1941). Santa Clara Cty. v. S. Pac. Ry., 118 U.S. 394 (1886). Seals-McClellan v. Dreamworks, Inc., 120 Fed.Appx. 3 (9th Cir. 2004). Step-Saver Data Systems, Inc. v. Wyse Technology, 939 F.2d 91 (1991 U.S. App. 16526). Toho Co., Ltd. v. William Morrow and Co., Inc., 33 F.Supp.2d 1206, 1215 (C.D.Cal. 1998). Two Pesos, Inc. v. Taco Cabana, Inc., 505 U.S. 763, 769 (1992). Universal City Studios v. Reimerdes, 111 F.Supp.2d 294, 327 (S.D.N.Y. 2000). Urantia Found. v. Maaherra, 114 F.3d 955 (9th Cir. 1997). Waits v. Frito-Lay, 978 F.2d 1093, 1098–100 (9th Cir. 1992). Walt Disney Prod. v. Air Pirates, 581 F.2d 751 (9th Cir. 1978). Warner Bros. Inc. v. Am. Broadcasting Cos, Inc., 720 F.2d 231, 241 (2d Cir. 1983). Washington Speakers Bureau, Inc. v. Leading Authorities, Inc., 33 F.Supp.2d 488, 493 (E.D. Va. 1999). Wendt v. Host Intern., Inc., 125 F.3d 806, 810 (9th Cir. 1997). White v. Samsung, 971 F.2d 1395, 1397–99 (9th Cir. 1992). Wrenh LLC v. Taco Bell Corp., 256 F.3d 446 (6th Cir. 2001). Yow v. National Enquirer, Inc. 550 F.Supp.2d 1179, 1183 (E.D.Cal. 2008). 

Statutes Berne Convention Implementation Act of 1988, Pub. L. No. 100-568, 102 Stat. 2853 (1988). Berne Convention for the Protection of Literary and Artistic Works, art. 6bis (Sept. 9, 1886). Cal. Civ. Code §§ 44, 45a, and 46. Defamation Act of 1996, Great Britain. Federal Trademark Dilution Act of 1995, Pub. L. No 104-98, 109 Stat. 985 (Jan. 16, 1996). Lanham Act § 43(a). Restatement (Third) of Unfair Competition, § § 13, 16, 17. Sen. Treaty Doc. 99-127. Trademark Act of 1946, codified at 15 U.S.C. § 1125 et seq. US Const. art. I, § 8, cl. 8. U.S.T. Lexis 160. 1 B.D.I.E.L. 715. 113 S. Ct. 198 (1992). 282 U.S. 902 (1931).

BARFIELD PRINT.indd 42

13/11/2018 11:52

The law of virtual reality and increasingly smart virtual avatars  43 470 U.S. 1052 (1988). 17 U.S.C. § 101 (1976). 17 U.S.C. § 102(a) (2005). 17 U.S.C. § 106A, 113, 301, 501(a) (1990). 17 U.S.C. § 117 (1996). 47 U.S.C. § 230(c) (2000).

BARFIELD PRINT.indd 43

13/11/2018 11:52

2.  Starting up in virtual reality: examining virtual reality as a space for innovation Crystal Nwaneri*

1. INTRODUCTION: VIRTUAL REALITY AS AN INNOVATION SPACE The Oxford Dictionary defines virtual reality as a “computer-generated simulation of a three-dimensional image or environment that can be interacted with in a seemingly real or physical way by a person using special electronic equipment, such as a helmet with a screen inside or gloves fitted with sensors.”1 In some respects, commercial virtual reality (“VR”) has been around since at least the late 1950s but has recently seen a precipitous rise as a result of modern desktop and mobile devices. For a time relegated to laboratories at academic institutions, VR is once again a growing, budding commercial industry with applications in education, health care, military training, and theme parks. This new beginning for the industry is based primarily on advances in “the helmet with a screen inside,” that is, in head-mounted displays (“HMD”). Such displays contain sensors that detect and track the orientation and tilt of a user’s head, which allows images on the helmet’s screen to move corresponding to those head movements. The helmets consist of a mask with a screen large enough to encompass more than 90 degrees of a user’s field of view. The “virtual reality” industry had a lift in 2014 when Facebook spent $2 billion dollars to purchase Palmer Luckey’s Oculus Rift VR system. The Rift’s key distinctions were its lower price point compared to other products on the market, high resolution, wide 100degree field of view, and lower latency rate. Latency refers to the rate at which a visual display changes or updates its screen in order to “trick” the human eye into having the impression that it is viewing a continuous simulation. If the update rate is fast enough to increase a user’s sense of presence in the immersive virtual environment, it may also reduce the incidence of motion sickness that may occur if the visual display is unable to align with the speed at which the brain processes images. The Oculus system is able to capture and track movements 1,000 times per second using its “Adjacent Reality Tracker,” composed of a magnetometer, gyroscope, and accelerometer. The Oculus Rift system is now part of a growing hardware field consisting of mobile and desktop devices that allow more people than ever to use and own their own VR devices. Along with improved hardware and corresponding increases in the use of VR technology have come interesting legal issues in areas ranging from commercial law to intellectual property.

*  This chapter is based in part on past work by the author, such as Crystal Nwaneri, Ready Lawyer One: Legal Issues in the Innovation of Virtual Reality, 30(2) Harvard J. L. & Tech. 601 (2017). 1   Oxford Living Dictionaries, definition of virtual reality, https://en.oxforddictionaries.com/ definition/virtual_reality.

44

BARFIELD PRINT.indd 44

13/11/2018 11:52

Starting up in virtual reality  45 Just as funders have invested in the engineering of modern HMDs and position tracking systems, they have also invested in the creation of content within this growing industry. Generally, VR devices allow users to choose content through application (“app”) stores that work similarly to the Google Play Store and Apple App Store. Oculus has a proprietary app store; its competitor HTC’s desktop VR hardware works with content distributed through an existing game distribution platform called Steam. Content developers range from indie gaming studios,2 on the one hand, to dedicated media studios making VR experiences, on the other.3 One of the most prominent entities to invest heavily in VR content development is the New York Times, which offers 360-degree news experiences.4 The continued development of VR hardware and content creates multiple points of liability. This chapter will review the legal implications of VR as a field for innovation. After going over the history of VR, the chapter will examine the common legal pitfalls that VR companies may face in retaining confidential information and obtaining funding as hardware or software startups. It will then consider new privacy risks that come from the collection of biometric data necessary to make VR function. It will next explore the potential cognitive and physical implications of allowing users to immerse themselves in alternative universes while still physically inhabiting this one. It will end by discussing how to protect the intellectual property of these new companies and the legal rights of others who may have their intellectual property reproduced without authorization in VR. In innovating in virtual reality, developers and the attorneys advising them must think critically regarding both the new and old legal risks posed by this nascent industry.

2.  MODERN VIRTUAL REALITY 2.1  History of VR VR has a long history, which is worth examining in more detail before going on to discuss the legal implications of the technology’s current iterations. In 1957, Morton Heilig developed what some argue is the first instance of VR—a 3D head-mounted display and a Sensorama machine that allowed users to immerse themselves in an experience with artificial vision, sound, smell, and vibration. A decade later (extending his work from the early 1960s), Ivan Sutherland created a stereoscopic HMD, commonly called the “Sword of Damocles,” that was suspended from the ceiling above a user’s head during operation.5 Experimentation with VR had been around for at least 40 years before Jaron Lanier coined the phrase “virtual reality” in the 1990s. Lanier himself was a VR innovator. His company, VPL Research, created new VR peripherals, including HMDs and data gloves used to track hand motions to enable 2   See Feltham, 11 VR Games Developers to Watch, VentureBeat (Oct. 1, 2016), https:// venturebeat.com/2016/10/01/11-vr-game-developers-to-watch/. 3   See, e.g., StartVR, https://startvr.co/. 4   NYTVR, www.nytimes.com/marketing/nytvr/ (last visited Aug. 11, 2017). 5   Ivan Sutherland, A.M. Turing Award, http://amturing.acm.org/award_winners/­sutherla​ nd_3467412.cfm

BARFIELD PRINT.indd 45

13/11/2018 11:52

46  Research handbook on the law of virtual and augmented reality interaction with the immersive environment. One of the goals of VR technology is to increase one’s sense of presence in a computer-generated simulation. Early investigators of this concept were Tom Sheridan at MIT and Woodrow Barfield, then at the University of Washington. Presence is the sense of actually being or existing in a virtual world.6 Also of importance to the VR experience were the development of the peripherals just mentioned, such as HMDs and data gloves, which track a user’s position in order to effect change in the digital display of the 3D environment that replicates the change in the physical world. These changes in display and position within a 3D environment increase presence and are what make VR immersive in the first place. The key tool in immersive VR is the HMD, which surrounds the user with a computergenerated virtual environment. An HMD “tricks” the human eye and thus the brain into making a user believe she is present in the virtual world by enclosing the user in a blackout blindfold to reduce external distractions while presenting each eye with a slightly separate or offset view of the virtual scene in order to mimic human binocular vision. The human eye will instinctually fuse the distinct images in order to create 3D depth perception. Stereo or surround speakers may be added to increase the sense of immersion. The HMD tracks head position (or in some cases user movement) by using an accelerometer—such devices, whether in one’s cellphone or in a VR display, allow for navigation in reference to the device in which it is contained. When an accelerometer is placed in an HMD, navigation is based on the position of the user’s head. In this latest era of VR technology, two general platforms have emerged: mobile and desktop VR, distinguished by the computers used to effectuate the VR experience. 2.2  Desktop VR Consumer desktop VR is experienced through the use of an HMD attached to a high-powered computer or video game console. Some early iterations of this type of VR technology are the HTC Vive, Oculus Rift, and Sony PlayStation VR. These VR technologies require either sensors or cameras set up around the room to input and track the user’s position. The latter of these systems hooks up to the Sony PS4 video game console while the former systems connect to computers. This is the traditional version of commercial VR technology. The presumed advantage of these types of systems is their greater functionality, larger displays, higher refresh rates, and overall greater computing power. This allows content developers to create high-fidelity games.7 Yet, this requires a powerful VR-ready computer and space in order to set up a system that properly tracks one’s head, hand, and body positions. This creates a significant barrier of entry for interested consumers—not only are these VR kits expensive in their own right, but in order to use them one must already own a VR-ready PC or console, or would need to invest in one separately.8 6   See, e.g., W. Barfield and S. Weghorst, The Sense of Presence within Virtual Environments: A Conceptual Framework, in Human-Computer Interaction: Applications and Case Studies (ed. G. Salvendy and M. Smith), Elsevier, 699–704, 1993. 7   Matthew Handrahan, The Vanishing Difference between Mobile and Desktop VR, GameIndustry. biz (Jul. 25, 2016), www.gamesindustry.biz/articles/2016-07-25-the-vanishing-differences-betweenmobile-and​-desktop-vr. 8   Id.

BARFIELD PRINT.indd 46

13/11/2018 11:52

Starting up in virtual reality  47 2.3  Mobile VR A user experiences mobile VR through the connection of smartphones to mobile HMDs. Some popular early mobile HMDs are the Samsung GearVR and Google’s Daydream. The former HMD runs with versions of the manufacturer’s smartphone released concurrently and the latter with a select group of smartphones running Alphabet’s Android operating system. The phrase “mobile VR” is in some ways a misnomer. Although the smartphones and HMDs themselves are wireless and can be carried easily to different locations—unlike a desktop VR system—the user must nevertheless remain stationary while using such a device for reasons of safety. The devices are mobile, but while they are in operation, the user is not. Although the computing power, and thus the display power, of a smartphone and that of a fully rigged gaming PC are nowhere near each other in terms of capabilities, many industry insiders feel that mobile VR is the future of the industry. Despite the few inputs that mobile VR devices support compared to more elaborate setups, developers still say that it is easier and cheaper to develop content for mobile VR platforms.9 By the end of 2016 more than 100 million VR systems had sold, but 88 million of those sales were of Google’s low-end HMD, Cardboard. Mobile VR has a larger audience and has made VR as a whole a more acceptable cultural product.10 The fact that mobile VR is cheaper has resulted in it becoming far more commonplace than higher-end VR systems, and allows users to be in VR anywhere they have their HMD and smartphone. This poses significantly greater personal risks for those who choose not to exercise care.

3.  VR STARTUP CONCERNS 3.1  New Legal Issues in Funding a VR Startup Not only is the technology for modern VR different compared to past iterations, but it is also distinct from VR’s past era in its financial origins. Because modern VR is commercial, smaller companies attempting to obtain funding to develop their experiences or products are looking to private funding rather than research grants to innovate in the field. While the information below is widely applicable to the high-tech startup space rather than VR specifically, it is worth analysis given that it describes the context in which a modern VR startup will arise. Back in 2012, Oculus proved that VR was a viable investment through a highly successful crowdfunding campaign and its later acquisition by Facebook. Founder Palmer Luckey developed an early version of the Rift in 2012 and that year it was named one of the best products at the Electronic Entertainment Expo, an annual conference showcasing upcoming developments in the gaming world. Based on the interest generated at the Expo,

  Id.   Id; Paul Armstrong, Just How Big Is the Virtual Reality Market And Where Is It Going Next? Forbes (Apr. 6, 2017), www.forbes.com/sites/paularmstrongtech/2017/04/06/just-how-big-is-thevirtual-reality-market-and-where-is-it-goi​ng-next/#1334d404834e.  9 10

BARFIELD PRINT.indd 47

13/11/2018 11:52

48  Research handbook on the law of virtual and augmented reality Oculus started a Kickstarter campaign, which garnered over two million dollars—among the largest amounts raised in crowdfunding history at the time, and far above the amount the company was seeking. Oculus proved to venture capitalists and large electronics manufacturers alike that VR had a commercially viable future. Oculus and other successful high-tech hardware crowdfunding campaigns also revealed that crowdfunding’s “backers” were viable sources of capital for innovative early stage startups. Backers in a crowdfunding campaign are a distinct source of financial capital. Backers provide a relatively small amount of capital to a project listed on a crowdfunding platform’s website in exchange for whatever promise a founder makes based on the price tier of the donation. This promise can range anywhere from heartfelt thanks to early access to a new product or, if the donation is large enough, early access to a prototype of the hardware at issue. Yet even these so-called gifts may come with obligations. First, public pressure may cause a VR or other hardware company to owe a nonlegal obligation to backers. The clearest example is Oculus. After the company’s successful Kickstarter campaign, which allowed it to begin to develop a state of the art HMD, Facebook bought Oculus for $2 billion dollars. Although the sale took place two years after the Kickstarter campaign, it caused a backlash among some of the company’s early supporters. Oculus had already fulfilled its Kickstarter promises to backers, yet in 2016 it sent the first commercial release of its Rift HMD as part of a freebie bundle to backers who had provided the original campaign with significant funds. While Oculus had no legal obligation to provide an additional freebie, it likely made the decision to stay in its supporters’ good graces despite the fact that, in financial terms, the company had become wildly successful. It is unclear what rights, if any, a crowdfunding backer could assert against a company in which she has no stake if that company moves in a direction the supporter does not like. If a company violates a promise made to backers on Kickstarter, while there is no formal legal relationship between the company and the backer, such an event may still constitute fraud and lead to an enforcement action taken by the Federal Trade Commission (“FTC”). Section 5 of the FTC Act allows the FTC to take enforcement action against companies that engage in “unfair and deceptive acts and practices in or affecting commerce.”11 In FTC v. Erik Chevalier, after accepting over $122,000 in Kickstarter campaign funds in order to make a board game, the defendant canceled the project after 14 months and used the funds to pay for personal expenses.12 The FTC charged the defendant with engaging in deceptive acts in violation of the statute.13 In the end, the court ordered that the defendant was restrained and enjoined from misrepresenting the purposes for which funds raised from a crowdfunding campaign may be used, from failing to honor a crowdfunding service’s refund policy, and awarded the FTC a little over a suspended $111,000 due to the defendant’s financial state.14 This was the first case to be brought based on fraud occurring in a crowdfunding campaign—it shows that consumers who have participated in such campaigns have protection even if they may not have any legal claim against a company.   15 U.S.C. § 45(a).   FTC v. Erik Chevalier also d/b/a The Forking Path, Co., Case 3:15-cv-01029-AC (D. Or. 2015). 13   Id. 14   Id. 11 12

BARFIELD PRINT.indd 48

13/11/2018 11:52

Starting up in virtual reality  49 Based on the Jumpstart Our Business Startups Act (JOBS Act) of 2012,15 some of these supporters may have an actual stake in a company and have actual rights to assert. Formal investors in a company provide capital to a company in order to receive equity. The JOBS Act permits individuals who are only able to make a small investment, subject to certain conditions in the statute, to obtain a financial stake in a company they support offering securities. Specifically, the statute makes securities-based crowdfunding legal. This allows a startup to obtain crowdfunding in exchange for actual securities. In order to obtain funding, the company must use an intermediate funding portal rather than soliciting funding directly from investors and investors are limited in the amount of funds they may invest based on their income.16 While this Act has allowed companies to diversify their funding, taking too much capital from smaller investors may actually have a negative impact on the future growth of the company, deterring angel investors and venture capitalists from future rounds of financing. Some view this alternative form of financing as a positive for the innovative hardware industry, allowing developers to generally retain control of their company while still getting outside investments. However, as shown above, crowdfunding in both forms poses its own set of problems to hardware startups. Lawyers advising such companies should be aware of all promises made to backers and, in the case of JOBS Act investors, must meet all of the Act’s requirements for crowdfunding investments. There are many new ways in which nascent companies can obtain funding—the next big innovation in VR may not use a traditional funding method and there are a number of legal considerations for both the entities and their lawyers to consider if they take an alternative path. 3.2  Protecting Trade Secrets During Development The first challenge of developing innovative hardware in the tech space is retaining talented developers and protecting company secrets. New VR companies launching soon will have to protect trade secrets in the development process of their hardware and software. While this is a general rule of thumb for startups, some highly publicized legal battles in the retail VR field reinforce this need. In one legal battle, Total Recall Technologies (“Total Recall”) sued Facebook and its subsidiary Oculus Rift, LLC (“Oculus”) for a breach of confidentiality with regard to the technology behind Total Recall’s HMD hardware.17 Palmer Luckey, the founder of Oculus Rift, had signed a nondisclosure, exclusivity, and payments agreement with Thomas Seidl stating he would have exclusive use of Luckey’s HMD designs for $10,000 per year. Thomas Seidl formed Total Recall in partnership with Ron Igra in order to develop a 3D head-mounted display. The agreement granted Seidl rights to Luckey’s designs and also held that all commercially valuable information was protected and confidential. After Luckey had sent a prototype to Seidl for Total Recall, he announced the commercial sale of Rift, outside the agreement. Total Recall

  H.R. 3606—112th Congress: Jumpstart Our Business Startups Act; 112 P.L. 106.   Updated: Crowdfunding and the JOBS Act: What Investors Should Know, FINRA, www. finra.org/investors/alerts/crowdfunding-and-jobs-act-what-investors-should-know. 17   See Total Recall Techs. v. Luckey, No. C 15-02281 WHA (N.D. Cal. 2016). 15 16

BARFIELD PRINT.indd 49

13/11/2018 11:52

50  Research handbook on the law of virtual and augmented reality itself was not a party to the agreement but the court permitted it to sue based on its standing as a third party beneficiary and as an undisclosed principal to the agreement between Luckey and Seidl. Luckey asserted that he could not have been in breach since the contract had not taken effect, and that the contract violated California law prohibiting noncompete agreements. The defendant’s position was undermined further when it was discovered that Luckey had allegedly agreed in conversation with Seidl to share the prototype at issue through open source methods. Total Recall claimed that the defendant committed fraud when he decided to commercialize the HMD. It is unclear whether Luckey and Oculus would have been able to succeed based on their assertions—the case was dismissed when the court discovered that the lawsuit was unauthorized given that both partners had not agreed to it and the plaintiff partner failed to cure this issue by obtaining authorization or ratification.18 But this all shows the importance of the complex agreements that can be formed when one is working in a nascent field, and the necessity of writing clear contract agreements. None of the parties realized how popular the Oculus system would ultimately be, but things said in casual conversation were later brought up in litigation regarding who truly owned the information used to reinvigorate a multibillion-dollar industry. Developers must seek out advice when working with promising technology and lawyers must advise clients on the promises they should and should not make, and help them determine the proper scope of even preliminary working agreements. The issue of ambiguity in contracts concerning development and production has arisen in two other cases in the VR field. In the first, ZeniMax Media Inc. and id Software LLC (collectively “ZeniMax”) sued defendants including Luckey, Oculus, and Facebook for, inter alia, copyright infringement, tortious interference with a prior contract, misappropriation of trade secrets, unfair competition, and breach of contract.19 The defendants argued that the plaintiff’s claims rested on an unenforceable nondisclosure agreement that did not define the scope of the confidential information it was meant to protect. The case centered on an ambiguous agreement. This argument over what information was protected by the agreement between the parties cost the defendants a significant amount of money. At the end of a trial a jury found in favor of ZeniMax over Oculus and its executives and awarded the plaintiffs $500 million dollars for breach of contract, copyright infringement, and false designation. In a second case, Fawzy Amer Deghedy v. Viztek, Inc., the Southern District of Iowa granted a permanent injunction against the plaintiffs in a contractual dispute regarding the “contractual rights and obligations arising from the development, distribution and sales of a three-dimensional anatomical visualization system of the human body,” or a VR-like program used to teach anatomy to medical students.20 Although the jury found the defendants guilty of unjust enrichment and breach of implied contract, the jury also found the plaintiff guilty of “defamation, intentional interference with   Total Recall Techs. v. Luckey, No. C 15-02281 WHA, at *11 (N.D. Cal. Mar. 9, 2017)   Zenimax Media, Inc. v. Oculus VR, LLC, 166 F. Supp. 3d 697 (N.D. Tex. 2015). 20   Fawzy Amer Deghedy v. Viztek, Inc., No. 3:12-cv-00048-HCA, 2016 U.S. Dist. LEXIS 181097 *2 (S.D. Iowa June 10, 2016); see Cyber-Anatomy Med VR, Cyber-Anatomy, http://­cyberanatomy.com/product_CAHA_medVR.php (last visited Aug. 12, 2017); Global Medical VR, Global Medical VR, http://globalmedicalvr.com/ (last visited Aug. 12, 2017). 18 19

BARFIELD PRINT.indd 50

13/11/2018 11:52

Starting up in virtual reality  51 prospective business, misappropriation of trade secrets, copyright infringement and unfair competition.”21 Based on this ruling, the court enjoined the plaintiff from further infringement of the defendants’ copyrights in their Cyber-Anatomy Med VR, misappropriation of the defendant’s trade secrets and confidential information, and from generally using the defendants’ trade secrets in any commercial manner.22 The first defense against the disclosure and misuse of a new company’s confidential, proprietary information is the confidential agreements that such companies have employees and external partners sign. These agreements must be clear and must properly define what information is protected and who is entitled to use it and in what capacity. Considering the financial stakes, it must be clear if technology will be open source or if it will be considered a trade secret. Even then, the provision of labels stating information’s confidentiality may not establish that such information constitutes a trade secret—what constitutes a trade secret is determined by looking at the information’s value, whether or not the information can be obtained some other way, and the extent to which a party imposes measures to protect trade secrets.23 The virtual reality industry is growing and the information within the industry is valuable—lawyers must be cognizant of the value of this information even in working with early stage startups and take necessary measures to truly protect their client’s work from disclosure.

4.  RISKS OF VR 4.1 Privacy Lawyers and developers should consider the legal risks of VR content and hardware companies not only as startups but also as platforms that bring new risks to data privacy. It is a VR company’s responsibility to inform consumers about the range of data such technology will collect, as well as to provide them with information about how such data will be used. This applies to top-level hardware manufacturers and content developers alike. Just as applications on smartphones each have their own set of permissions regarding the information they can access on phones and collect about users, each VR experience may collect different sets of information from users, and users have a right to be informed of what that may be. Virtual reality functions through the collection of data regarding a user’s physical movements. Peripherals, whether an HMD or a data glove, are likely to collect information about a user’s position in space and therefore information about a user’s physical movements. As peripherals become more advanced, they will likely be able to collect more particularized biometric information such as a person’s heart rate, hand movements, or even gaze. Yet, of the major existing hardware manufacturers, only Oculus directly addresses in its privacy policy the fact that it may collect information on a user’s   Fawzy Amer Deghedy v. Viztek, Inc.   Id. 23   Jonathan Rubens, Early-Stage IP Protection: A Primer and Overview for Working with the Startup, Bus. L. Today, July 2016, at 1, 1, www.americanbar.org/content/dam/aba/publications/ blt/2016/07/ip-protection-201607.authcheckdam.pdf. 21 22

BARFIELD PRINT.indd 51

13/11/2018 11:52

52  Research handbook on the law of virtual and augmented reality physical movement. Like other manufacturers, such as Sony, Samsung, Google, and HTC, Oculus’s privacy policy says it will collect information provided by users on how they access the company’s services, about the applications installed on users’ devices, and about how such information may be shared with third parties. Oculus is unique in that it is the only major dedicated VR hardware company with its own dedicated privacy policy. Other companies generally direct users of their VR devices to the company’s general privacy policy for all its services, and this may be inadequate given the discrete information that an HMD and other VR peripherals may collect. Oculus’s privacy policy in particular states it will collect “[i]nformation about [a user’s] physical movements and dimensions when [she uses] a virtual reality headset.”24 Users may access the data associated with their Oculus accounts by emailing the company.25 Facebook, Oculus’s parent company, has previously taken advantage of its legal rights in using user data. Due to the social network’s broad data use policy, its company data scientists were able to run an experiment on users seeing how removing all negative or positive posts affected a user’s mood. This led to the publication of a scientific paper on the subject in 2014.26 Users should note that Oculus similarly retains the right to use the information it collects to conduct research on the way consumers use its services. Although Oculus may be the only company that explicitly states it will collect such information, if collection of movement and biometric data is in the nature of VR hardware and software, companies trying to enter the space may all be collecting unique information that has never been collected in such a widespread manner before. If this is coupled with passive or personally identifiable data given to these companies, such entities may have the ability to determine a user’s physical characteristics or health data by connecting formerly unconnected data points. While users should be conscious of the amount of information collected about them from these companies, it is unlikely that the companies will intentionally misuse that data. But users would have a right to be wary of the possibility that such information might fall into the wrong hands after a cyberattack or be misused by a third party to facilitate discrimination or engage in some type of surveillance. Although major hardware players in the VR industry currently do not create distinct privacy policies for their new devices, because this is an entirely new category of information being granted to commercial entities, it is likely that it will prove worthwhile to craft a new privacy policy that clearly tells a user how such data will be collected and retained. Such a policy is not only important for users but would also be useful internally so that employees within a company can design with notice of the company’s goals regarding user privacy. If a company does not adhere to its privacy policy—by collecting more data than disclosed or by violating its privacy policy—it may cause a backlash among users or run afoul of state privacy and wiretapping laws, or of the FTC itself. Section 5 of the FTC Act allows the FTC to take enforcement action against companies that engage in “unfair and deceptive acts and practices in or affecting commerce” or for   Oculus, Legal Documents, Oculus Privacy Policy, www.oculus.com/legal/privacy-policy/.   Id. 26   Adam D.I. Kramer, Jamie E. Guillory, and Jeffrey T. Hancock, Experimental Evidence of Massive-Scale Emotional Contagion through Social Networks, 111.24 Proceedings of the National Academy of Sciences 8788 (2014). 24 25

BARFIELD PRINT.indd 52

13/11/2018 11:52

Starting up in virtual reality  53 violating other federal laws relating to consumer privacy and security.27 This includes taking action against a company for failing to live up to its privacy promise or failing to maintain security for sensitive consumer information. The FTC has also held companies secondarily liable for the advertising and telemarketing practices of business partners and this may be extended to a contracting third party’s failure to uphold their privacy policies. This means that VR content developers should make sure they fulfill their privacy promises and that contracting hardware or platform providers ensure applications have a privacy policy in place. 4.2  Cognitive and Physical Risks In its product safety and regulatory guide for the Vive, HTC warns: Content viewed using the product can be intense, immersive, and appear very life-like and may cause your brain and body to react accordingly. Certain types of content (e.g. violent, scary, emotional, or adrenaline-based content) could trigger increased heart rate, spikes in blood pressure, panic attacks, anxiety, PTSD, fainting, and other adverse effects. If you have a history of negative physical or psychological reactions to certain real life circumstances, avoid using the product to view similar content. Similarly, it is important to remember that simulated objects, such as furniture, that may be encountered while using the product do not exist in the real world, and injuries may result when interacting with those simulated objects as if they were real, for example, by attempting to sit down on a virtual chair.28

Similar language is used in Google’s Daydream View Health and Safety Information guide.29 Although users may approach VR as if it were just another game, VR companies are cognizant of the short and long-term physical and cognitive harm that immersive sensory experiences may cause users. The first harm mentioned in HTC’s safety guide is psychological harm. VR companies should be aware that the lifelike nature of their products may cause real life psychological trauma and illness. Depending on the VR experience, users could either suffer the effects of post-traumatic stress disorder after experiencing something triggering, or experience something stressful and new that could cause the condition. For example, the armed forces do not use the Oculus Rift system for military training because, despite its advancement, it still has a high enough latency rate that it could trigger nausea or cause a user to act unpredictably, compared to the higher-end models currently in use by the armed services. It has been proposed that long-term exposure to VR could even undermine an individual’s ability to process and respond appropriately to threats if she repeatedly subjects herself to catastrophic experiences but walks away unharmed. Decades of research has shown that VR can affect the way people act and think—lawyers and developers must consider this facet of VR as they roll out more and more widely available commercial products.30 27   Enforcing Privacy Promises, Federal Trade Commission, www.ftc.gov/news-events/mediaresources/protecting-consumer-privacy/enforcing-privacy-promises (last visited Aug. 11, 2017). 28   Safety and Regulatory Guide, HTC, http://dl4.htc.com/vive/safty_guide/91H02887-08M%20Re​ v.​A.PDF. 29   Daydream View Health and Safety Information, Google, https://support.google.com/day​ dream/answer/7185037?hl=en&ref_topic=7184600. 30   See Brian Crecente, “We’re running with scissors”: Why Some Experts Worry about VR Dangers, Polygon (Apr. 7, 2017), www.polygon.com/features/2017/4/7/15205366/vr-danger-close.

BARFIELD PRINT.indd 53

13/11/2018 11:52

54  Research handbook on the law of virtual and augmented reality In response to such harm, a user could attempt to file a tort suit against a VR platform provider or app developer. An absolute defense to intentional torts is consent, and a claim in that category is not likely to succeed if the user participated in the experience willingly. A user could attempt to file a negligence or strict products liability claim but precedent shows that such a claim is unlikely to succeed against today’s VR companies if the harm is premised on content rather than some hardware defect. In Sanders v. Acclaim Entertainment, Inc., family members of the victims of the 1999 Columbine high school shooting sued movie and video game producers for negligence and strict products liability, claiming that their violent products influenced the shooters.31 The federal court dismissed both claims, reasoning that the manufacturers had no duty to the victims’ families for the negligence claim and that the intangible ideas and thoughts contained in the media at issue were not subject to the strict liability doctrine. If a user were to bring a similar claim against a VR company, that company could argue that while their product is immersive, it too is only a collection of intangible ideas and thoughts. Although VR promises to be an immersive, tangible experience, it is unclear if, despite the industry’s progress, it has advanced far enough for its content to legally be considered tangible enough to be subject to strict liability doctrine. Some users still suffer from motion sickness caused by latency issues and may not feel fully immersed even with today’s high-tech peripherals. VR has yet to reach a point where it can be said to be an experience entirely divorced from the video game industry. As long as VR is still closely connected in content and experience to video games, the industry can likely rely on such precedents to defend itself against similar legal claims. History shows that technology advances faster than researchers predict and courts may need to set new precedents sooner than one thinks to keep up with the psychological issues that may stem from immersion in a 3D VR environment. The second harm mentioned in HTC’s safety guide is physical harm. The language mentions that users could suffer harm from trying to interact with virtual objects that do not exist in the real world. Serious physical harm can also stem from general carelessness. In order to be immersed, users have their entire field of vision blocked either by a screen or by a blackout blindfold. The technology intentionally tricks a user’s brain into believing that what they are seeing is the world in which the user exists. It is thus no surprise that a user may accidentally inflict harm on herself or another, or cause other property damage. There are a number of ways in which a VR company could protect itself from liability. The first defense seen is the use of adequate warnings enclosed in safety guides. In Elvig v. Nintendo of America, Inc., the District Court for the District of Colorado granted Nintendo’s motion for summary judgment after the mother of a user sued the company when her son’s Wii remote strap broke and damaged her television set.32 More than 900 other reports to Nintendo had been made regarding property and personal injuries stemming from the defective strap. The mother had claimed that Nintendo violated Colorado’s Consumer Protection Act (CCPA) by engaging in deceptive trade practices by distributing inadequate safety straps, breached an implied warranty that the safety straps were fit for a particular purpose under the Uniform Commercial Code (UCC), and violated the

  Sanders v. Acclaim Entm’t, Inc., 188 F. Supp. 2d 1264 (D. Colo. 2002).   Civil Action No. 08-cv-02616-MSK-MEH, 2010 U.S. Dist. LEXIS 100643 (D. Colo. Sep. 23, 2010). 31 32

BARFIELD PRINT.indd 54

13/11/2018 11:52

Starting up in virtual reality  55 Colorado Product Liability Act (CPLA) in distributing straps that were unreasonably dangerous and defective. The court granted summary judgment to Nintendo for the plaintiff’s CCPA claim because she failed to carry her burden to provide proof that Nintendo falsely advertised its system and the remote’s athletic use without disclosing that the straps had a propensity to break. Similarly, the court granted Nintendo summary judgment for the UCC claim since the plaintiff failed to establish that she had obtained her system for a particular purpose other than that for which it would customarily be used as required for such a claim or that the strap was not fit for ordinary purpose. Likewise, the court granted Nintendo summary judgment for the CPLA claim because a safety card included in the system gave precise warnings that the same misuse of the controller engaged in by the plaintiff’s son would result in the harm caused by that misuse. This shows that precise warnings within a safety guide or a safety card may act as a sufficient defense if there is an external hardware failure that causes some type of harm. Another defense is for companies to allow users to assume these risks through the terms and conditions for use. While many of the major manufacturers have safety guides and some include information about the need to use HMDs in safe environments, generally the actual applications are only accessible if one has internet access. Since online services generally have terms of service, users can agree to assume the risk to these physical harms when they agree to use a platform’s web-based services. This requires a platform and app provider to add new sections to a terms of service agreement rather than relying on boilerplate language that does not account for the fact that physical harm may actually stem from a web-based service. Another option would be for competitors to form a coalition early on and lobby Congress for new laws that will limit a fledgling company’s liability for its users’ carelessness. The extent of damage that VR may cause, if any at all, is unclear, and it would be unsurprising if companies that are investing billions into the industry want to protect their investments preemptively, before an extreme case such as an accidental death causes another nuclear winter in terms of further development in the industry.33 A fourth option would be to create both software and hardware safety solutions. Lawyers can advise VR clients to put technical safety measures in place within the products themselves. At the platform and app store level, companies such as Oculus already provide a safety warning every time their platform is loaded up. This acts as a defense to failure to warn products liability claims, as outlined above in the Elvig case. However, content developers themselves can provide further, more specific warnings when an application is opened. Content developers can develop quick, creative ways for users to interact with and become cognizant of the dangers of use before they are able to participate in the experience itself. Content developers could even follow Niantic’s lead from the augmented reality Pokémon Go mobile game and limit the circumstances in which their content can be used. In that game, users attempt to capture randomly generated fictional creatures called Pokémon that appear to exist in different locations in the physical world through the screen of the user’s mobile device. As a safety measure, new Pokémon will not randomly populate or appear in the game’s map for capture if a user is moving at driving speed.   Id.

33

BARFIELD PRINT.indd 55

13/11/2018 11:52

56  Research handbook on the law of virtual and augmented reality At the hardware level, other external companies are creating peripherals that make the use of VR safer. While users tend to rely on swivel chairs so they can be stationary while still taking advantage of the content’s 360-degree immersion, various startups, such as Roto VR, have developed VR-enabled chairs to enhance users’ experience. Roto VR’s particular good is a motorized chair that was specifically built to allow for a more immersive experience while still allowing a user to be safely seated.34 This is just one example of hardware built for modern VR technology that has safety in mind. While content developers may not want to enter the hardware space, other hardware developers should consider making their own complementary safety peripherals and consider how the peripherals they already manufacture put their users in harm’s way. VR poses new risks and safety should be considered at every level of production.

5.  INTELLECTUAL PROPERTY The growth of virtual reality will allow users to connect with and use real-world products in a fashion that may never have been previously conceived of. VR companies will need not only to protect their own intellectual property (“IP”) in their software and hardware but also to take care not to infringe on the IP rights of other manufacturers, authors, or brand owners. Although replicating real-world goods may make an immersive 3D environment feel more lifelike, VR developers must be warned of the potential liability inherent in using another’s IP rights. 5.1  Patent Rights Patents are a negative right that allow an owner to prevent others from making, using, selling, offering for sale, and importing their claimed invention for a statutory, limited period of time. Under U.S. patent law, in order to claim this right the owner’s invention must be useful, novel, and non-obvious to a person having ordinary skill in the art. Another person will have infringed the owner’s patent if she uses, makes, or sells the claimed invention without authorization. The patent owner may then sue for infringement. Only a few prominent patent infringement cases have arisen in the VR context. One such case is Lamson v. United States.35 There, the patent holder for a method for “Virtual Reality Immersion Therapy for Treating Psychological, Psychiatric, Medical, Educational and Self-Help Problems” sued the United States in federal claims court for unauthorized use of methods covered by the patent. Although the court construed the phrase “immersive virtual reality environment” in favor of the plaintiff, defining it as a computer-generated three-dimensional immersive environment that allows movement and navigation,36 it ultimately granted summary judgment to the government, stating that the Patent Act’s medical use immunity protected the government from liability for its actions in this specific case.37   Roto Chair, www.rotovr.com/.   Lamson v. United States, 117 Fed. Cl. 755, 763 (2014). 36   Lamson v. United States, 110 Fed. Cl. 691, 704 (2013). 37   Under 35 U.S.C. § 287(c), if a medical practitioner engages in a medical activity that would be considered direct or inducement patent infringement, a patent holder may not bring a civil 34 35

BARFIELD PRINT.indd 56

13/11/2018 11:52

Starting up in virtual reality  57 In order to protect a developer’s own innovations in VR, she must file a patent to protect any new processes used to create the hardware or any innovations in the manufacture of the VR peripheral itself. In the VR industry, many different manufacturers are producing different but similar peripherals in order to produce the best VR experience. Thus, a claim in a VR patent must refer back to existing VR patents, but also must clearly outline what is different about the hardware that the developer seeks to protect. A developer may want to file a utility patent or a patent for a new or an improvement on an existing machine or manufacture.38 A developer may also want to file a design patent, which may be “issued for a new, original, and ornamental design embodied in or applied to an article of manufacture.”39 In addition to protecting her client’s novel invention, a lawyer must also consider the new ways in which a patent may be infringed within a VR environment. The first new issue is whether or not reproducing a real-world good in a VR environment would infringe on the patent owner’s rights to the real-world good. In this case, a virtual good will likely only explicitly violate a patent if the attorney filing the patent extended the patent to cover computer embodiments of her client’s claimed invention. One way to do this is to write additional patents or claims in a patent in order to include a digital reproduction of the product or process. While the explicit violation of patent seems unlikely in a 3D environment, how patent applications are being drafted could change to accommodate new technology. If that is the case, developers must know that they can be held liable for direct infringement. A patent owner may also assert a claim against a VR content producer through the doctrine of equivalents. This doctrine allows a patent owner to sue an alleged infringer for the use, production, or sale of a good that carries out a substantially similar function in a substantially similar way with a substantially similar result to her claimed invention. This may extend to a virtual representation of a good if it operates in the same way within the applicable laws of physics. The difference between the representation and the invention may be deemed trivial by the court and the producer may be found liable. VR platform owners such as HTC or Oculus may also be found to be secondarily liable for inducement infringement. An entity may be secondarily liable for infringement if it actively induces infringement of a patent.40 To prove inducement infringement, a plaintiff must show that (1) a third party actually infringed, that (2) the inducer knew of the patent, and (3) that there was knowing inducement of the infringing acts with a specific intent to encourage infringement by the third party.41 If platform owners include clauses in their terms of service stating that content producers may not reproduce patent protected goods in VR without obtaining proper permission, this will undermine any assertion that the platform owner knowingly induced the infringing acts. Besides p ­ rotecting her own

action or seek damages from them or any related health care entity for the conduct. 35 U.S.C. § 287(c). This is worth keeping in mind considering virtual reality’s long history of use for medical treatment. 38   Types of Patents, www.uspto.gov/web/offices/ac/ido/oeip/taf/patdesc.htm (last visited Aug. 11, 2017). 39   Id. 40   35 U.S. Code § 271(b). 41   Minn. Mining & Mfg. Co. v. Chemque, Inc., 303 F.3d 1294, 1304–5 (Fed. Cir. 2002).

BARFIELD PRINT.indd 57

13/11/2018 11:52

58  Research handbook on the law of virtual and augmented reality patents, a developer must also make sure not to infringe another’s patent either in the development of her own hardware and software or in the reproduction of goods within the 3D environment. 5.2  Trademark Rights Besides patents, a VR developer must also ensure that she protects her own brand and does not infringe upon the brand of others. A trademark is a word, symbol, device, or name used to distinguish between the goods or services of one seller and another.42 A party infringes another’s trademark when she uses the trademark owner’s mark in commerce and that use causes or is likely to cause confusion as to the source of the good or service. Once a company has made an improvement on VR technology or produced unique VR content, it is imperative that the developer protect the brand with a federal and international trademark registration. Although this process has a number of steps and is not mandatory, in order to aid in the growth of a company, registration is important to provide “notice to the public of the registrant’s claim of ownership of the mark, [obtain] legal presumption of ownership nationwide, and [obtain] exclusive right to use the mark on or in connection with the goods/services listed in the registration.”43 This should be one of the first steps taken by a new VR company. It is also important to be cognizant of the new ways in which a producer may infringe on another’s trademark within VR content. One way to create a more realistic immersive world is to reproduce the goods found in the physical world and the brands that produce them. Replication of a mark in a virtual world may facially seem like a straightforward case of infringement. However, VR developers have three plausible arguments that such replication is not infringement. The first defense is based on the First Amendment and the right to make permissible parodies of marks.44 Use of another’s mark is not infringement if the alleged infringer can show that use of the mark does not cause consumer confusion in light of trademark law’s multi-factor test for confusion. Such a use is fair use if a reasonable viewer would not think the mark owner was a source of the parodying good. A representation of a mark is a parody if it simultaneously presents itself as the original and communicates that it is not the original through some articulable element of satire or ridicule.45 The right to make parodies is protected by the First Amendment. For example, in E.S.S. Entertainment v. Rock Star Videos, Inc., the Ninth Circuit held that use of a mark in an artistic work would not violate the Lanham Act unless the mark has no artistic relevance to the underlying work. If it does have artistic relevance, it will only violate the Act if it explicitly misleads consumers as to the source of the work. In that case, the court found that the parodic reproduction of an East Los Angeles nude dance club in a video game had artistic relevance but did not explicitly mislead consumers as to the source of either   See 15 U.S.C. § 1127.   Trademark Basics, USPTO, www.uspto.gov/trademarks-getting-started/trademark-basics (last visited Aug. 11, 2017). 44   Louis Vuitton Malletier, S.A. v. My Other Bag, Inc., 156 F. Supp. 3d 425 (S.D.N.Y. 2016). 45   Louis Vuitton Malletier, S.A. v. My Other Bag, Inc., 156 F. Supp. 3d 425, 434–35 (S.D.N.Y. 2016). 42 43

BARFIELD PRINT.indd 58

13/11/2018 11:52

Starting up in virtual reality  59 the game or the club—thus the video game publisher did not infringe on the trade dress of the club under Section 43(a) of the Lanham Act and its acts were protected by the First Amendment. Trade dress refers to the overall image of a good or service which may receive the same protection as a trademark if it serves to identify and distinguish the source or a good or service.46 While this case sets a useful precedent against claims of trademark infringement in future immersive worlds, its relevance may be limited. The court’s determination was made based on the facts of the case and future cases will likely also be subject to the same specific analysis. In future cases, those making more realistic 3D worlds may have to take greater caution, as a hyperrealistic rather than parodic world may make a consumer more likely to believe that the trademark owner was somehow involved in production. As a second defense, a trademark owner’s claim of infringement may also fail if the defendant can prove that the owner failed to establish that consumers were confused as to the owner’s affiliation with or sponsorship of the good or service in the virtual world. A defendant will likely not be held to have falsely presented an affiliation or sponsorship if the defendant has either criticized the mark, identified the mark’s owner, or even disavowed affiliation with the mark. Similarly, a defendant may also claim that their use constitutes nominative fair use. Under this doctrine, use of a mark is not infringing if the mark is only used to describe the goods or services of the trademark owner or its geographic origin in good faith. A use is not infringing if the VR developer is attempting to describe or comment on a good in good faith. There can also be no claim of improper affiliation if there is a true affiliation. Just as brands do with video games, VR content producers can intentionally enter into product placement deals. Furthermore, there are a number of cases where trademark infringement claims against video game manufacturers were dismissed because the court found that it was unlikely for consumers to believe that they were purchasing a copy of the trademark owner’s goods when they purchased the relevant game.47 These cases will likely inform the first era of litigation in regard to misuse of trademarks in VR. The third defense that an alleged infringer may assert is that the use of the mark was not in commerce, as required by the Lanham Act. Use in commerce is defined as a bona fide use in the ordinary course of trade.48 If the mark is used on virtual goods that a user can sell or exchange solely for virtual currency, it is unclear whether such a transaction constitutes use in commerce. It is also unclear if the exchange of fictional, virtual ­currency—even if it is made across state lines through an internet connection—is something that a federal statute may govern. A trademark owner will have an even weaker case if the good or service with which the mark is associated is not part of any transaction at all. A trademark owner could plausibly argue that being part of any sort of VR content that a consumer obtains is use in commerce, but this application of the Lanham Act would be overly broad. Such an application would allow any party to attempt to apply the   1-2A Gilson on Trademarks § 2A.01.   Brittany Frandsen, Is Using Call of Duty in This Comment Infringement? BYU L. Rev. 295, 301 (2016), http://digitalcommons.law.byu.edu/cgi/viewcontent.cgi?article=3031&context=lawreview. 48   15 U.S.C. § 1127; Alexandra MacKay, Defining Use in Interstate Commerce in Trademark Applications, INTABulletin (Oct. 1, 2016), www.inta.org/INTABulletin/Pages/Defining_Use_in_ Interstate_Commerce_7117.aspx. 46 47

BARFIELD PRINT.indd 59

13/11/2018 11:52

60  Research handbook on the law of virtual and augmented reality Lanham Act to all virtual commercial transactions, whether those transactions concern real goods or are simply a component of a fictional immersive world. Likely, whether or not the use of a mark in VR constitutes “use in commerce” will have to be litigated before any final resolution can be found. Attorneys working with VR platform providers can set a precedent by requiring that content producers have a license to use a trademark before reproducing it in VR. Nevertheless, if a license is not obtained, the First Amendment and trademark fair use have been shown to provide some protection for the reproduction of real life brands. 5.3 Copyrights With the production of new and old goods in 3D virtual environments primarily through software code, VR developers must also take the time to protect their copyrights and respect the copyrights of others. Article I, Section 8, Clause 8 of the US Constitution grants Congress the power to “promote the Progress of Science and useful Arts, by securing for limited Times to Authors and Inventors the exclusive Right to their respective Writings and Discoveries.” The Copyright Act, derived from this clause, specifically grants a specific set of exclusive rights to authors whose works are fixed in a tangible medium of expression.49 The Act grants authors the exclusive right to reproduce, create derivatives, distribute, publicly perform, and display their original works.50 Under Section 102 of the Copyright Act, the subject matters of the works include literature, music, plays, choreography, pictorial or sculptural works, audiovisual material, sound recordings, and architectural works.51 VR content likely garners protection in multiple ways. First, if the experience is put together through the development of software code, the authors will have a copyright in the code itself. However, considering that VR is an immersive experience, content developers will likely also be able to assert copyrights in the other audiovisual elements of the experience itself—the sounds, images, and any video content. A copyright is created in the instant when the original work is fixed in a tangible medium of expression and is able to be perceived. In order to obtain certain damages in the event of a lawsuit for infringement, an author must register her work with the Copyright Office. Because of the complex intersection of different types of copyright-protected material, it is likely worthwhile for a VR content developer or even a platform provider who creates an immersive portal to other VR content to take an expansive approach and register copyrights for all relevant material. To protect against the improper violation of copyright by users rather than competitors, VR content producers could also use technological protection measures. VR content has limited utility outside existing platforms and hardware devices. Yet, creators could still use encryption or the digital rights management software to stop the replication and use of their content in unauthorized hardware systems. As for the reproduction of real-world copyright-protected content within VR experiences themselves, copyright owners may have claims against VR developers if their works

  17 U.S.C. § 102.   17 U.S.C. § 106. 51   17 U.S.C. § 102. 49 50

BARFIELD PRINT.indd 60

13/11/2018 11:52

Starting up in virtual reality  61 are reproduced in the 3D immersive environment without authorization. Yet, if VR follows precedent set in the gaming industry, it is unlikely that copyright owners will be able to protect their work to the same extent as they would be able to in the physical world, since it may not be financially worthwhile to pursue infringers and limiting reproduction of digital content is much more difficult than limiting reproduction of physical works. In order to prove that an owner’s copyright has been infringed, the owner must show (1) that she has a valid copyright and (2) that actual copying of original elements of the plaintiff’s work occurred.52 This is a fact-intensive inquiry and can cost parties a significant amount if they choose to take a case to trial. Furthermore, damages are limited—a plaintiff may generally only collect statutory damages if the copyright was registered before the infringement took place, or actual damages and a defendant’s additional profits. If the alleged infringer makes either low or no profits from the use of the virtual goods or the distribution of an infringing VR application, the lawsuit as a whole may not be worthwhile. Such a suit would then only be pursued to make an example of the infringer. For some larger corporations this may be something that the company can invest in, but it likely will not be a practical remedy for copyright owners with more limited means. In regards to infringement within the VR industry itself, this state of affairs provides a recourse to the industry’s larger players while limiting the options of smaller companies that produce pioneering content. Enforcement of copyright will likely take the creation of open communications channels between VR platform providers and app distributors, such as HTC and Steam, and content producers. Copyright owners would likely have a higher chance of having infringing content removed if they make a claim to the platform provider that the infringer has violated the clause asserting that she has not used anyone else’s intellectual property without permission, which is commonly found in most terms of service or end-user license agreements for today’s online services. This will likely take place through the filing of a claim under the Digital Millennium Copyright Act (DMCA),53 enacted in 1998. Title II of the bill, called the “Online Copyright Infringement Liability Limitation Act,” added Section 512 to the Copyright Act, which limited service providers’ liability for potentially infringing material posted online by third parties through their services. In order to protect themselves from liability for the infringing acts of users, online service providers must have a notice and takedown procedure meeting Section 512’s statutory requirements.54 This allows copyright owners to request to have infringing content removed. In order for this procedure to be applicable, VR platform providers who control the distribution of VR content in both mobile and desktop systems must be considered online service providers. Online service providers are “provider[s] of online services or network access, or the operator of facilities therefore.”55 This has been interpreted broadly to include services existing purely on the web, such as social networks, and will most likely extend to the app store-like platforms controlled by VR’s most prominent hardware developers. However, the possibilities of VR complicate matters. Who is the service provider

    54   55   52 53

4 Nimmer on Copyright § 13.01. Digital Millennium Copyright Act, 1998 Enacted H.R. 2281, 112 Stat. 2860, 105 P.L. 304. 17 U.S.C. § 512(c). 17 U.S.C. § 512(k).

BARFIELD PRINT.indd 61

13/11/2018 11:52

62  Research handbook on the law of virtual and augmented reality if a user creates infringing content within an application? Whom does a copyright owner reach out to first? Must there be a notice and takedown procedure at each level, even for a small VR social network that has not gained any prominence yet? Established gaming companies, famous studios, and new startups are all competing for attention and usage within the space—this poses unique and serious challenges to copyright owners attempting to assert their rights. It is unclear if the DMCA and copyright litigation will provide adequate remedies.

6. CONCLUSION A host of new legal issues and challenges to existing law arise when considering VR as a new growth industry. Lawyers advising on VR may look to the case law concerning video games in the near future, since it is likely the subject with the most closely related fact situations. Yet, video game companies have not traditionally had to consider the biometric data being collected from users. Nor have they had to consider the harm and liability that stems from having users participate in fictional but increasingly tangible, lifelike digital immersive experiences. Both forms of new media likely require a more targeted reinterpretation of intellectual property law, since they involve complex multimedia that grows more realistic in both its visuals and its immersive nature as time passes. VR has never been used just for entertainment and as it starts to affect our healthcare, education, and business dealings, its own body of law will develop as it gains wider distribution. Until then, lawyers working with growing virtual reality companies must understand both the familiar and unique challenges faced by a company in this sector.

SOURCES Cases Civil Action No 08-cv-02616-MSK-MEH, 2010 U.S. Dist. LEXIS 100643 (D. Colo. Sep. 23, 2010). Fawzy Amer Deghedy v. Viztek, Inc., No. 3:12-cv-00048-HCA, 2016 U.S. Dist. LEXIS 181097 *2 (S.D. Iowa June 10, 2016). FTC v. Erik Chevalier also d/b/a The Forking Path, Co., Case 3:15-cv-01029-AC (D. Or. 2015). Lamson v. United States, 110 Fed. Cl. 691, 704 (2013). Lamson v. United States, 117 Fed. Cl. 755, 763 (2014). Louis Vuitton Malletier, S.A. v. My Other Bag, Inc., 156 F. Supp. 3d 425, 434–35 (S.D.N.Y. 2016). Minn. Mining & Mfg. Co. v. Chemque, Inc., 303 F.3d 1294, 1304–5 (Fed. Cir. 2002). Sanders v. Acclaim Entm’t, Inc., 188 F. Supp. 2d 1264 (D. Colo. 2002). Total Recall Techs v. Luckey, No C 15-02281 WHA (N.D. Cal. 2016). Total Recall Techs v. Luckey, No C 15-02281 WHA, at *11 (N.D. Cal. Mar. 9, 2017). Zenimax Media, Inc. v. Oculus VR, LLC, 166 F. Supp. 3d 697 (N.D. Tex. 2015).

Statutes Digital Millennium Copyright Act, 1998 Enacted H.R. 2281, 112 Stat. 2860, 105 P.L. 304. H.R. 3606—112th Congress: Jumpstart Our Business Startups Act; 112 P.L. 106. 15 U.S.C. § 45(a). 15 U.S.C. § 1127.

BARFIELD PRINT.indd 62

13/11/2018 11:52

Starting up in virtual reality  63 17 U.S.C. § 102. 17 U.S.C. § 106. 17 U.S.C. § 512(c). 17 U.S.C. § 512(k). 35 U.S.C. § 287(c) 35 U.S. Code § 271(b).

BARFIELD PRINT.indd 63

13/11/2018 11:52

3.  Virtual rule of law Michael Risch*

1. INTRODUCTION The rule of law generally requires that governments announce and follow the laws of land. This allows citizens to know what to expect from their government and to make investments accordingly. Entrepreneurs want to know what activities are legal, whether the government can interfere in business, whether agreements are enforceable, and whether harmful actions by others will be stopped. For this reason, many consider the rule of law a catalyst for economic development, and there is reason to believe that it will be equally important in virtual economies, despite their differences with the real world. As more people turn to virtual worlds for both fun and livelihood, the rule of law will become prominent in encouraging investments in virtual business.1 This chapter considers whether virtual worlds provide a rule of law that sets expectations for virtual life. It is not surprising that—for the most part—they do not; virtual worlds currently lack many of the elements of the rule of law. Which aspects fail is more surprising, however. Provider agreements and computer software, the sources of regulation that are most often criticized as “antiuser,” provide the best theoretical hope for achieving the rule of law, even if they currently fail in practice. On the contrary, widely proposed “reforms,” such as community norms, self-regulation, and importation of real world law, face both theoretical and practical barriers to implementation of the rule of law in virtual worlds. These conclusions follow from a four-step analysis. Section 2 is a brief introduction to virtual worlds. Virtual worlds allow their users nearly unlimited choices in their actions and interactions with other users and the virtual environment. As a result, business and social interactions permeate virtual worlds just as they do elsewhere. If a user can deliver something of value, other users will trade something else of value (including real money) for it. Section 3 introduces the rule of law and its importance. However, the analysis does not attempt to determine just how important the rule of law might be; virtual worlds are not conducive to such normative determinations. Instead, this part examines the literature

*  This chapter is based on several of the author’s works and an earlier version can be found at 112 W. VA. L. REV. 1 (2009). The author thanks participants in that symposium, as well as Joshua Fairfield, Eric Goldman, James Grimmelman, Greg Lastowka, Anne Lofaso, Brian Tamanaha, Francis Taney, and Tim Wu for their helpful comments. Helpful research assistance was provided by Nate Griffith. 1   John Rawls, A Theory of Justice 235 (1971) (“A legal system is a coercive order of public rules addressed to rational persons for the purpose of regulating their conduct and providing the framework for social cooperation. When these rules are just they establish a basis for legitimate expectations”).

64

BARFIELD PRINT.indd 64

13/11/2018 11:52

Virtual rule of law  65 and abstracts a generalized positivist test of ten rule of law indicia. Rule by law must be: (1) nonarbitrary, (2) stable, (3) public, (4) nondiscretionary, (5) comprehensible, (6) prospective, (7) attainable, (8) consistently enforced, (9) impartially applied, and (10) adjudicated in a factually neutral way. These indicia, however, do not include traditional notions of “liberal” rule of law, including democracy and personal rights. Just as normative measurements of the rule of law’s impact on virtual business are unhelpful, so too is an examination of liberal ideals in the context of virtual worlds. Technical limitations and user preferences render notions such as democracy and legitimacy difficult to apply in virtual worlds. Furthermore, to the extent that setting expectations is the most important feature with respect to business, liberal values are less important. Thus, the analysis is descriptive and formalistic, such that providers and users can objectively determine whether the rule of law is present in a world and make decisions based on such determinations. Section 4 examines law and sovereignty in virtual worlds. The four primary sources of law follow from general cyberlaw principles developed by Professor Lessig and others. Law is a constraint on behavior, and such constraints are imposed by four sources: the market (such as provider agreements), code (the virtual world software), norms (community defined rules), and real-world law (legislation and case law). Three potential sovereigns impose this law: real government, the virtual-world provider, and the virtual community. Further, the law is imposed on both the user and her online persona—her avatar. While these dimensions imply that law may come in 24 different flavors, this part narrows the consideration down to a single question: “How does the law, from whatever source, affect the end user?” This analysis focuses on human, rather than avatar, wellbeing. Section 5 draws upon the previous discussion to critically examine the rule of law in virtual worlds. The chapter argues that agreements and code offer the best possibilities for implementing the rule of law in virtual worlds, even though they currently fall short. Provider agreements and code can exhibit all of the indicia of the rule of law. They can be nonarbitrary, stable, and public. However, these sources often fail in practice. Agreements are not consistently enforced, and amendments might be arbitrary or retroactive. Code, on the other hand, is perfectly enforced but is frequently hidden and potentially arbitrary and unstable. Other sources, such as norms and real-world laws, are unlikely to provide the rule of law in virtual worlds either in theory or in practice. Community norms are often unwritten, unspoken, and partial. The very nature of community enforcement is based on vigilantism, the antithesis of rule by law. The application of real-world rules is also unlikely to fulfill the indicia, especially in the short run. The nonphysical nature of virtual conduct, as well as game rules that allow activity that would certainly be disallowed if perpetrated among humans, means that determining which laws should apply is impossible. Many have suggested looking to sports rules for the answer, but the analogy is too attenuated to provide hope that the rule of law will be present in theory or in practice. Based on the general finding that the rule of law is currently absent but theoretically achievable in virtual worlds, the chapter concludes with some suggestions about how the rule of law might be enhanced. The suggestions focus on some of the key shortcomings that might practically be changed, such as neutral factfinding, more detailed in-game rules with finely tuned penalties, and contract enforcement by users. It may be, however, that time may best improve the rule of law. As more disputes are resolved, a body of law will grow that can apply more faithfully to conduct in virtual worlds.

BARFIELD PRINT.indd 65

13/11/2018 11:52

66  Research handbook on the law of virtual and augmented reality

2.  VIRTUAL WORLDS For the uninitiated, a virtual world is an interactive computer software program running on a central server (in modern times, over the internet). Many human users control characters—called “avatars”—in the game software from any computer attached to the network. Participating in a virtual world is like playing a video game, except that the other characters are also humans who control their own avatars.2 A more recent version of such worlds is “augmented” reality worlds, in which computer-generated elements overlay real ones. Virtual worlds allow avatars to do a range of things: walk, fly, build virtual buildings on virtual real estate, and otherwise interact. Figure 3.1 shows several avatars enjoying a live concert played by another avatar in the virtual world Second Life.3 Second Life’s software allows users to transmit live or prerecorded performances to the virtual world, which are then broadcast to other avatars in the same virtual location. And children

Figure 3.1  Virtual concert in Second Life

2   F. Gregory Lastowka & Dan Hunter, Virtual Crimes, 49 N.Y.L. Sch. L. Rev. 293, 309 (2004), hereafter Virtual Crimes. 3   Graphic located at: http://s3.amazonaws.com/static-secondlife-com/screenshots/ web/int_open​ mic.jpg (available at https://perma.cc/95PG-YGMB).

BARFIELD PRINT.indd 66

13/11/2018 11:52

Virtual rule of law  67 everywhere play Minecraft, a pixelated world where a primary goal is to build and destroy structures. Many worlds have some form of virtual money. Many allow avatars to obtain virtual property and to transfer that property to others. Many also incorporate varying levels of social status, especially worlds that involve combat with computer-controlled enemies; the more enemies killed the higher the “level” an avatar might achieve. Rarer property and higher levels are valuable to those who do not want to spend their own time appropriating them. Virtual worlds are home to serious business.4 One study suggests that virtual economies may reach the size of small countries.5 Businesses are varied, from mining virtual gold to real gambling and anything inbetween.6 As a brief example, the concert depicted in Figure 3.1 might include a few types of businesses. A concert promotion business might pay the virtual land owner for the right to use the space for a concert. A fledgling musician might pay the promotion company for the right to play a show—advertising for real-world music. Conversely, the promotion company might pay an established musician to play in the venue, so that the patrons might pay for the right to listen to the music. All of these payments are made using either real money or virtual money (which can be exchanged for real money or other items of value).

4   Edward Castronova, Synthetic Worlds 255 (2005) (describing shift of economic activity to virtual worlds); David G. Post, In Search of Jefferson’s Moose: Notes on the State of Cyberspace 182 (2009) (“Many hundreds or thousands or hundreds of thousands of people (precise statistics are not easy to come by here) are, at the moment, earning some or all of their living in Second Life”); Andrea Vanina Arias, Life, Liberty, and the Pursuit of Swords and Armor: Regulating the Theft of Virtual Goods, 57 Emory L.J. 1301, 1301 (2008) (market may reach $2 billion). 5   Edward Castronova, Virtual Worlds: A First-Hand Account of Market and Society on the Cyberian Frontier 33 (CESifo Working Paper Series No 618, 2001), at http://ssrn.com/ abstract_id=294828 (last visited May 25, 2009). Though often cited, this calculation is not the best estimator of the size of virtual worlds. The study compares per capita production, which may not be comparable to the real world. The real world is home to many “unproductive” citizens: infants and children, the disabled and elderly, and voluntarily and involuntarily unemployed adults. In virtual worlds, every avatar is a productive member of society. 6   Greg Lastowka & Dan Hunter, The Laws of the Virtual Worlds, 92 Cal. L. Rev. 1, 10 (2004), hereafter Laws (describing different purchases one might make in a virtual world); Post, supra note 4, at 181 (“They make stuff—clothing and jewelry for their avatars, huge buildings, paintings to put on the walls of those buildings, automobiles or airships that can transport them from one ‘place’ to another in the virtual world, videos . . . and they exchange what they make with others; if you like the virtual clothing or the virtual jewelry I’m wearing, or the virtual picture I’ve painted, or the virtual building or virtual airship that I’ve created, you can try to persuade me to give it to you. Or sell it to you. For money. Not ‘real money,’ of course – play money, game money . . . But here’s the thing: It turns out that it is real money. Linden Dollars can be exchanged for things of value, including . . . U.S. dollars”).

BARFIELD PRINT.indd 67

13/11/2018 11:52

68  Research handbook on the law of virtual and augmented reality

3.  THE RULE OF LAW AND VIRTUAL ECONOMIES 3.1  The Tie between Law and Virtual Business It is relatively well settled that economies thrive under the rule of law.7 Some commentators go further, arguing that economies will fail without the rule of law.8 Regardless of how strong the tie may actually be, governments throughout the world favor the rule of law as a harbinger of economic stability.9 Virtual economies are no different in theory, and the rule of law will become more important as virtual commerce expands.10 When planning business investments, every entrepreneur will ask questions: Can I operate this business now? Will I be able to do so in the future? What are the penalties if I break a rule? Will the rules be enforced against others? How do I ensure that my customers follow the rules? Professor Post adds: How many people are going to give their hard-earned money—real money!—to Chiaretta Charron [a virtual banker] without some assurance that she (or he, or it) will behave reasonably with it? How many people will extend credit to anyone else without some way to enforce the obligation? How many people will invest large amounts of time or effort or money in any substantial undertaking—building a law school, say, or organizing a recording studio—without some assurance that it won’t be destroyed by other participants in the “game,” or by the ­operators of the virtual world themselves, for “no good reason” at all.11

The presence (or absence) of the rule of law helps answer these questions.12  7   Brian Z. Tamahana, On the Rule of Law: History, Politics, Theory 119 (2004) (“A growing body of evidence indicates a positive correlation between economic development and formal legality that is attributable to these characteristics [of the rule of law]”) and 53 (describing Montesquieu’s view of the rule of law as facilitating commerce by “facilitating their transactions, enforcing their agreements, protecting their property, and otherwise leaving them be”). See also, e.g., O. Lee Reed, Law, The Rule of Law, and Property: A Foundation for the Private Market and Business Study, 38 Am. Bus. L.J. 441 (2001); Kevin J. Fandl, The Role of Informal Legal Institutions in Economic Development, 32 Fordham Int’l L.J. 1 (2008); Kevin E. Davis & Michael J. Trebilcock, The Relationship Between Law and Development: Optimists Versus Skeptics, 56 Am. J. Comp. L. 895 (2008); Norman L. Greene, Perspectives from the Rule of Law and International Economic Development: Are there Lessons for the Reform of Judicial Selection in the United States, 86 Denv. U. L. Rev. 53 (2008); Todd J. Zywicki, The Rule of Law, Freedom, and Prosperity, 10 Supreme Court Economic Review 1, 22 (2003).  8  Arias, supra note 4, at 1339 (real world social order could break down without virtual law).  9   Rachel Kleinfeld Belton, Competing Definitions of the Rule of Law: Implications for Practitioners, Carnegie Paper No 55 (Carnegie Endowment 2005), available at http://www.­carnegieen​ dowment.org/publications/index.cfm?fa=view&id=16405. See also, John Locke, Second Treatise of Government s.202 (Hackett Publishing 1980) (“Where-ever law ends, tyranny begins”). 10   Jack M. Balkin, Virtual Liberty: Freedom to Design and Freedom to Play in Virtual Worlds, 90 Va. L. Rev. 2043, 2044 (2004) (“Even at this early stage of technological development, people have simply invested too much time, energy, and money in virtual worlds to imagine that the law will leave these worlds alone”). Compare James Grimmelmann, Virtual World Feudalism, 118 Yale L.J. Pocket Part 126 (2009), hereafter Feudalism (describing virtual worlds as feudal societies) with Tamahana, supra note 7, at 29–31 (describing the rise of rule of law and the related demise of feudalism). 11  Post, supra note 4, at 184. 12  Post, supra note 4, at 184 (“The answer, I think, is: not nearly as many people as would do

BARFIELD PRINT.indd 68

13/11/2018 11:52

Virtual rule of law  69 However, this chapter is unconcerned with just how important the rule of law might be in a given virtual economy, because there are so many to choose from. Virtual-world providers are free to establish the type of system they deem best. Providers have done so—there are hundreds of virtual worlds, ranging from free-for-all worlds targeted at fraud and fighting to tightly controlled worlds targeted at children. Similarly, users are not burdened with any particular world by birth and may choose which virtual world to join, if at all. User wellbeing is determined in large part by individual happiness-maximizing choices in the market, of which only a small portion might relate to business income.13 Thus, users choose worlds based on their own perception of the rule of law, among other things.14 As a result, normative considerations about the level of rule of law have no place in this particular discussion, for four reasons.15 First, users might well prefer a lawless virtual society for many reasons, including the reckless abandon that comes with shedding real-world risks. These users should not be told that their choices are morally wrong or socially nonoptimal,16 nor should the virtualworld provider be criticized for offering lawless outlets.17 so if there were a functioning legal system in place”); Jack Goldsmith & Tim Wu, Who Controls the Internet? 129 (2006) (“eBay quickly learned that to prevent fraud, enforce its contracts, and ensure stability in its auction services, it would depend critically on government coercion and the rule of law”) and 145 (eBay’s refusal to expand into Russia, which “suffers from private harms gone unchecked: insecurity of private property, corporate fraud, a failed criminal law system . . . and ineffective respect for and enforcement of contract rights”). 13   Raph Koster, Declaring the Rights of Players, The State of Play: Law, Games, and Virtual Worlds 63 (Jack M. Balkin & Beth Simone Noveck, eds 2006) (“[T]he common good is that which increases the population of a [world] without surrendering core social tenets or mores”); Phillip Stoup, The Development and Failure of Social Norms in Second Life, 58 Duke L.J. 311, 313 (2008) (“The optimal mix between code-created rules and real-world regulations could be determined by finding the ‘mix that provides optimal protection at the lowest cost’”). Section 2.4 discusses this concept further with respect to the role of liberty in the rule of law. 14  Castronova, supra note 4, at 262 (“The ideal future would have a broad portfolio of worlds for us to visit, and we would all be able to spend time in the worlds we prefer, whether or not their governments are legitimate”); Richard A. Bartle, Virtual Worldliness, The State of Play: Law, Games, and Virtual Worlds 37 (Jack M. Balkin & Beth Simone Noveck, eds 2006); Yochai Benkler, There is No Spoon, The State of Play: Law, Games, and Virtual Worlds 182 (Jack M. Balkin and Beth Simone Noveck, eds 2006); Caroline Bradley & A. Michael Froomkin, Virtual Worlds, Real Rules, 49 N.Y.L. Sch. L. Rev. 103 (2004) (suggesting that different virtual worlds might be used to test effectiveness of different sets of laws); Balkin, supra note 10, at 2050. But see Lastowka & Hunter, Laws, supra note 6, at 61–2 (questioning whether free choice works in practice). 15   Contra, e.g., Joseph Raz, About Morality and the Nature of Law, 48 Am. J. Juris. 1, 12 (2003); Matthew Sundquist, Online Privacy Protection: Protecting Privacy, the Social Contract, and the Rule of Law in the Virtual World, 25 Regent U. L. Rev. 153 (2012) (“The law should serve the common interest and secure values that will be broadly useful to society”). 16   Orin S. Kerr, Criminal Law in Virtual Worlds, 2008 U. Chi. Legal F. 415, 427 (2008), hereafter Criminal (“More broadly, a strong regime of criminal enforcement would threaten one of the foundational strengths of virtual world games: the ability of each virtual world to define its own terms and to appeal to specific users who want that virtual environment instead of another”); cf. Bradley & Froomkin, supra note 14, at 128–30 (discussing choices users make about player versus player combat). 17  Tamahana, supra note 7, at 5 (“Like all ideals, there are certain socio-cultural contexts for

BARFIELD PRINT.indd 69

13/11/2018 11:52

70  Research handbook on the law of virtual and augmented reality Second, in this market context, it is difficult if not impossible to normatively balance the needs of users and providers.18 One author, for example, argues that—for the good of the world—providers must have carte blanche to change game rules and code at any time, and with retroactive effect.19 This suggestion is contrary to the rule of law, but may be normatively superior for worlds that harbor the goals the author discusses, namely world evolution, user achievement, and exclusion of the real world.20 Further, while efficiency-maximizing economic analysis might yield normative answers about the entire market, such analysis is less helpful in assessing whether the rule of law in any particular world is normatively justified.21 Similarly, use of the rule of law to further moral rights is difficult in virtual worlds. For example, the formal rule of law described below might be present in authoritarian regimes where some rights are impinged.22 As discussed below, many—and perhaps most—virtual-world providers might be considered rights-limiting authoritarians. However, stripping providers of the right or ability to act authoritatively in the name of moral rights might mean that certain worlds close or never even open. Whether user “rights” outweigh the moral good that comes from the ability to be users in the first place is an unanswerable question. Third, a little lawlessness may be a preferable way to encourage innovation and improve wellbeing in virtual worlds. Schumpeter, for example, argued that innovation thrives with disruptive activities by entrepreneurs.23 Presumably, some lawlessness and breaking of pre-existing norms encourages Schumpeter’s “creative destruction.” For example, lawbreaking can create efficiencies and help identify unjust and inefficient rules.24 Further, use of others’ creative work can increase creativity as a whole.25 Because lawlessness may have a positive effect on entrepreneurship, categorical normative pronouncements about the rule of law would be incomplete.26 Fourth, in any event, there is no consensus about which aspects of the rule of law are

which [the rule of law] is ill suited, and it must be weighed against and sometimes give way to other important social values”). 18   Cf. Goldsmith & Wu, supra note 12, at 141 (discussing differing needs of and influential power of different groups). 19  Bartle, supra note 14, at 43. 20   Id. 21   See, e.g., Castronova, supra note 4, at 261 (arguing that whether users should own virtual property depends on the type of world and how isolated it is from the real world). 22  Tamahana, supra note 7, at 93, points out that “formal legality” is often used to justify authoritarian practices. 23   See generally Joseph A. Schumpeter, The Theory of Economic Development: An Inquiry into Profits, Capital, Credit, Interest, and the Business Cycle (Redvers Opie trans., Harvard University Press 1951) and Joseph A. Schumpeter, Capitalism, Socialism, and Democracy (3rd ed. 1950). 24   Eduardo Moises Penalver & Sonia K. Katyal, Property Outlaws, 155 U. Pa. L. Rev. 1095, 1103 (2007). 25   See generally Lawrence Lessig, The Future of Ideas: The Fate of the Commons in a Connected World (2001). 26   But see A.V. Dicey, Introduction to the Study of the Law of the Constitution xli (8th ed. reprinted 1923) (discussing moral problems of lawlessness by individuals “pursuing some end to which to him or to her seems to be just and desirable”).

BARFIELD PRINT.indd 70

13/11/2018 11:52

Virtual rule of law  71 normatively superior.27 In some societies, for example, preannounced rules may not be normatively superior.28 Therefore, this chapter focuses on the positive, rather than normative, aspects of the rule of law in virtual worlds.29 The question is not whether the rule of law should be implemented, but rather whether the rule of law is implemented. Armed with that information, providers, users, legislators, and judges can make their own choices.30 3.2  Defining the Rule of Law There are innumerable ways to define the “rule of law.” Some formulations require more than others do, and some conflict with each other.31 Thus, defining the rule of law is the subject of many books. One commentator recently noted: “Read any set of chapters discussing the rule of law, and the concept emerges looking like the proverbial blind man’s elephant—a trunk to one person, a tail to another.”32 The analysis here begins with the trunk and the tail—two possible definitions of the rule of law: (1) government must follow the law, or (2) rules must be announced beforehand and applied nonarbitrarily.33 While one could argue for either of these two formulations, among others, this chapter opts for the latter, for two reasons. First, examination of rulemaking provides for a richer examination with respect to virtual worlds. Second, focus on rules arguably incorporates the requirement that the government follow those rules. Hayek describes the combination: Stripped of all technicalities, [the rule of law] means that government in all its actions is bound by rules fixed and announced beforehand—rules which make it possible to foresee with fair

27   Gianluigi Palombella, The Rule of Law and Its Core, Relocating the Rule of Law 35 (Gianluigi Palombella & Neil Walker, eds 2009). 28  Tamahana, supra note 7, at 95 and 139–40. One potential exception is the argument that a rule of law defined by a government being bound by its own rules is a universal moral good. Tamahana, supra note 7, at 115 and 137 (describing government bound by law, which also requires limits on how the law can be changed by the government). Even if true, this view of the rule of law would limit the analysis in this chapter to just one data point; this chapter seeks to be more comprehensive. 29  Tamahana, supra note 7, at 94, discusses a variety of reasons why one might normatively prefer to examine the rule of law in formalistic terms. See also Palombella, supra note 27, at 35 (rule of law requisites in the positivist scheme are morally neutral). 30  Palombella, supra note 27, at 35 (“All these main prerequisites derive from the essential objective of the law, which is that of guiding behaviour”) (emphasis in original). 31   David Beatty, Law’s Golden Rule, Relocating the Rule of Law 99 (Gianluigi Palombella & Neil Walker, eds 2009) (confusion about what the rule of law means leads to its decline). 32  Belton, supra note 9, at 5. 33  Locke, supra note 9, at s.137; Tamahana, supra note 7, at 33; Belton, supra note 9, at 3; Dicey, supra note 26, at 180, 189, 198, and 323; Fuller, infra note 38, at 209–10 (“Surely the very essence of the Rule of Law is that in acting upon a citizen . . . a government will faithfully apply rules previously declared as those to be followed by the citizen and as being determinative of his rights and duties. If the Rule of Law does not mean this, it means nothing”); Rawls, supra note 1, at 235 (“[T]he conception of formal justice, the regular and impartial administration of public rules, becomes the rule of law when applied to the legal system”). The second view extends even further to the “liberal” rule of law—that individuals have rights and that government must be democratic. Rawls, supra note 1, at 235 (“The rule of law is closely related to liberty”). Section 2.4 discusses these extensions.

BARFIELD PRINT.indd 71

13/11/2018 11:52

72  Research handbook on the law of virtual and augmented reality certainty how the authority will use its coercive powers in given circumstances and to plan one’s individual affairs on the basis of this knowledge.34

3.3  Indicators of the Rule of Law Hayek’s combined definition is a blunt instrument; the most insignificant failure would technically negate the rule of law. Indeed, any sweeping definition is insufficient to fully consider the presence or absence of the rule of law.35 Fortunately, many have identified elements of the rule of law that might contribute to an analytical scalpel. Unfortunately, this chapter cannot examine every element, nor can it resolve every conflict between political philosophers about which elements are most important. Thus, the chapter aggregates several points of view, treating a variety of considerations as indicators of legal rule. The analysis thus assumes that no single element is either necessary or sufficient to prove that law rules in a virtual world. Rather, the more indicators, the more a virtual world is governed by law.36 Multifactored analysis makes intuitive sense as well; many readers might be surprised by the absence of several factors in their own countries that were previously thought to follow the rule of law. While there are theoretically dozens of indicators, the following ten provide a reasonable foundation for examining virtual worlds:  (1) Nonarbitrary:37 Laws should not be arbitrary or contradictory.38 Methods of measuring arbitrariness will differ from person to person,39 but rules should usually be general rather than ad hoc.40 For example, bills of attainder directed at a single person or small group have no place in a sovereignty governed by the rule of law.   Friedrich A. Hayek, The Road to Serfdom 72 (1944).  Tamahana, supra note 7, at 3 (“Notwithstanding its quick and remarkable ascendance as a global ideal, however, the rule of law is an exceedingly elusive notion”); Belton, supra note 9, at 26 (because the rule of law has many different ends, it is impossible to establish a unitary measurement of the rule of law); Dicey, supra note 26, at 183 (“[W]e may safely conclude that . . . whenever we talk of Englishmen as loving the government of law, or the supremacy of law . . . [we] are using words which, though they possess a real significance, are nevertheless to most persons who employ them full of vagueness and ambiguity.”). 36   See, e.g., Palombella, supra note 27, at 35 (“When dealing with the rule of law, legal theory concentrates typically on the features which law generally needs in order to rule”); Belton, supra note 9, at 6–7 (“The ends are the reason why we value the rule of law and are what most people mentally measure when determining the degree to which a country has the rule of law. Another type of definition describes the institutions a society must have to be considered to possess the rule of law. Such a society would have certain institutional attributes”). 37   “Laws,” “rules,” and “regulations” are used interchangeably here for variety. Section 3 discusses the sources of regulation in virtual worlds, whatever they might be called. 38  Locke, supra note 9, at s.137; Lon Fuller, The Morality of Law 39 (revised ed. 1969); Dicey, supra note 26, at 198. 39  Hayek, supra note 34, at 73–4, for example, argues that if the government must decide what buses to run, then it is arbitrary. See also Dicey, supra note 26, at xxxviii (judicial authority given to officials connected with elected government shows decline in rule of law); Tamahana, supra note 7, at 64–5 (discussing A.V. Dicey’s criticism of the administrative state as a degradation of the rule of law). 40  Fuller, supra note 38, at 39. 34 35

BARFIELD PRINT.indd 72

13/11/2018 11:52

Virtual rule of law  73  (2) Stable: Changes should be infrequent; frequent changes render laws arbitrary.41  (3) Public: The rules must be publicly available and, in modern times, written.42  (4) Nondiscretionary: Further, laws must be applied with minimal discretion.43  (5) Comprehensible: Regulations must be discernable and understandable.44  (6) Prospective: The sovereign must not punish an offense without a law proscribing that offense, and new laws should not be applied retroactively.45  (7) Attainable: The law must not require more than the affected parties have the power to accomplish.46  (8) Consistent Enforcement: Laws must be enforced regularly, accurately, and in accordance with their text.47  (9) Impartial Application: Regulations must be applied impartially, such that similar circumstances are treated similarly.48 (10)  Neutral Factfinding: A neutral arbiter of the truth should determine when the law has been broken. Some argue this indicator requires an independent judiciary,49 but the minimum is an independent and impartial factfinder.50 41  Locke, supra note 9, at s.137; Fuller, supra note 38, at 39; Belton, supra note 9, at 17; Tamahana, supra note 7, at 33. 42   Montesquieu, The Spirit of the Laws, Book XI, ch. 6 (Anne M. Cohler, et al., eds, Cambridge University Press 1989); Fuller, supra note 38, at 39; Belton, supra note 9, at 7; Tamahana, supra note 7, at 33; but see Dicey, supra note 26, at 196 (describing certain guarantees of freedom as simply “the law of the land” rather than being guaranteed in a written document). 43  Montesquieu, supra note 42, at Book XI, ch. 6; Dicey, supra note 26, at 198; Hayek, supra note 34, at 72–3 (“[T]he discretion left to the executive organs wielding coercive power should be reduced as much as possible”). This traditional requirement is interesting, as discretion might be considered normatively valuable in a free society. Montesquieu, for example, suggests that the legislature have the power to “moderate the law in favor of the law itself by pronouncing less rigorously than the law.” Book XI, ch. 6. 44  Fuller, supra note 38, at 39. 45  Dicey, supra note 26, at 198; Tamahana, supra note 7, at 33; Rawls, supra note 1, at 238 (“there is no offense without a law”); Fuller, supra note 38, at 39. 46  Fuller, supra note 38, at 39. 47  Montesquieu, supra note 42, at Book VI, ch. 3; Fuller, supra note 38, at 39; Belton, supra note 9, at 3, 17; Tamahana, supra note 7, at 33; Rawls, supra note 1, at 235 (“One kind of unjust action is the failure of judges and others in authority to apply the appropriate rule or to interpret it correctly. It is more illuminating in this connection to think not of gross violations exemplified by bribery and corruption, or the abuse of the legal system to punish political enemies, but rather of the subtle distortions of prejudice and bias as these effectively discriminate against certain groups in the judicial process.”); Dicey, supra note 26, at 203 (rights allowed by law are nominal unless their “exercise is secured”). 48  Montesquieu, supra note 42, at Book XI, ch. 6; Dicey, supra note 26, at 198; Belton, supra note 9, at 3; Tamahana, supra note 7, at 33; Rawls, supra note 1, at 237 (“The rule of law also implies the precept that similar cases be treated similarly”). Hayek, supra note 34, at 76–7, again pushes this indicator to the limit, arguing that any government provided service cannot be impartial. 49  Montesquieu, supra note 42, at Book XI, ch. 6; Tamahana, supra note 7, at 35; Rawls, supra note 1, at 239 (“While there are variations in these procedures, the rule of law requires some form of due process: that is, a process reasonably designed to ascertain the truth . . . as to whether a violation has taken place and under what circumstances. For example, judges must be independent and impartial, and no man may judge his own case”); Belton, supra note 9, at 17. But see Fuller, supra note 38, at 81 (arguing that a judiciary can detract from the rule of law if laws are not enforced as written). 50  Tamahana, supra note 7, at 9 (Aristotelian rule of law requires “reason unaffected by desire”).

BARFIELD PRINT.indd 73

13/11/2018 11:52

74  Research handbook on the law of virtual and augmented reality 3.4  Liberty and the Rule of Law Missing from the foregoing ten traditional indicators of the rule of law are several values51 that many have come to associate with the rule of law. The following are five concepts that this chapter does not consider part of the rule of law:52 (1)  Democracy: Dictators can—though they often do not—live by the rule of law;53 the Magna Carta shows that a nonelected sovereign can obey the law.54 Further, Madison’s “tyranny of the majority” implies that democracy does not equal rule of law.55 As de Tocqueville pointed out:   A majority taken collectively is only an individual, whose opinions, and frequently whose interests, are opposed to those of another individual, who is styled a minority. If it be admitted that a man possessing absolute power may misuse that power by wronging his adversaries, why should not a majority be liable to the same reproach?56

Finally, the definition of democracy itself is slippery. Federalist voting concepts such as the US Senate and the winner-take-all Electoral College might seem highly undemocratic to citizens of a parliamentary state. Parliamentary systems are undemocratic to those who desire direct democracy. If democracy is ambiguous, then laws emanating from a democracy might be considered arbitrary, but no one takes the view that the US is lawless for this reason.57

  See generally, Greg Lastowka, Virtual Justice: The New Laws of Online Worlds (2011).  Tamahana, supra note 7, at 91, describes this as the difference between formal rule of law (as described in this chapter) and substantive rule of law. The indicators described above align most closely with a “formal legality” regime. 53   Lon Fuller, Positivism and Fidelity to Law—A Reply to Professor Hart, 71 Harvard L. Rev. 630, 660 (1958) (problem with Nazi rule was not the dictatorship, but instead that the dictatorship had no fidelity to law); Tamahana, supra note 7, at 10 (Aristotle and Plato believed that the “best government was by the best man, not rule by law, for law does not speak to all situations and cannot contemplate all eventualities in advance”); see also Tamahana, id. at 37. But see Castronova, supra note 4, at 206–7 (arguing that good government requires limited power: “The premise here tends to be that any individual will exploit any power to her own ends, to the maximum feasible extent, unless constrained in some way by a countervailing incentive”). 54  Tamahana, supra note 7, at 25–6. James Grimmelmann, Virtual Worlds as Comparative Law, 47 N.Y.L. Sch. L. Rev. 147, 175–6 (2004), hereafter Comparative, argues that similar limitations on dictatorial power might be possible in virtual worlds. 55   James Madison, The Federalist No 10 (1787) (“Complaints are everywhere heard . . . that measures are too often decided, not according to the rules of justice and the rights of the minor party, but by the superior force of an interested and overbearing majority”). 56   Alexis de Tocqueville, I Democracy in America 269 (Knopf 1945); see also Tamahana, supra note 7, at 37 (discussing arguments that a judiciary is needed to enforce the rule of law when democracy expands the laws). 57   See, e.g., de Tocqueville, supra note 56, at 272 (“If, on the other hand, a legislative power could be so constituted as to represent the majority without necessarily being the slave of its passions, an executive so as to retain a proper share of authority, and a judiciary so as to remain independent of the other two powers, a government would be formed which would still be democratic while incurring scarcely any risk of tyranny”); see also id. at 271 (American legislature elected 51

52

BARFIELD PRINT.indd 74

13/11/2018 11:52

Virtual rule of law  75 (2)  Freedom: Rules that violate personal autonomy might still satisfy the rule of law.58 Of the rights guaranteed by the US Bill of Rights,59 most are about personal freedom,60 not about whether the sovereign preannounces and follows rules.61 A country could make, for example, a law allowing search warrants to issue upon any application by law enforcement.62 So long as all knew the law and the government undertook no search without an application, the rule of law would be satisfied despite the abandonment of probable cause. (3)  Legitimacy: That the lawmaker has the consent of the governed is helpful for enforcement and political stability,63 but the rule of law might persist without legitimacy.64 Further, the rule of law does not necessarily follow from the legitimacy of lawmakers. As Hayek notes: “The fact that someone has the full legal authority to act in the way he does gives no answer to the question whether the law gives him power to act arbitrarily or whether the law prescribes unequivocally how he has to act.”65 (4)  Noncoerciveness: Even if everyone follows the law, the sovereign could—and, some would argue, should—still have coercive machinery.66

by majority, executive elected by majority, even judges elected by the majority in some states); Tamahana, supra note 7, at 34 (liberals argue that democracy leads to freedom). 58  Tamahana, supra note 7, at 37 (“A regime with oppressive laws can satisfy legal liberty [rule of law] by meticulously complying with those laws”); Palombella, supra note 27, at 36 (“It follows that a perspective of neutrality accepts that it is possible for rights and human dignity to be infringed, even when those requirements of the rule of law are satisfied”) (emphasis in original). 59   U.S. Const. Amends. I–X. 60  Tamahana, supra note 7, at 35 (describing role of personal autonomy in liberal view of freedom). 61   But see Hayek, supra note 34, at 82 (“Man is free if he needs obey no person but solely the laws”), quoting Immanuel Kant, Metaphysical Elements of Justice; Montesquieu, supra note 42, at Book XI, Ch. 3 (“Liberty is a right of doing whatever the laws permit”); Tamahana, supra note 7, at 34 (liberal argument is that if the government is constrained by preexisting rules, then freedom will ensue). 62  Tamahana, supra note 7, at 35, notes that the definition of minimal personal autonomy is imprecise. Thus, the proposed search rule might also satisfy a “freedom” requirement. This imprecision is one reason why the rule of law ideal should be separated from the freedom ideal. 63   Friedrich Kratochwil, Has the “Rule of Law” become a “Rule of Lawyers”? Relocating the Rule of Law 177–8 (Gianluigi Palombella & Neil Walker, eds 2009) (rule of law cannot be separated from those who create and implement it). 64   Compare Castronova, supra note 4, at 157 (arguing that limitations of rights in virtual worlds as potentially benefiting the community’s interests “would be more persuasive if the community’s interests had been validated and expressed through some unbiased consensus-building process”) and 208 (“What is interesting about [provider agreements] is that while they do solicit the consent of the governed . . . they offer no due process of enforcement or amendment”). 65  Hayek, supra note 34, at 82. 66  Rawls, supra note 1, at 240 (arguing that the sovereign must be able to enforce rules even if they are never violated: “[T]he existence of effective penal machinery serves as men’s security to one another”), citing Thomas Hobbes, Leviathan, ch. 17 (Liberal Arts Press 1958); Lawrence Lessig, Code and Other Laws of Cyberspace 189 (1999) (arguing that architecture for enforcement is important for any regulation to be effective); Belton, supra note 9, at 17 (enforcement is required for rule of law); Goldsmith & Wu, supra note 12, at 135 (coercion supplements community norms and cooperation); Lastowka & Hunter, Laws, supra note 6, at 55–9 (describing how wizards in

BARFIELD PRINT.indd 75

13/11/2018 11:52

76  Research handbook on the law of virtual and augmented reality (5)  Happiness: Laws need not improve citizen welfare.67 Every country that purports to follow the rule of law has a variety of laws that might hinder the welfare of some part of the population.68 These five points fly in the face of a rule of law interpretation that incorporates a requirement of liberty.69 This is a modern and decidedly Western view of the rule of law.70 Many countries have varying levels of civil liberties, yet still govern by preannounced and impartially applied rules.71 This is not to say that the liberal rule of law is normatively bad; it is certainly morally good in most cases, including in virtual worlds.72 Take, for example, the termination of Peter Ludlow’s account because of his blog “newspaper” in The Sims Online. The newspaper reported the abuse of community norms, such as new users cheated out of their cyberproperty and purportedly underage users running cyberbrothels.73 Such LambdaMOO needed to retain power even if the users ruled the virtual world by democracy); Orin S. Kerr, Enforcing Law Online, 74 U. Chi. L. Rev. 745, 751 (2007); cf. Tamahana, supra note 7, at 47. 67  Fuller, supra note 38, at 153 (rule of law is “indifferent toward the substantive aims of law”); Palombella, supra note 27, at 35 (“The rule of law is indeed to be distinguished from the rule of the ‘good’ law.”). See, e.g., James Madison, The Federalist No. 39 (1787) (arguing that elections and limited tenure are sufficient to control the legislature). Madison’s argument is morally neutral; it says nothing about what the legislature should do. William H. Riker, Liberalism Against Populism 9 (1982). But see Tamahana, supra note 7, at 9 (Aristotle and Plato believed that laws should further the good of the community). However, because Aristotle and Plato did not believe in democratic rule, the only way to constrain leaders was to constrain the moral goodness of law. 68   Goldsmith & Wu, supra note 12, at 141 (“There are systematic reasons why elected governments sometimes fail to do what is best for their citizens”); Castronova, supra note 4, at 206 (describing balance between restricted governmental power and social welfare). But see Rousseau, Social Contract, Book I, ch. 8 (majority rule will never oppress because the majority dictates society’s needs). Rousseau’s view implies that democracy is sufficient to be morally good. Riker, supra note 67, at 11. 69   Compare Tamahana, supra note 7, at 5 (“[T]he rule of law ideal initially developed in nonliberal societies”) with Belton, supra note 9, at 3 (human rights are part of the rule of law) and Dicey, supra note 26, at 191 (rule of law includes personal freedoms). 70  Tamahana, supra note 7, at 33 (“[W]hile liberal systems cannot exist without the rule of law. . .the rule of law can exist outside of liberal systems”); see also Tamahana, supra note 7, at 81 (describing modern “responsive law” arguments regarding progressive US Supreme Court decisions such as desegregation) and 99 (describing democracy as a Western ideal). 71  Tamahana, supra note 7, at 5 (“However, many of the non-Western societies that wish to implement the rule of law have no desire to become liberal, and many Western societies with the rule of law are committed to the social welfare state [that Hayek criticizes as arbitrary]”); Palombella, supra note 27, at 35 (arguing that formalistic rule of law might apply to welfare state despite Hayek’s belief to the contrary); Grimmelmann, Feudalism, supra note 10 (describing order of law in virtual “feudal” worlds). But see Lastowka & Hunter, Laws, supra note 6, at 32–3 (virtual worlds display trappings of modern Western property law, even if supposedly set in medieval times); Bradley & Froomkin, supra note 14, at 134 (ability to test alternative legal schemes is limited due to player “attitudinal constraints”). 72   See generally Koster, supra note 13, at 56 (discussing rights of avatars); but see Julian Dibbell, Owned! The State of Play: Law, Games, and Virtual Worlds 140 (Jack M. Balkin and Beth Simone Noveck, eds 2006) (describing virtual sweatshops used to generate virtual property for sale). 73  Castronova, supra note 4, at 305 note 5; Balkin, supra note 10, at 2075–6.

BARFIELD PRINT.indd 76

13/11/2018 11:52

Virtual rule of law  77 reporting provided important information to users, and the authoritarian world provider did not want such information disseminated because it discouraged users from joining. From a rule of law perspective, a rule banning all newspapers might satisfy the rule of law despite the moral good associated with the press. However, to the extent that banning this particular user was arbitrary and not in accordance with existing rules,74 the action would violate the rule of law. The newspaper example might imply that liberal rule of law should be a goal for virtual worlds.75 Nonetheless, applying liberal democratic principles to virtual worlds is infeasible for four reasons. First, gameplay in the virtual world may not lend itself to freedoms.76 For example, many worlds are dedicated to player versus player combat, and concepts such as “citizen happiness” depend on who wins the battle. Also, the computer sees all—and indeed must do so in order for the software to work properly. As a result, guarantees that the provider (as sovereign) will not “search” a user’s virtual possessions are pointless.77 Second, requirements of democracy and legitimacy cannot apply because virtual worlds are democratic dictatorships.78 Providers are in power by fiat,79 because they invest in the hardware and software to build the world.80 They gain sovereign rights only by attracting subscribers who vote with their feet and their wallets.81 While providers may listen to input 74  Balkin, supra note 10, at 2075–6 (“Ludlow argued that this was a pretextual enforcement of a technical violation of the TOS not regularly applied against other players”). 75  Balkin, supra note 10, at 2075–6 (discussing Ludlow and speech values); Lastowka & Hunter, Laws, supra note 6, at 51–2 (many new users want traditional liberal rights in virtual worlds); Koster, supra note 13, at 66 (players expect a certain level of treatment from providers, whether or not such “rights” are expressed as such or written in a document). 76   Lastowka & Hunter, Laws, supra note 6, at 59 (“[W]e might well conclude that virtual democracy and avatar rights are not ideals worth pursuing. The reasons for this are many, but certainly include arguments that the worlds are built and maintained out of the funds of a private entity, or that democracy and these worlds are not good bedfellows”). 77   Indeed, such protections would likely not even apply to real governments to the extent that a user’s virtual possessions are visible to other avatars in the world. See generally Joshua Fairfield, Escape Into the Panopticon: Virtual Worlds and the Surveillance Society, 118 Yale L.J. Pocket Part 131 (2009) (discussing privacy and government searches in virtual worlds), available at www. yalelawjournal.org/forum/escape-into-the-panopticon-virtual-worlds-and-the-surveillance-society. 78  Castronova, supra note 4, at 152 (describing providers as having dictatorial powers). 79   Lastowka & Hunter, Laws, supra note 6, at 55. 80   Lastowka & Hunter, Laws, supra note 6, at 59 (“The standard argument against avatar rights, therefore, is that wizards, by virtue of their private (and corporate) ownership of the computer equipment and substantial investments in creating the virtual world, should have a right to do exactly as they please”). 81   Compare Castronova, supra note 4, at 207 (“[O]ne does not find much democracy at all in synthetic worlds . . . The typical governance model in synthetic worlds consists of isolated moments of oppressive tyranny embedded in a widespread anarchy . . . There is a tyrant in place from the beginning, but an extraordinarily inactive one.”) with id at 261 (“[C]ompetitive pressures force companies to keep as many people as possible just happy enough to stay”). See also Grimmelmann, supra note 54, at 176–8 (describing that right of exit puts pressure on providers to consider user desires); Lastowka & Hunter, Laws, supra note 6, at 59; David G. Post, Anarchy, State and the Internet: An Essay on Law-Making in Cyberspace, 1995 J. Online L. art 3, ¶42 (“A kind of competition between individual networks to design and implement rule-sets compatible with the preferences of individual internetwork users will thus materialize in a new . . . market for rules”).

BARFIELD PRINT.indd 77

13/11/2018 11:52

78  Research handbook on the law of virtual and augmented reality provided by their users,82 the laws providers impose are not, with few exceptions,83 put in place by majority vote or any other form of democratic representation.84 Further, users cannot revolt to depose the provider’s rule if they disagree with how they are regulated.85 Thus, subscribers choose to have their avatars be subject to dictatorial laws.86 Third, freedom is further limited because virtual lives and property are not portable.87 Once an avatar lives in a world, it cannot leave except by death.88 The democratic choice of a particular dictatorship is locked in, and the more time and effort one spends in a world, the more it hurts to vote for another world.89 While some have suggested mandatory portability of virtual possessions as a way to improve avatar rights,90 at present these complications make analysis of modern liberal ideals virtually impossible in this context.91 Fourth, to the extent that the rule of law fosters investment by setting expectations, liberal ideals are less important. So long as expectations are fixed, users can allocate time, attention, and investments appropriately. Of course, liberal governance may maximize total wealth as compared to nonliberal governance, but given the other limitations discussed previously, fixed expectations are the most that can be achieved.

 Dibbell, supra note 72, at 142–3; Balkin, supra note 10, at 2051.   LambdaMOO, for example, tried a user democracy for a period of time. Castronova, supra note 4, at 217; Lastowka & Hunter, Laws, supra note 6, at 55–9. See also Grimmelmann, Comparative, supra note 54, at 176 n.115 (describing user referenda in “A Tale in the Desert”); Timothy Burke, Play of State: Sovereignty and Governance in MMOGs, at 12 (Aug. 2004) ­(providers are the sovereign), available at www.swarthmore.edu/SocSci/tburke1/The%20MMOG%20State. pdf and https://perma.cc/YS5Z-GKPY. 84   Bradley & Froomkin, supra note 14, at 143–6. 85  Grimmelmann, Comparative, supra note 54, at 176 (“[A]t least since Locke, the legitimacy of republican government has been intertwined with the right of revolution. If the government refuses to obey the results of an election, it must expect the citizenry to rise up and depose. But there is no way to depose the designers of a game”). 86  Lessig, supra note 66, at 201 (the amount of regulatory power of a site coincides with its market power). 87  Lessig, supra note 66, at 202 (because switching costs are high, users cannot effectively vote with their feet). More precisely, they can only vote once without having to reinvest in a new world. 88   An avatar can, however, be transferred to another user. 89  Castronova, supra note 4, at 152 and 261 (“If you disagree, and want to abandon the fruits of thousands of hours of work and effort, as well as all of your friendships, click ‘I Disagree’ [to the provider agreement] and go spend some time as a lonely hobo in some other world”); Lastowka & Hunter, Laws, supra note 6, at 61–2; Joshua A.T. Fairfield, Anti-Social Contracts: The Contractual Governance of Online Communities, 53 McGill L.J. 427, 470 (2008), hereafter Anti-Social (providers design worlds to create high switching costs); Balkin, supra note 10, at 2051. But see Grimmelmann, Comparative, supra note 54, at 178–81 (arguing that a user’s ability to sell an avatar limits the cost of leaving a world, thus making it easier to leave and exerting less pressure on providers to change their rules). Interestingly, then, providers appear to have little pressure, regardless of whether or not it is costly to exit. 90  Fairfield, Anti-Social, supra note 89, at 471. 91  Grimmelmann, Comparative, supra note 54, at 181–3 (providers acting as non-intervening dictators may be preferable than attempting to create a virtual democracy); Koster, supra note 13, at 63 (users have almost no power to seek rights, and providers do not want to surrender the control necessary to grant rights). 82 83

BARFIELD PRINT.indd 78

13/11/2018 11:52

Virtual rule of law  79

4.  DEFINING “LAW” IN RULE OF LAW The two primary components of the rule of law—sovereigns and rules—take on new meaning in the context of virtual worlds. For example, just who is the sovereign? Is it the virtual-world provider or the real-world government? Both? What are the rules? Are they set by contract? Through user democracy? By real-world legislatures? Answering these questions must precede consideration of the rule of law.92 4.1  Sources of Law and Sovereignty As Professor Lessig and others have pointed out, four different forces regulate behavior in virtual worlds:93 (1)  Market: The demand for services offered by and uses for each virtual world will affect what users can do in that world. Contracts thus define many of the rules that provide the basic framework for conduct in and relating to the virtual world.94 If consumers want a particular set of rules, they will flock to worlds that provide the preferred combination and away from worlds that do not.95 If providers attempt to change the rules, users may leave, or just threaten to do so.96 For example, with respect to the concert depicted in Figure 3.1, the Second Life terms of service do not forbid the playing of live music. However, if Linden Lab ever modified the agreement to ban live music, users might leave for another world. (2)  Architecture (Code): The ability for virtual world providers to control what avatars can and cannot do by fiat is a perfectly enforceable regulation.97 As Professor Lessig notes, “Code is law.”98 Code provides private parties with a previously unavailable coercive force.99 For example, Linden Lab may disable the software that allows for transmission of live music to the Second Life servers.

92   Lastowka & Hunter, Laws, supra note 6, at 9–10 (“Since people expect places to be governed by some law, we should attempt to fashion some decent answer to the question of what laws might (or should) apply to virtual worlds”). 93  Lessig, supra note 66, at 88. 94  Rawls, supra note 1, at 236 (“[T]he law defines the basic structure within which the pursuit of all other activities takes place”); Koster, supra note 13, at 66 (terms of use can describe limitations on both user and provider conduct); Dibbell, supra note 72, at 143–4; Kerr, Criminal, supra note 15, at 422; Margaret Jane Radin, Regulation by Contract, Regulation by Machine, 160 J. Inst. & Theoretical Econ. 1, 6–7 (2004) (under an effective contract, “for a large subset of the social order . . . the law of the state has been superseded by the promulgated contractual regime, the ‘law’ of the firm”). 95  Fairfield, Anti-Social, supra note 89, at 471–3; Dibbell, supra note 72, at 144. 96  Grimmelmann, Comparative, supra note 54, at 176–8 (discussing several market factors that might affect whether users leave games and whether providers will consider user requests); Dibbell, supra note 72, at 142–3. 97   Jonathan Zittrain, The Future of the Internet and How to Stop It, 168–9 and 172 (2008). 98  Lessig, supra note 66, at 6 (emphasis in original); but see Castronova, supra note 4, at 205 (code may be law, but more than code creates a virtual “state”). 99   See, e.g., Rawls, supra note 1, at 236 (“The constitutional agencies that [a legal system]

BARFIELD PRINT.indd 79

13/11/2018 11:52

80  Research handbook on the law of virtual and augmented reality (3)  Law: These are rules announced by and enforced by some sovereign with coercive power. Unlike code, enforcement may not be perfect. For example, copyright laws may forbid the public performance of someone else’s musical work in Second Life. (4)  Norms: These are rules, written and unwritten, followed by avatars in a particular community.100 Other users may or may not enforce the rules, and the sovereign may even enforce them.101 For example, gatherings with live music are unlikely events in worlds featuring player versus player combat. Each of these sources of regulation can theoretically come from two sovereigns—the real and the virtual—both of which can effectively regulate behavior.102 Even though community is not really a sovereign, community-based enforcement of norms might constitute an additional source of regulation. Of course, some types of regulation go together more naturally with a particular sovereign. It is unlikely that providers will pass any binding legislation, just as it is unlikely that the United States will write software code. However, overlap is not as farfetched as it may sound. A government could set, for example, default rules about required or banned contract terms and software functionality. Furthermore, a virtual world might enforce the laws of a government (for example, exacting in-world punishment for defamation). Thus, there are potentially 12 sources of enforced rules in any given world: three sovereigns (including the community) multiplied by four sources. These enforced rules might have an effect on both real users and virtual avatars, bringing the total number of potential applications to 24.

defines generally have the exclusive legal right to at least the more extreme forms of coercion. The kinds of duress that private associations can employ are strictly limited”). But see Orin S. Kerr, The Problem of Perspective in Internet Law, 91 Geo. L.J. 357, 372 (2003) (“Saying that the power of code is akin to the power of law is simply too loose a use of the word ‘law’ to be helpful. If code is law to an Internet user, then a sports referee’s calls are law to an athlete, and Steven Spielberg’s decisions about how to shoot a movie are law to a movie viewer”) and Tateru Nino, Code is Law (May 31, 2007) (“However, it’s not that code is a limiting factor for the most part. What it is, instead, is an enabling factor”) (emphasis in original), available at https://web.archive. org/web/20100808020653/http://www.secondlifeinsider.com/2007/05/31/code-is-law/ and https:// perma.cc/B8W9-GVUQ. 100  Fairfield, Anti-Social, supra note 89, at 459–61; Kerr, Criminal, supra note 16, at 422. 101   David R. Johnson, The New Visual Literacy: How the Screen Affects the Law, The State of Play: Law, Games, and Virtual Worlds 246 (Jack M. Balkin and Beth Simone Noveck, ed. 2006) (“We may soon take it for granted that the act of visiting a particular online space corresponds to submission to the special rules that apply to action in that context”); Zittrain, supra note 97, at 168 (“[S]ocial problems can be met first with social solutions—aided by powerful technical tools—rather than by resorting to law”). 102  Lessig, supra note 66, at 189 (“We can have an idea of sovereign power­—the right of the sovereign to regulate or control behavior—but our idea is only meaningful when we place it within a particular regulatory context, or within particular architectures of control”); Radin, supra note 94, at 7 (“Sovereignty has been abrogated in favor of whatever firm has promulgated the regime”); Grimmelmann, Feudalism, supra note 10 (the provider both grants property rights and sets the law relating to those rights); Burke, supra note 83, at 2.

BARFIELD PRINT.indd 80

13/11/2018 11:52

Virtual rule of law  81 Table 3.1  The dimensions of real / virtual law and real / virtual sovereignty

Real law Virtual law

Real sovereign

Virtual sovereign

Fraud Breach of contract

Penalty for defamation Penalty for rule violation

4.2  Simplifying the Analysis Despite apparent complexity, these 24 combinations can be simplified in two steps. First, they can be categorized in two dimensions. Second, the two dimensions can be collapsed into one test. 4.2.1  Two dimensions The combination of sovereigns and sources might be expressed as two dimensions: Real/ Virtual Law and Real/Virtual Sovereignty. Table 3.1 illustrates how these two dimensions work to create four possible applications of the rule of law. Each sovereign has the potential ability to enforce each type of law: The case of a Japanese Maple Story virtual-world user illustrates these dimensions.103 Distraught about the “divorce” of her and her virtual spouse’s avatars, she decided to exact revenge by logging into the virtual world using her virtual spouse’s username and password and deleting the spouse’s account. The deletion of the avatar potentially falls under all four combinations.104 First, the deleted avatar ceased to exist in-game, and thus a virtual murder was committed in the virtual world, even if no human was injured. Second, assuming that such virtual murder violates the provider agreement, a real sovereign might enforce virtual law in a breach of contract suit. Third, the unauthorized account access and destruction of (presumably) valuable electronic assets violates a real law enforced by the real sovereign. Fourth, to the extent that the provider agreement bars unauthorized account access, then the virtual sovereign would be enforcing real law. There are subtle differences between the second and fourth combinations. In the second (virtual murder enforced in real courts), the “law” that is broken is in-game—the virtual murder. In the fourth (real tampering enforced by the virtual sovereign), the “law” that is broken is outside the game; further, even though the provider might need to sue for breach of contract in a real court, in this abstraction the real-world law is still being imposed by the virtual sovereign rather than by a real-world government. 4.2.2  Collapsing the dimensions It is tempting to consider each of the 24 or even four possible permutations with respect to the rule of law, but, in the end, such complexity is neither warranted nor fruitful. 103   Woman Arrested for “Killing” Her Virtual Husband, NY Times (Oct. 24, 2008), available at www.nytimes.com/2008/10/24/world/asia/24iht-virtual.1.17223730.html. 104   Virtual marriage and divorce falls in the virtual sovereign/virtual laws dimension, though they likely fall under community norms.

BARFIELD PRINT.indd 81

13/11/2018 11:52

82  Research handbook on the law of virtual and augmented reality Because the law applies to individual entities, it should be examined from the standpoint of individual entities. For example, when considering the rule of law in America, we consider the effect of state and federal law together, even though laws of one jurisdiction are often enforced in the courts of the other, and even if the laws might conflict. What matters is the cumulative effect of the law on its subjects. Thus, this chapter considers a projection, to continue the geometric metaphor, of multiple dimensions into one question: “What constraints are imposed on real people in their activities relating to virtual worlds?” These constraints are surely overlapping, such that questions of conflict and preemption must be considered to determine which regulations will govern the user.105 A focus on the user is consistent with the effect of the rule of law, or lack thereof, on digital entrepreneurship.106 Humans are a better focus than avatars for three reasons. First, humans, not avatars, make decisions based on rules.107 Even automated decision processes must be initially programmed by humans. As a result, humans direct much of the debate and political deliberation about virtual worlds both inside and outside the virtual world. Inside the world, avatars perform protests, while outside they complain to customer service (and threaten to take their business elsewhere).108 The observation that most of the “democratic” discussion about what users want in their virtual world occurs in the real world reinforces a human focus.109 Second, human wellbeing is more important than virtual wellbeing,110 though perhaps some might disagree.111 As a society, we care about what happens to people, not what happens to their avatars.112 If Linden Lab closed Second Life, millions of avatars would “die.” Their owners might be unhappy, but their wellbeing would suffer little more than 105  Lessig, supra note 66, at 192; Joshua Fairfield, Mixed Reality: How the Laws of Virtual Worlds Govern Everyday Life, 27 Berkeley Tech. L.J. 55 (2012) (discussing how virtual rules affect real-world users). 106   Joshua A.T. Fairfield, The Magic Circle, 11 Vanderbilt J. Ent. & Tech. L. 823, 829–30 (2009), hereafter Magic Circle (“The fundamental question of online regulation is not the balance of power between nation-state sovereigns. Rather, it is the balance between sovereign and citizens”); cf. Castronova, supra note 4, at 151 (“It is frankly impossible to deny that the gold pieces of fantasy worlds are money, just like the money in your pocket. They are sustained by exactly the same social mechanisms and perform exactly the same functions”). 107  Lessig, supra note 66, at 190 (“Whenever anyone is in cyberspace, she is also here, in real space. Whenever one is subject to the norms of a cyberspace community, one is also living within a community in real space. You are always in both places if you are there, and the norms of both places apply”) (emphasis in original). 108  Castronova, supra note 4, at 152–3. 109   Id. 110   Jonathon W. Penney, Understanding the New Virtualist Paradigm, 12(8) J. Internet L. 3,4 (2009) (“Virtual people in virtual communities are real people with real-life concerns”). 111   Lastowka & Hunter, Laws, supra note 6, at 63–5, for example, discuss the projection of the user into an avatar, rather than the projection of the avatar onto the user. Despite differing terminology, the authors still consider the user’s wellbeing as the end goal: “Yet while an avatar’s owner may be perfectly comfortable with killing the avatar when she grows sick of it, she may feel genuine anger when a more powerful avatar decides to use her avatar for target practice.” Id at 63. 112   Cf. Lastowka & Hunter, Laws, supra note 6, at 9 (“One does not study the labor market because work is holy and ethical; one does it because the conditions of work mean a great deal to a large number of ordinary people. By the same reasoning, economists and other social scientists will

BARFIELD PRINT.indd 82

13/11/2018 11:52

Virtual rule of law  83 losing at a video game.113 If, however, such a shutdown meant that real people would lose real money (and productive time) invested in Second Life, then those negative effects are worth considering.114 Third, humans might control several avatars, in the same or multiple virtual worlds.115 The rule of law in one world might counteract lawlessness in another, and the net effect on the user (and the user’s choice to participate in multiple worlds) is the important and interesting point of study. Thus, if code, norms, law, or markets exert control on real people, then this chapter questions whether the rule of law governs those controls.

5.  VIRTUAL RULE OF LAW The foregoing background provides a foundation to consider the rule of law in virtual worlds. As a starting point, the fact that regulation comes from multiple sources should not alone vitiate the rule of law. Given the complexity of virtual worlds, regulation from multiple sovereigns is to be expected.116 Real-world jurisdictions are subject to multiple levels of jurisdiction as well; US residents are subject to federal laws, federal administrative regulations, state laws, state administrative regulations, county ordinances, city ordinances, and local rules imported from other sources, such as building codes. Just as the rule of law might be present in the United States, it might still be present in virtual worlds. As discussed above, the analysis considers the constraints (from whatever source) on users. The results are unexpected. Maligned sources of law such as agreements and code hold the most theoretical hope for the rule of law, despite often currently falling down in practice. Lauded sources of law, such as community norms and importation of real-world law, appear to fail not only practically but also theoretically.

become more interested in Norrath and similar virtual worlds as they realize that such places have begun to mean a great deal to large numbers of ordinary people”). 113  Castronova, supra note 4, at 154–5, rightly points out that “fun” is part of the economic utility (compensation, jollies, or utils) that one gets from playing in the virtual world. Ironically, though, the less “fun” a character is, the more valuable it is, because there are fewer of them. Id. There is a limit to this argument—people will choose not to play games that are no fun. The argument discussed here is slightly different—this chapter places less value on “fun” where the world is used for entertainment only and has no business component. 114  Castronova, supra note 4, at 152 (“Should something happen to the conditions of being a wizard—say, a formerly powerful spell gets weakened—all those who are settled into wizarding as an occupation experience a genuine loss of well-being”); Balkin, supra note 10, at 2071 (“If virtual items have real-world equivalent values, though, the game designer may be destroying a considerable amount of value by turning off the game, and the more value that is destroyed, the less likely the law will stand for it”). 115   Tateru Nino, Can an Avatar Sign a Contract? (May 4, 2009) (pointing out the “shocking” fact that humans are not, in fact, the same as their avatars), available at http://dwellonit.taterunino. net/2009/05/04/can-an-avatar-sign-a-contract/ and https://perma.cc/M7HF-6FFV. 116  Zittrain, supra note 97, at 168 (“One sovereign cannot reach every potentially responsible entity on a global network, and while commercial forces can respond well to legal incentives, the amateur technology producers that are so important to a generative system are less likely to shape their behavior to conform to subtle legal standards”).

BARFIELD PRINT.indd 83

13/11/2018 11:52

84  Research handbook on the law of virtual and augmented reality 5.1  Market Regulation via the User Agreement For better or worse, provider agreements define almost all of the rules that users must obey. Many object to the notion that such agreements—which are undoubtedly contracts of adhesion—should govern behavior within the worlds.117 Despite objections, contractual sovereignty is the fact of the matter.118 If an agreement states that users do not own the “virtual property” that their avatars “possess,” then users do not own such virtual possessions, no matter how seemingly unfair, inefficient, or unfortunate the announced rule might seem.119 In theory, contractual regulations may exhibit the ten indicators discussed in section 2. Contractual rules can be general, announced in advance, nonarbitrary, neutral, and so forth.120 Further, contracts fulfill the simpler definition that rule of law only requires the sovereign to behave in accordance with law. Real courts of law will almost certainly hold providers to the explicit terms of their agreements.121 While it is the case that many users agree to such agreements without reading them,122 this does not change the result;123 most people have not read the US Code, state codes, and court cases that will affect them.

117   See generally Fairfield, Anti-Social, supra note 89. See also Balkin, supra note 10, at 2071–2 (predicting that courts will not enforce all user agreements); but see Dibbell, supra note 72, at 143 (To the extent that users have input with providers, “[t]he EULA starts to look less like a contract of adhesion . . . and more like a social contract”). 118  Fairfield, Anti-Social, supra note 89, at 432; Goldsmith & Wu, supra note 12, at 137 (even with informal dispute resolution, contract enforcement is critical to online business); but see Lessig, supra note 66, at 197 (contracts in cyberspace are entered nearly costlessly, which might warrant real-world limitations on the rights that might be signed away by contract); Stoup, supra note 13, at 338–9 (provider implemented rules are more effective than community policed norms); but see Christopher Cifrino, Virtual Property, Virtual Rights: Why Contract Law, Not Property Law, Must Be the Governing Paradigm in the Law of Virtual Worlds, 55 B.C. L. Rev. 235 (2014) (arguing that contracts are better than property for regulation of virtual worlds). 119  Castronova, supra note 4, at 151 (“Users are a community of interests who are affected by the decisions of a coding authority”); Grimmelmann, Comparative, supra note 54, at 150 (property rights are possessory—one has exclusive control as to other game players only so long as one holds onto the property); Fairfield, Anti-Social, supra note 89, at 441 (“The question of whether these private-law contracts can deliver the public goods they promise is the centrepiece of this chapter”) and 443–5 (describing why contracts cannot create all background law for virtual worlds); Kerr, Criminal, supra note 16, at 428 (breach of agreements should not be criminalized, because terms of such agreements are “arbitrary, and can reflect the whims and biases of whoever sets them”). But see Balkin, supra note 10, at 2065 (restricting ownership of virtual property can enhance game play). 120  Dibbell, supra note 72, at 143 (describing importance of provider agreement in defining acceptable behavior and problems of uncertainty caused by ignoring the agreement). 121   But see Dibbell, supra note 72, at 144 (providers might overstep their own agreements in suits against users); Michael Meehan, Virtual Property: Protecting Bits in Context, 13 Rich. J.L. & Tech. 7, 26 (2006) (provider agreements allow for arbitrary revocation of property). 122  Fairfield, Anti-Social, supra note 89, at 468–9. 123  Fairfield, Magic Circle, supra note 106 (users consent to package of game rules even without reading them, including agreements and community norms).

BARFIELD PRINT.indd 84

13/11/2018 11:52

Virtual rule of law  85 5.1.1  Contract modification Of course, in practice the rule of law may not apply. For example, the provider can modify most agreements at any time,124 but even this does not negate the rule of law.125 The question is whether such modifications are frequent, arbitrary, ex post facto, or targeted at specific users or groups.126 Anecdotal evidence suggests that modifications do not suffer from such problems much of the time. Indeed, providers seem to respond to the desire of their users (and the corresponding market pressure) to make no changes that will upset a large portion of the population, but there are also exceptions.127 Four such modifications illustrate the point. First, Second Life decided to start taxing virtual property.128 A tax made great business sense; as virtual property proliferated, revenues on such property would grow. Of course, users did not want to pay additional money, and their avatars staged virtual protests to make the point. Second Life eventually rescinded the change. Second, in 2007, Second Life outlawed gambling.129 On its face, this would appear to  violate rule of law principles because those who joined Second Life to start a ­gambling business were now out of business,130 but closer examination reveals the opposite. The ban was not a “frequent” change; it was not as if Second Life banned entire lines of business and then reinstated them on a regular basis. The contract amendment was not arbitrary; gambling is illegal in many jurisdictions. The rule had no ex post facto effect; it imposed no penalty for past gambling. Additionally, the change was not targeted; it was a general rule with general application. Based on these changes, the contractual law against gambling was no different from any legislative ban on real-world gambling, in accordance with the rule of law. Third, in March of 2009 Blizzard announced new rules regarding the use of software “addons” used in its World of Warcraft world. These rules purport to limit who can use addons and how. Addon developers may not charge for addons nor may they seek donations in-game, they must make the source code visible to all, and Blizzard may disable any addon at any time.131 This rule has a definite ex post facto effect—time, 124   David P. Sheldon, Claiming Ownership, but Getting Owned: Contractual Limitations on Asserting Property Interests in Virtual Goods, 54 UCLA L. Rev. 751, 768 (2007). 125   Users must typically agree to new terms before every game session or forfeit their account, much like credit card agreements. But see Castronova, supra note 4, at 208 (user agreements provide no “due process” of enforcement or amendment). 126   See, e.g., Nicolas Suzor, Order Supported by Law: The Enforcement of Rules in Online Communities, 63 Mercer L. Rev. 523 (2012) (discussing how contracts in virtual worlds should be administered). 127  Castronova, supra note 4, at 152–3; Dibbell, supra note 72, at 143. 128   Wagner James Au, Tax Revolt in Americana! New World News (Sept. 12, 2003), available at http://nwn.blogs.com/nwn/2003/09/tax_revolt_in_a.html and https://perma.cc/LFW4-9BGV. 129   Susan W. Brenner, Fantasy Crime: The Role of Criminal Law in Virtual Worlds, 11 Vand. J. Ent. & Tech. L. 1, 65–6 (2008). 130   Many likely went “underground.” 131   Adam Holisky, New Add-On Policy Makes Selling Add-Ons Against the Rules (Mar. 20, 2009), available at https://web.archive.org/web/20100407034821/http://www.wow.com​ /200​ 9/​03/20/new-add-on-policy-makes-selling-add-ons-against-the-rules/ and https://perma.cc/​7MJ​ D-X5HN.

BARFIELD PRINT.indd 85

13/11/2018 11:52

86  Research handbook on the law of virtual and augmented reality effort, and money used to develop this software is now worthless if the addon cannot be sold as anticipated.132 Further, though apparently general, the rules may target specific users.133 Fourth, modification may be arbitrary and targeted at specific individuals. As discussed previously, The Sims Online banned Peter Ludlow because of a particular newspaper blog he published.134 To the extent that the provider agreement did not allow such a ban, the action was arguably a modification of the agreement as it applied to Ludlow and, as such, it was arbitrary. By extension, every action that a customer service agent takes with respect to a user, if not specifically allowed by the agreement, is a potentially arbitrary “modification” of that agreement.135 Even where a company modifies the agreement in response to customer requests,136 such frequent changes violate one of the rule of law indicators. 5.1.2 Vagueness A bigger problem than modification—indeed, a critical problem in practice—is that provider agreements are notoriously vague.137 These agreements become even vaguer when they incorporate community standards as enforceable promises. For example, the Second Life terms of service state that violation of community norms, including lack of “tolerance” and “respect,” is punishable by account termination.138 In other words, the potential penalty for “intolerance” is death of the avatar. The rule of law problem with this is not the death penalty—if announced beforehand, users can avoid Second Life if they want to avoid penalties. Instead, the problems are: (1) intolerance is ill-defined;139 (2) there is no way that such a contract term can (or will) be enforced neutrally and consistently; and (3) there is no independent factfinder to determine whether behavior is intolerant. Even so, these shortcomings may not negate the rule of law if the provider could announce that the contractual rules are never enforced—users would know what the likely outcome is and plan accordingly.140 Of

132   Mike Schramm, Addon Devs Respond to Blizzard UI Policy Changes (Mar. 23, 2009) (“Already, the creator of the popular QuestHelper has responded on his changelog, saying that the addon is ‘dead’”), available at https://web.archive.org/web/20090606075349/http://www.wow. com:80/2009/03/23/devs-respond-to-addon-changes and https://perma.cc/6LVV-HSWE. 133   Id. (“word is going around that Blizzard released these new policies to put the kibosh on [one particular add-on provider]”). 134  Castronova, supra note 4, at 127. See note 74, supra, for discussion about whether such change complied with the agreement. 135   James Grimmelmann, Virtual Power Politics, The State of Play: Law, Games, and Virtual Worlds 148 (Jack M. Balkin and Beth Simone Noveck, ed. 2006), hereafter Power (describing a variety of targeted responses to user conduct); Fairfield, Anti-Social, supra note 89, at 468-9. 136  Dibbell, supra note 72, at 143 (provider agreement for EverQuest “effectively renegotiated on a daily basis”). 137  Fairfield, Anti-Social, supra note 89, at 436–8 (contracts create confusion); Dibbell, supra note 72, at 144 (there are questions that agreements cannot answer). 138  Stoup, supra note 13, at 319 n.43. 139   Tateru Nino, Policies. . . (“New and ambiguous content policies from Linden Lab left everyone in doubt as to what was and was not allowed in Second Life”), available at http:// dwellonit-comic.taterunino.net/archive/25 and https://perma.cc/43SN-QE68. 140  Tamahana, supra note 7, at 90 (indeterminate legal systems need not be unpredictable).

BARFIELD PRINT.indd 86

13/11/2018 11:52

Virtual rule of law  87 course, other users might attempt to enforce the terms, which would dilute any provider announcement.141 Vagueness detracts even further from the rule of law when the penalty shifts from unenforced avatar death to enforced real-world imprisonment. Criminal liability for violating the Computer Fraud and Abuse Act by breaching a user agreement is subject to continuing battle in the courts.142 Thus, an avatar’s intolerance and lack of respect—conduct that could never be criminalized in the real world—might lead to jail time if expressed in breach of a provider’s user agreement. This outcome—even the uncertainty generated by its mere possibility—is contrary to any number of the rule of law indicators discussed above.143 5.1.3  Lack of enforceability and enforcement These vagueness concerns highlight a more general practical problem: provider agreements are not enforced or enforceable consistently or equally, such that users cannot rely on agreement-based law, even if announced beforehand.144 One explanation for uneven enforcement is that providers do not include a wide variety of sanctions in their user agreements in case of breach.145 This means that any wrongdoing receives the maximum penalty: avatar death. Providers are loath to terminate user accounts, especially paying accounts, for every little transgression that might not comply with the terms of the provider agreement.146 Providers have even less incentive to enforce contractual terms in real-world courts, for cost and publicity reasons. Furthermore, it is quite difficult for users to enforce provider agreement breaches against other users, because users do not enter into contracts with each other—only with the provider. Some argue that such enforcement is impossible;147 at the very least, it is 141   This may be a reason why providers disfavor the third party beneficiary doctrine. See generally Michael Risch, Virtual Third Parties, 25 Santa Clara Computer & High Tech. L.J. 415 (2009). 142   United States v. Nosal 844 F.3d.1024 (2016) (access without permission can constitute violation); United States v. Lowson, No. CRIM. 10-114 KSH, 2010 WL 9552416, at *6 (D.N.J. Oct. 12, 2010) (upholding indictment for bypassing Ticketmaster’s Captcha). See generally, Orin Kerr, Cybercrime’s Scope: Interpreting “Access” and “Authorization” in Computer Misuse Statutes, 78 N.Y.U. L. Rev. 1596, 1643 (2003) (discussing concerns about imprisoning individuals for breaching website terms of service). 143   Courts seem to recognize this. United States v. Nosal, 676 F.3d 854, 863 (9th Cir. 2012) (“We need not decide today whether Congress could base criminal liability on violations of a company or website’s computer use restrictions. Instead, we hold that the phrase ‘exceeds authorized access’ in the CFAA does not extend to violations of use restrictions”). 144  Castronova, supra note 4, at 208 (user agreements provide no “due process” of enforcement or amendment); Radin, supra note 94, at 4 (consistent enforcement of contracts only a hypothetical); Nicolas Suzor, The Role of the Rule of Law in Virtual Communities, 25 Berkeley Tech. L.J. 1817 (2010). 145   These sanctions could be written in the form of game rules, which might technically be considered provider-enforced liquidated damages. 146  Fairfield, Anti-Social, supra note 89, at 472–3 (strict rules are needed to keep the riff-raff out even though a majority do not want such rules, leading to limited targeted enforcement); Stoup, supra note 13, at 331–2. Indeed, lack of enforcement may be a deliberate choice to maximize profits. Grimmelmann, Comparative, supra note 54, at 181 (“If driving away five $10 per month players causes a $100 per month player to double her spending, the game administration will happily make such a trade”). 147  Fairfield, Anti-Social, supra note 89, at 436–8.

BARFIELD PRINT.indd 87

13/11/2018 11:52

88  Research handbook on the law of virtual and augmented reality difficult,148 and it is inefficient.149 Further, even if a provider agreement is enforceable by users, equitable relief may not be available.150 As a result, if the virtual-world provider does not enforce the contract, at best the “laws” might be enforceable, and at worst they will remain unenforced.151 Contractual regulations with limited and uncertain enforceability fail several enforcement-related indicators of the rule of law. Further, lack of enforcement means that provider agreements are almost never tested in court. In turn, lack of testing provides no pressure to write clear contractual rules with varying penalty levels for varying misconduct. If there were a variety of penalties short of avatar death, providers might be more willing to define and enforce contractual rules in-game. 5.2  Community Norms as Law Lack of contractual enforcement has led to community enforcement of norms.152 Community enforcement can be anything from a decrease in status, to gossip, to harassment, to attacking the offending avatars.153 Some have applauded community efforts as a successful response to what would otherwise be lawlessness.154 For example, some suggest that laws in virtual worlds might mimic rules relating to sports. Professor Fairfield uses hockey as an example; the rulebook disallows fighting, but community norms supplement background tort law by allowing some amount of fighting without being legally actionable battery.155 From a rule of law point of view, however, matters are hardly so simple. First, enforcement of community norms, even if effective some of the time, does not provide sufficient widespread stability to be considered the rule of law.156 Players have no coercive power

 Risch, supra note 141.  Fairfield, Anti-Social, supra note 89, at 429. 150  Fairfield, Anti-Social, supra note 89, at 451. 151  Fairfield, Anti-Social, supra note 89, at 448–9. 152  Grimmelmann, Feudalism, supra note 10 (“Second Life land barons have no recourse but self-help” to deal with disputes among tenants). 153  Grimmelmann, Comparative, supra note 54, at 159–60; Lastowka & Hunter, Laws, supra note 6, at 51–2; Balkin, supra note 10, at 2062; Stoup, supra note 13, at 319–20. 154   Lastowka & Hunter, Laws, supra note 6, at 69–71; Fairfield, Anti-Social, supra note 89, at 459–61 and 471–2 (adverse selection caused by lax rules causes only the most abusive members to stay in a world). 155  Fairfield, Anti-Social, supra note 89, at 459–61; Fairfield, Magic Circle, supra note 106 (“penalties” are breaches of game rules that are not legally actionable); Kerr, Criminal, supra note 16, at 422. 156  Balkin, supra note 10, at 2044 (too much money is at stake for the law to allow community norms to settle all disputes); Goldsmith & Wu, supra note 12, at 135 (reputation model of eBay did not “scale up” to large number of users); Lessig, supra note 64, at 159 (norms may not be effective where the community of enforcers does not include those who bear the cost of the norms); Stoup, supra note 13, at 328–30 (too many users to regulate); Arias, supra note 4, at 1340–41 (community norm enforcement fails to regulate users); Matthew Williams, Virtually Criminal: Crime, Deviance and Regulation Online 138 (2006). But see Fairfield, Magic Circle, supra note 106 (to the extent that sovereigns consider norms, then norms can have real-world effect). 148 149

BARFIELD PRINT.indd 88

13/11/2018 11:52

Virtual rule of law  89 to enforce the rules, other than through player versus player harassment.157 Furthermore, norms may simply be arbitrary.158 Thus, while community enforcement can shape user behavior, it is often discretionary, inconsistently enforced, partial, and not determined by a neutral factfinder.159 As a result, community enforcement fails six of the ten indicators based on enforcement alone.160 Second, community norms are, overall, more vague and more secret than provider agreements.161 New users have no idea what the community expects,162 vague notions of tolerance and respect are often unspoken,163 norms do not effectively propagate to noncentrally located users,164 and norms shift over time based on new user influx,165 as well as complaints that are not aired publicly.166 Additionally, it is unclear what level of violation of community norms will rise to the level of independently actionable torts.167 Thus, users certainly consent to being governed by community norms,168 but they often have no idea what they are consenting to and, more importantly, no ability to find out other than through trial and error. Thus, there are no preannounced, publicly available, attainable, written, forwardlooking, impartially enforced rules. Using the hockey-fight example, new players may not know how much fighting is acceptable, and many players will be unaware that the fight has gone too far until it is too late.169 Selective community enforcement through online violence might seem like the right 157  Castronova, supra note 4, at 217; Burke, supra note 83, at 17; John Rothchild, Protecting the Digital Consumer: The Limits of Cyberspace Utopianism, 74 Ind. L.J. 893, 967–8 (1999) (LambdaMOO’s self-regulation failed because community norms are unenforceable); Stoup, supra note 13, at 328; but see Grimmelmann, Comparative, supra note 54, at 170–71 (discussing importance of guilds in forming and enforcing social norms despite their technical weakness). 158  Bartle, supra note 14, at 37. 159  Castronova, supra note 4, at 214 (“Even when guilds become powerful, their use of power rarely feels legitimate in the sense of being in the service of the community as a whole. Rather, they tend to act like a family of mighty people whose projection of their own power happens, coincidentally, to keep the peace on occasion”). 160   But see Zittrain, supra note 97, at 143 (arguing that community enforcement in Wikipedia, though not coercive, satisfies several indicia of the rule of law). 161  Bartle, supra note 14, at 35 (discussing unwritten rules of virtual worlds). 162   See, e.g., Grimmelmann, Comparative, supra note 54, 155–6 (Describing complexities of different communities dealing with “kill stealing,” a practice whereby one person or group attacks a monster that others are already fighting). The author recalls feeling incredibly awkward at having said the wrong thing or entered the wrong building without permission in an early non-graphical virtual world. 163  Grimmelmann, Comparative, supra note 54, at 180–81, suggests that providers might want such vagueness so that they may cater to their users by terminating accounts of unwanted users, or sometimes to allow a user to harass others if they pay a higher fee than users that might be driven away. 164  Stoup, supra note 13, at 332–3. 165  Stoup, supra note 13, at 333–4. 166  Balkin, supra note 10, at 2067. 167  Fairfield, Anti-Social, supra note 89, at 440–41. 168  Fairfield, Magic Circle, supra note 104 (consent is key to determining allowable conduct in virtual worlds). 169   “McSorley Found Guilty of Assault,” Associated Press, available at www.deseretnews. com/article/786934/McSorley-found-guilty-of-assault.html and https://perma.cc/27R3-6B5P;

BARFIELD PRINT.indd 89

13/11/2018 11:52

90  Research handbook on the law of virtual and augmented reality idea in the absence of other law, but the real world calls such behavior mob vigilantism.170 It may be necessary for enforcement, but it is hardly the rule of law. Further, it does not encourage more business in the virtual world.171 These concerns may transcend practical difficulties to present a theoretical bar to the rule of law. Communities shift by nature, and enforcement will always be at the whim of the group. Unless community norms can be “codified” in some form (such as an agreement) that is enforceable in a regular, consistent, and impartial manner, community norms will always be a barrier. 5.3 Factfinding An overriding problem with enforcement of market- or community-based regulations is the lack of any factfinding, let alone neutral factfinding, with respect to in-game violations.172 To be sure, parties can obtain a neutral determination in the real world, but there is no real-world right of action for many in-world wrongs and there are few real-world checks on the provider’s ability to terminate an account or otherwise to curb activity through software programming.173 Additionally, uncertainty about the enforcement of provider agreements may divert focus away from important regulation issues.174 Even if communities perform their own factfinding, there is no guarantee that providers will take any action based on the findings.175 Finally, the cost of obtaining a neutral real-world factfinding will likely outweigh the benefits of such adjudication even if the harm is actionable. However, difficulties in factfinding should be separated into those that affect the rule of law and those that do not. Factfinding that is absent due to lack of adjudication and Balkin, supra note 10, at 2068–9. But see Fairfield, Anti-Social, supra note 89, at 460 (“The process is just as simple as the one determining that tackling is not a tort, but high-sticking is”). 170  Williams, supra note 156, at 138; Tamahana, supra note 7, at 97 (following a rule is generalized while ad hoc actions are not); Zittrain, supra note 97, at 198–9 (discussing need for rule of law to avoid the abuse of community power when “[v]irtue [gives] way to narrow self-interest and corruption”); Grimmelmann, Comparative, supra note 54, at 168 (“When we look at the mechanisms by which players might enforce their notions of fair play and good behavior, an odd paradox emerges. The set of unpleasant and wrongful acts players might wish to deter is identical to the set of unpleasant and effective sanctions available as deterrence. To prevent violence, annoyance, and non-cooperation, players can engage in violence, annoyance, or non-cooperation”); Stoup, supra note 13, at 330–31 (shaming ineffective). 171  Castronova, supra note 4, at 209 (describing drop in world population when player versus player combat was allowed). 172  Grimmelmann, Comparative, supra note 54, at 173–4 (discussing limitations of provider designed political systems and norms enforcement). 173   Lastowka & Hunter, Laws, supra note 6, at 50–51 (provider agreements make simple inworld dispute resolution difficult). 174  Dibbell, supra note 72, at 144 (“Ruling the EULA to be a valid contract, on the other hand, would have sent the question back where it belonged—into the much more finely tuned evaluative process that is the ceaseless, grinding struggle between players and designers over the shape of the game”). 175  Arias, supra note 4, at 1340–41 (“Unfortunately, internal regulation methods have proven ineffective, because after most internal investigations, the MMORPG developer does not restore the stolen virtual goods to the victim”).

BARFIELD PRINT.indd 90

13/11/2018 11:52

Virtual rule of law  91 enforcement mechanisms implicates the rule of law.176 Factfinding rendered impractical because the cost of obtaining resolution is higher than the harm at issue concerns the rule of law far less. Access to justice can sometimes be costly, whether real or virtual. While inexpensive or even free adjudication may be normatively preferable, in most cases cost should not negate the rule of law, so long as it is not the sovereign imposing the cost. Thus, filing fees might negate the rule of law while attorneys’ fees would not. This distinction is relevant to provider agreements that require costly arbitration; where the providersovereign requires a large upfront cost such as arbitrator fees, an arbitration clause may negate the factfinding and enforcement indicia.177 5.4  Code as Law Provider-implemented code that constrains behavior in the virtual world suffers from few of the shortcomings just discussed. There are no vague contract terms. The community neither defines nor enforces the rules.178 There is no need for factfinding. Instead, the provider defines exactly what avatars can and cannot do—a reality often called the “wizard” or “god” phenomenon.179 5.4.1  Code and the rule of law The rule of law problem here is the inverse of user agreements as law. Instead of being vague, partial, and unevenly enforced, code is the ultimate impartial rule: It is clear, unambiguous, perfectly enforceable (if there are no bugs),180 treats everyone equally, and lacks all discretion.181 If a world disables the software that allows avatars to transfer 176   Lucille Ponte, Leveling up to Immersive Dispute Resolution (IDR) in 3-D Virtual Worlds: Learning and Employing Key IDR Skills to Resolve in-World Developer-Participant Conflicts, 34 U. of Ark. L. Rev. 713 (2012) (discussing need for in-world dispute resolution). 177   See, e.g., Bragg v. Linden Research, Inc., 487 F. Supp. 2d 593, 609 (E.D. Pa. 2007) (“Such [arbitration] schemes are unconscionable where they ‘impose[] on some consumers costs greater than those a complainant would bear if he or she would file the same complaint in court’”) ­(citation omitted). 178   But see Bradley & Froomkin, supra note 14, at 130–31 (code can influence social structure in virtual worlds). 179   Lastowka & Hunter, Laws, supra note 6, at 53 (“As we explain, even when these owners are not wholly adverse to democratic governance within the virtual spaces they maintain, their exclusive ability to exert absolute control over these environments hopelessly complicates attempts to map traditional notions of democratic governance onto these settings”); but see Lessig, supra note 66, at 192–3 (arguing that there is a battle for supremacy between real world laws and other forms of regulation in cyberspace). 180  Grimmelmann, Comparative, supra note 54, at 175–6 (code can implement new laws without having to “bribe, bedazzle, brainwash or bully” anyone in the way a dictator would, such as the ability to revoke property without any kind of resistance by the avatar possessing it); Grimmelmann, Feudalism, supra note 10 (“What offline governments can do only after lengthy legal proceedings, Linden does unilaterally, just by changing an entry in a database”), available atwww.yalelawjournal. org/forum/virtual-world-feudalism; Radin, supra note 94, at 11. 181  Grimmelmann, Comparative, supra note 54, at 153–4 (“But in the world of software, increasing complexity does not bring with it increased discretion. This is so because software operates by itself . . . Even as code-based property rights become increasingly complex, with more exceptions and special cases, they never become any less hard-nosed in their application”).

BARFIELD PRINT.indd 91

13/11/2018 11:52

92  Research handbook on the law of virtual and augmented reality virtual property to others, then avatars can no longer transfer virtual property inside the world. In itself, unilateral code is not a problem under the rule of law; as discussed previously, even a dictator can follow the rule of law. Indeed, depending on the administration, in theory code could be the perfect rule of law.182 Unfortunately, in practice code will often fail several indicators. Code changes are frequently hidden,183 unannounced,184 and potentially very arbitrary.185 These traits are potentially exacerbated if users provide the code.186 If administered in this way, code, too, would violate the rule of law.187 5.4.2  The interaction between code and agreements If a world’s code disallows an activity expressly permitted by contract, the rule of law will suffer from the provider not following its own rules. This renders the law contradictory/ arbitrary, vague, and nonpublic. Of course, there may be debate about whether an action is expressly permitted, though a vague agreement diminishes the rule of law as well. The opposite scenario, where code allows avatars to take actions otherwise barred by agreement, raises concerns, too. For example, a world may bar player versus player combat in its agreement, yet the software design may allow one avatar to attack another.188 Similarly, the code may allow an avatar to obtain another avatar’s virtual property through 182   Cf. Grimmelmann, Comparative, supra note 54, at 159–60 (describing how code allows for low-risk contracting in sale of goods due to software mechanisms that allow for clean transactions and nondefective goods). 183  Burke, supra note 83, at 9 (“Who decides, how they decide, and when they decide, are almost always unknowns”). 184  Bartle, supra note 14, at 39 (describing complaints when code reduces number of killable monsters); Burke, supra note 83, at 8 (“Most MMOG developers have treated communication with their customers as something of an afterthought, and have regarded it as a specialized activity to be managed independently from the development of the game itself ”). 185   Lastowka & Hunter, Laws, supra note 6, at 55; Burke, supra note 83, at 7–8 (“A precious few such changes are entirely neutral or positive, affecting all players equally”); Castronova, supra note 4, at 152 (describing how different segments plead to providers to enhance the benefits of their group, such as warriors); Dibbell, supra note 72, at 143 (describing user complaints leading to frequent changes in virtual world); Balkin, supra note 10, at 2051; Lessig, supra note 66, at 199 (“My main concern is accountability—these architectures and the values they embed should be architectures and values that we have chosen”); Bartle, supra note 14, at 39 (discussing repercussions of changes to coded prices for goods); Radin, supra note 94, at 13 (discussing industry group capture of intellectual property and digital rights management laws). 186  Zittrain, supra note 97, at 168–9 and 172 (discussing “rough justice” of a single private party providing code that regulates what actions others can take on the internet: “These private programs are serving important functions that might otherwise be undertaken by public authorities—and their very efficiency is what might make them less than fair”). 187  Grimmelmann, Comparative, supra note 54, at 181 (“One of the most frequently given pieces of advice is that almost nothing is more destructive to a thriving game community than regular intervention by its designers. Since designers are not bound by the same ‘laws’ of code as regular players, their presence can be destabilizing. When designers engage in conduct not available to players, it highlights their distance from players and their apparent unaccountability”). 188   Lastowka & Hunter, Virtual Crimes, supra note 2, at 309–10 (Ultima Online allows “harassment” by code, but game rules disallow such harassment, excluding theft and player killing, from the definition of harassment).

BARFIELD PRINT.indd 92

13/11/2018 11:52

Virtual rule of law  93 fraud. Stopping barred activity through code varies in difficulty; for example, stopping player combat is far simpler than programming against deception. The rule of law becomes relevant because many believe that if the software allows something, then it must be “legal.”189 Such users would argue that the rules are arbitrary because they are contradictory: one source of regulation bars an activity while the other allows it. The answer lies in the complexity of code—it is impossible to both predict and effectively block every type of action that might occur in a virtual world.190 For example, game rules and code provide almost no method for making complex agreements between avatars.191 The real world has an analog: door locks and fences.192 These codelike devices keep humans from entering real property, but their existence does not mean that one breaks no law by entering someone’s unlocked home. The rules are not contradictory simply because both locks and laws regulate behavior. Thus, the fact that code and agreements might allow different things is not a real problem for the rule of law. Both constrain users, and users must look to both as the source of regulation.193 There are exceptions, of course. For example, where code allows an action due to a program error—an action not otherwise barred by agreement—the provider might “undo” the results of the “exploit” retroactively.194 This retroactivity may be illustrated by two examples. First, a user of Second Life sued the provider for revoking property the user purchased through an auction based on its suspicion that the user exploited a loophole in the software.195 Second, a provider in South Korea reverted possession of a virtual castle

189  Nino, supra note 99 (“people aren’t used to the idea of having to make their own behavioral choices within software. They simply assume that that which is not permissible is simply not coded, and therefore not possible . . . They don’t understand. ‘If it’s not allowed,’ they demand to know, ‘then why does it let me?’”) (emphasis in original). Bartle, supra note 14, at 36. 190  Grimmelmann, Comparative, supra note 54, at 151–2 (discussing complexity of programming code for property rules); Nino, supra note 99 (“Every time you encode a restriction, exceptions crop up. Every exception you encode lets a few cases slip through the cracks. Soon your code starts to look like the US Tax Code, with more exceptions and encoded governance than environment simulation”); Bartle, supra note 14, at 36; Burke, supra note 83, at 4 (“In many cases, this is because the real-world capacities of human agents essentially outstrip the technical capacities of law-as-code”). 191  Grimmelmann, Comparative, supra note 54, at 158 (“While online games have strong protections for property, they have nothing that we would recognize as a comparable body of contracts law. Most games have no way to draft any contract more complex than an immediate sale of goods for cash”); Nino, supra note 115 (avatars are not legal entities that can make a contract). 192  Radin, supra note 94, at 12 (describing code as more like a fence than a trespass law). 193  Grimmelmann, Comparative, supra note 54, at 153 (“Code may be clear and free from doubt, but where ‘loopholes’ come up, the need for rules is present—humans will take advantage of the shortcomings of code, and there are always shortcomings”) (emphasis in original); Nino, supra note 99 (“The absence of code to allow you to perform a specific task or action does not necessarily constitute governance. The presence of code to allow you to perform a specific task or action does not necessarily constitute governance”). 194  Grimmelmann, Power, supra note 135, at 147–8 (describing provider responses to user exploits). 195   Bragg v. Linden Research, Inc., 487 F. Supp. 2d 593, 597 (E.D. Pa. 2007) (“Linden sent Bragg an email advising him that Taessot had been improperly purchased through an ‘exploit.’ Linden took Taesot away. It then froze Bragg’s account, effectively confiscating all of the virtual property and currency that he maintained on his account with Second Life”).

BARFIELD PRINT.indd 93

13/11/2018 11:52

94  Research handbook on the law of virtual and augmented reality to prior owners, claiming that the battle for that castle should not have been allowed by program code.196 To the extent that these actions were not barred by agreement (that is, the auction and battles were “proper” under game rules and the users were unaware of the program errors), then the retroactive changes were surely contrary to the rule of law. Then again, for the users who lost property due to computer bugs, the errors were arbitrary application of code as law.197 As a result, exploited program errors will likely diminish the rule of law in one way or another.198 5.5  Legislated Lawlessness and the “Rules of the Game” It is common for virtual worlds to allow unconstrained behavior—even behavior that might harm others. In such worlds, fraud, theft, avatar combat, and deception are “part of the game.”199 Professors Lastowka and Hunter analogize claims of actionable harm to a basketball player suing for “theft of a basketball on the court” and argue that avatars of a given world accept mandated lawlessness.200 The question is whether one can have a law that there is no law. This chapter concludes that mandated lawlessness is not rule of law. First, it is unlikely that new users would know before joining that what might otherwise be actionable wrongdoing, such as fraud, is legally acceptable within the game.201 Second, even if such “rules” are announced in some form, they are not concrete enough to plan affairs by. Instead, they increase costs and unsettle expectations just like vague and arbitrary rules.202 Mandated lawlessness is still lawlessness—few argue that the rule of law ­governed the Wild West.

  Lastowka & Hunter, Laws, supra note 6, at 72.  Bartle, supra note 14, at 41 (discussing complexity of determining whether benefits of an exploit should be reversed by provider: “Who decides it’s wrongful? What makes some actions in the virtual world ‘exploits’ when other, similar actions, aren’t?”); Grimmelmann, Power, supra note 135, at 148 “[E]very change . . . will privilege some players while hurting others”) and 149 (“There is nothing ‘wrong’ with the exploit, as far as the software is concerned”); Balkin, supra note 10, at 2051 (providers should be free to adapt game to preserve quality game play). 198  Grimmelmann, Power, supra note 135, at 147 (user agreements allow providers to do whatever they want in response to an exploit); Lastowka & Hunter, Virtual Crimes, supra note 2, at 315 (those harmed by exploits include the provider). 199   See, e.g., Grimmelmann, Comparative, supra note 54, at 150 n.11 (Ultima Online has a “steal” function built into the code); Fairfield, Anti-Social, supra note 89, at 460 (“For example, one virtual world called EVE Online is a science-fiction world of corporate fraud, yankee trading, and piracy. The game’s designers have openly stated that deception is part of the game—that, in fact, ‘fraud is fun’”). 200   Lastowka & Hunter, Virtual Crimes, supra note 2, at 304–5; Kerr, Criminal, supra note 16, at 419; Balkin, supra note 10, at 2062; but see Brenner, supra note 129, at 60 (arguing that real-world crimes map directly to virtual crimes). 201  Balkin, supra note 10, at 2067. 202   See, e.g., Jason T. Kunze, Regulating Virtual Worlds Optimally: The End User License Agreement, 7 Northwestern J. Tech. & Intell. Prop., at 4 (2008) (describing how user bank that absconded with deposited funds in EVE Online was not punished because such fraud was part of the game). 196 197

BARFIELD PRINT.indd 94

13/11/2018 11:52

Virtual rule of law  95 5.5.1  Legislated rules and the magic circle Presumably, real-world conduct can be regulated by real-world laws,203 such as the case where a user logged in using another person’s password and sold his virtual property.204 Logging into another’s account is a real-world act that might be punishable regardless of what the game rules are.205 For example, hacking into a computer to steal gold might be punishable, even if stealing gold within the game is allowed by the game rules.206 However, where the conduct is in-world and not clearly allowed by game rules, the analysis becomes murkier207—especially because of the complexity relating to hardcoding in-game rules.208 Vague agreements and norms exacerbate the conclusion that wide open game rules do not comport with the rule of law;209 regulation is either incomprehensible or nonexistent. To overcome these shortcomings, virtual worlds must either engage in much more thorough legislative activity with respect to allowable conduct in the virtual world,210 or more specifically rely on real-world background law.

203   Lastowka & Hunter, Virtual Crimes, supra note 2, at 298; Kerr, Criminal, supra note 16, at 417–19 (most crimes cover actions taken with respect to physical reality rather than what happens to avatars). 204  Castronova, supra note 4, at 237; Lastowka & Hunter, Virtual Crimes, supra note 2, at 298; but see Fairfield, Magic Circle, supra note 106 (if one shares password information, there can be no liability for unauthorized access). 205  Koster, supra note 13, at 56 (“It’s pretty clear that there are some rights which leak over from the real world”); Kerr, Criminal, supra note 16, at 423 (describing unauthorized use of password as crime); Balkin, supra note 10, at 2067–8 (consumer protection laws will apply to shopping in virtual worlds); Brenner, supra note 129, at 52–3. 206   Edward Castronova, The Right to Play, The State of Play: Law, Games, and Virtual Worlds 73 (Jack M. Balkin and Beth Simone Noveck, eds 2006), hereafter Right to Play. 207   Lastowka & Hunter, Virtual Crimes, supra note 2, at 296–7; Kerr, Criminal, supra note 16, at 421 (“Yet if the rules of the game trump, this raises the important question of what ‘the rules of the game’ in a virtual world actually are . . . [V]irtual worlds are open. In a practical sense, a user has an infinite number of moves he can make. Which moves are permitted and which are not may be difficult to know”). 208  Grimmelmann, Comparative, supra note 54, at 167–8 (“[T]hanks to the wildly varied set of rules and conventions for games, it is not possible to identify specific acts as right or wrong in a way that holds true across games . . . The right response to the question of ‘what is a virtual crime?’ turns out to be ‘any activity that genuinely bothers most players of the game in question’”). 209   Lastowka & Hunter, Laws, supra note 6, at 71 (“Nonetheless, in urging courts to avoid recognizing virtual law, the cyberskeptics may have a point. Given the complexity of ascertaining a virtual world’s emerging legal rules and balancing them with avatar rights and wizardly omnipotence, the prospect of real-world courts entertaining virtual disputes is in some ways not very appealing”); Fairfield, Anti-Social, supra note 89, at 438 (“[A] combination of resistance on the part of virtual-world creators (expressed through their EULAs) and reticence on the part of courts to take on unfamiliar technologies has left these communities without the legal rules they need to thrive”). 210  Koster, supra note 13, at 56 (Legislation has not “caught up to the notion of virtual spaces very well”); Post, supra note 4, at 185.

BARFIELD PRINT.indd 95

13/11/2018 11:52

96  Research handbook on the law of virtual and augmented reality 5.5.2  In-world legislation A few sites have even attempted more thorough in-world legislation.211 LambdaMOO, for example, at one time had a detailed set of rules based on majority voting.212 Other sites have instituted detailed community standards rules that are different from real-world law. These efforts show that nonvague, nonarbitrary lawmaking is possible, but even this lawmaking is not without problems. Neutral factfinding and impartial enforcement, for example, continue to elude, and online communities may exhibit arbitrary and kneejerk reactions to perceived wrongs.213 More important, despite the fact that in-world regulation may be normatively preferable,214 legislating an entire criminal and civil law for each virtual world is doomed to fail. World-specific laws can effectively capture the vagaries of that world,215 but no provider or population has the time, energy, or will to create a detailed set of laws.216 Those laws would be exceedingly complex, including real and personal property, tax, insurance, torts, contracts, secured transactions, governance, and dispute resolution, among other things.217 These varied areas keep dedicated real-world governments busy; game players are unlikely to pay for the opportunity to work at creating, managing, and enforcing such

211  Castronova, supra note 4, at 152–3 (discussing political decisions in virtual worlds); Dibbell, supra note 72, at 142–3 (describing user input to providers for rules). See also Castronova, Right to Play, supra note 206, at 68 (suggesting that real-world laws be used to require all in-world activity to be separated from real-world activity); Lastowka & Hunter, Virtual Crimes, supra note 2, at (suggesting that virtual communities should define virtual crimes); Stoup, supra note 13, at 338–9 (recommending that virtual worlds implement all-encompassing legislation); Burke, supra note 83, at 16. 212   Lastowka & Hunter, Laws, supra note 6, at 55–9; but see Castronova, supra note 4, at 217 (democracy failed and the provider retook authority). 213   But see Lastowka & Hunter, Laws, supra note 6, at 70–71 (discussing reasoned deliberations in LambdaMOO after virtual sexual assault). Of course, that deliberation imposed a penalty for an offense that was not defined beforehand, which presents a separate rule of law problem. 214   Lastowka & Hunter, Laws, supra note 6, at 71 (“Perhaps, therefore, it would be best to require that the laws of the virtual worlds develop within their own jurisdiction. Perhaps, even if we accept that real lives, economic values, and substantial investments are at play within virtual worlds, the wiser course may be for courts to keep their distance”); Post, supra note 4, at 185 (“So why not begin by recognizing their right—perhaps even their inalienable right?—to govern themselves as they see fit? Why not let those who choose to enter, and to interact within, these online communities make their own law, deciding for themselves how they’d like to order their affairs?”). 215  Fairfield, Anti-Social, supra note 89, at 460–61; Dibbell, supra note 72, at 144; Balkin, supra note 10, at 2073–4; Stoup, supra note 13, at 337–8. 216   Lastowka & Hunter, Laws, supra note 6, at 55–9 (wizards did not have enough time to handle all disputes); Dibbell, supra note 72, at 142–3. But see Beth Simone Noveck, Democracy— The Video Game: Virtual Worlds and the Future of Collective Action, The State of Play: Law, Games, and Virtual Worlds 258–60 (Jack M. Balkin & Beth Simone Noveck, eds 2006) (democracy is more likely to thrive in virtual worlds than in other internet-based applications due to social interaction); Post, supra note 4, at 185–6 (if lawmaking institutions were created in virtual worlds, self-governed laws will follow). 217   Bradley & Froomkin, supra note 14, at 139–43 (describing different areas that could be legislated differently in virtual worlds); Fairfield, Anti-Social, supra note 89, at 429 and 432 (“Instead, they fail because contracts cannot cheaply create default rules that bind large and shifting populations”).

BARFIELD PRINT.indd 96

13/11/2018 11:52

Virtual rule of law  97 a detailed legislative system.218 Furthermore, providers cannot effectively implement rules that the majority may want, because such rules might lead to a disproportionate amount of abuse by those who take advantage of such rules.219 These concerns are more than mere practical hurdles. The complexity of an entirely self-legislated world, especially when combined with the inherent problem that desired rules might incentivize behavior by a few people that can significantly harm the world, renders autonomous self-regulation theoretically defunct. 5.5.3  Importation of real-world law The failure of complete virtual legislation means that providers and users must rely, at least in part, on well-developed real-world laws.220 Many sites do so, outlawing activity that would otherwise be illegal in the real world. This satisfies several rule of law indicators, even if the enforcement-related indicators are unmet. However, real-world laws are often ill-suited to game rules.221 Laws against murder, for example, are irrelevant where player combat is the norm.222 Even simpler crimes, such as theft, have no relevance in a world where theft is encouraged.223 The challenge is for each world to legislate exactly which real-world laws might apply to in-world wrongdoing, and how.224 A world might outlaw defamation but accept theft, for example.225 218   But see Bradley & Froomkin, supra note 14, at 143 (virtual worlds are much less complex than the real world with respect to areas that need to be legislated). 219  Fairfield, Anti-Social, supra note 89, at 472 (“[M]any communities currently have strict anti-harassment rules. Consider what would happen if a new virtual community were to relax those strict rules. If the new community were to state that it will have a higher tolerance for harassment, then a higher proportion of harassers would enter the community, eventually driving off even those who initially had a higher tolerance for such behaviour”); Rothchild, supra note 157, at 967–8 (LambdaMOO’s self-regulation failed because community punishment is unenforceable); Grimmelmann, Feudalism, supra note 10 (pre-emption of contracts “could leave [providers] powerless against abusive users who spoil the experience for others”). 220  Fairfield, Anti-Social, supra note 89. 221  Castronova, Right to Play, supra note 206, at 73; Fairfield, Anti-Social, supra note 89, at 459; Kerr, Criminal, supra note 16, at 419 (“But misconduct that draws social significance from its meaning in virtual reality normally will have no resonance with criminal statutes”); Lessig, supra note 66, at 198 (“As the rules that govern real space compete, cyberspace increasingly wins out”). 222   Lastowka & Hunter, Virtual Crimes, supra note 2, at 303 (“Indeed, one might reasonably predict that since Ultima Online is commonly understood to be a computer game, the gut reaction of state and federal prosecutors would be to view the legal status of thefts that take place in Ultima Online as similar to the gruesome murder of PacMan at the hands of Inky, Blinky, Winky, or Clyde”). This is especially true where players do not really die, but are instead simply weakened and sent far away. Castronova, supra note 4, at 305, note 3. But see Brenner, supra note 129, at 87–9 (problem with virtual murder is that it may incite users to commit real acts of violence). 223  Castronova, Right to Play, supra note 206, at 73; Lastowka & Hunter, Virtual Crimes, supra note 2, at 301–3; see also Arias, supra note 4, at 1306–8 (describing several types of theft: embezzlement, deception, pickpocketing, and hacking). 224  Fairfield, Anti-Social, supra note 89, at 459; Fairfield, Magic Circle, supra note 106 (“The proposed model of self-governance is not sovereign lawmaking, but industry self-regulation and the scope of consent in online play”). 225  Balkin, supra note 10, at 2063 (communications torts will apply in virtual worlds as well as real-space).

BARFIELD PRINT.indd 97

13/11/2018 11:52

98  Research handbook on the law of virtual and augmented reality However, world-by-world compromises about which law to import, though facially helpful, solve few rule of law problems inherent with a permeable barrier between real and virtual worlds—a broken magic circle.226 This is most true where behavior is allowed in-game, but where such behavior has real-world consequences.227 For example, even if in-world rules allow false statements about avatars, such statements might cause real harm to the avatar’s user that real-world defamation law will vindicate despite game rules.228 Game rules will not provide real-world immunity, because they only apply to in-game harm. Indeed, arguments that users cannot sue the provider for actions of other users,229 or sue other users for breach of the provider contract,230 make the problem worse. If users are not contractually bound to each other, then the game rules cannot immunize conduct outside the game vis-à-vis other users.231 More generally, the mechanism by which public law values might regulate private world behavior is exceedingly complex.232 It is unclear whether and which real-world laws will preempt game rules and provide a cause of action for an aggrieved user.233 This highlights one of the most pressing concerns: nobody seems to have a firm understanding about how real-world laws will apply to virtual wrongdoing.234 Many have contributed outstanding

226   “Magic circle” is a term of art that describes the barrier between in-world activities and realworld activities. Fairfield, Magic Circle, supra note 106. This circle, however, is routinely breached. Fairfield, Anti-Social, supra note 89, at 434. 227   In-game prohibitions are not as problematic, because behavior that is barred in-game because of real-world illegality will also bear real-world consequences. 228  Balkin, supra note 10, at 2063 (communications torts will apply to conduct in virtual worlds); Brenner, supra note 129, at 54–5. Edward Castronova argues that any permeability between virtual and real worlds will necessarily harm human well-being. Castronova, Right to Play, supra note 206, at 68. 229   47 U.S.C. § 230 (immunizing providers for content provided by users). 230   See notes 148 and 149, supra, and accompanying text. 231   As discussed in the conclusion, allowing users to exercise third party beneficiary rights may enhance the rule of law. 232  Zittrain, supra note 97, at 172–3 (discussing complexity of applying “company town” jurisprudence from Marsh v. Alabama, 326 U.S. 501 (1946), to private actors in cyberspace); Balkin, supra note 10, at 2076–9 (same); Lastowka & Hunter, Laws, supra note 6, at 59–60; but see Benkler, supra note 14, at 182 (virtual worlds are not company towns, and users are free to leave virtual worlds). 233   Lastowka & Hunter, Virtual Crimes, supra note 2, at 305–6 (discussing conflict between sports rules and tort laws); Fairfield, Anti-Social, supra note 89, at 459–61 (discussing sports rules and exceeding allowed conduct); Fairfield, Magic Circle, supra note 106; Balkin, supra note 10, at 2046 (determining how law will apply to worlds is important because they can preempt norms) and 2066–7 (discussing whether fraud laws should apply to in-game fraud); Tal Zarsky, Privacy and Data Collection in Virtual Worlds, The State of Play: Law, Games, and Virtual Worlds 221–2 (Jack M. Balkin and Beth Simone Noveck, eds 2006) (discussing privacy rights in virtual worlds and real-world preemption); Lessig, supra note 66, at 199 (“Government should push the architecture of the Net to facilitate its regulation, or else it will suffer what can only be described as a loss of sovereignty”); Brenner, supra note 129, at 67–9 and 79–81 (discussing “victimless” cyber crimes such as prostitution and “consensual” virtual rape). 234   Joseph Bernstein, Virtual Reality Isn’t Ready To Handle Abusive Trolls, BuzzFeed (Nov. 4, 2016), www.buzzfeed.com/josephbernstein/virtual-reality-isnt-ready-to-handle-abusive-trolls and https://perm.cc/ULJ5-EQ76 (discussing limitation of liability on providers for acts of users); Lastowka & Hunter, Virtual Crimes, supra note 2, at 311 (“Courts and legislators may conceivably

BARFIELD PRINT.indd 98

13/11/2018 11:52

Virtual rule of law  99 work toward answering this question, but actual legislation and court decisions are insufficient to determine what laws will apply.235 Until the laws applying to virtual worlds are far more developed, behavioral crossover with the real world will reduce the rule of law to a shambles.236 No matter how well developed a game’s rules might be (whether through agreement, code, or norms), the user will not know what their rights, remedies, limits, or penalties are. This is by definition an arbitrary rule. In a sense, this is both the easiest and the most difficult point of analysis. It is easy because the lack of any clear rules makes a finding of no rule of law straightforward.237 It is difficult, though, because simply dismissing the potential for real-world laws is not a terribly palatable answer. Surely, there must be some possible ironclad analysis applying laws to behaviors in virtual worlds: harassment, theft, spam, conversion of intangible property, and so forth. Such analysis is the subject of many books and chapters but, other than a smattering of cases internationally, few rulings have provided answers. As virtual worlds grow, more answers about whether the real world will provide the rule of law for virtual-world participants may appear. 5.6  Jurisdiction and Real-World Law Some have argued that real-world territorial uncertainty violates the rule of law. For example, Professor Post argues that participants in cyberspace—and virtual worlds in particular—face this uncertainty.238 Post describes cases in which a foreign country imposed laws on noncitizens who had no contact with the country other than online activities about which the countries seemed to care.239 This, he argues, violates the rule of law: “It’s a strange kind of law being served up by the Unexceptionalists—law that only gets revealed to the interacting parties ex post facto, and which can therefore no longer guide the behavior of those subject to it in any meaningful way.”240 It is certainly true that countries might apply their laws in an arbitrary way against noncitizens. This, however, is not a problem of the rule of law in virtual worlds, but rather a problem of rule of law in those specific countries. Countries that try noncitizens refuse to defer to the private orderings created by contract and software. But we cannot, at this point, predict under what circumstances legislatures and courts will be willing to depart from the current default rule of nearly absolute wizardocracy”); Fairfield, Anti-Social, supra note 89, at 440–41; Post, supra note 4, at 182–4 (describing complications of regulating virtual banks and determining which laws apply to activities); Balkin, supra note 10, at 2071–2 (discussing uncertainty about enforceability of provider agreements). 235  Fairfield, Anti-Social, supra note 89, at 435. 236  Dibbell, supra note 72, at 144 (“Considering the novelty of this realm, we might reasonably hope for future case law and legislation to do a better job of it, I suspect it will be a long time before enough of those ambiguities are ironed out to make a difference”). 237   Cf. Post, supra note 4, at 184. 238  Post, supra note 4, at 163 (“The tricky part, though, is: Which law? Whose law? . . . [O]n the inter-network, information moves in ways that seem to pay scant regard to [national] boundaries, and mapping them onto network activity is a profoundly difficult challenge”). 239  Post, supra note 4, at 164–5. As a result, those users might be subject to penalties if they ever enter that country. 240  Post, supra note 4, at 169.

BARFIELD PRINT.indd 99

13/11/2018 11:52

100  Research handbook on the law of virtual and augmented reality in absentia for activities that do not touch on the country surely fail several of the indicia of the rule of law with such arbitrary, targeted, and partial enforcement of laws. To be sure, virtual worlds make user conduct more visible to the authorities in such countries, but real-world television, radio, publishing, and banking activities will result in the same uncertainty with respect to knowing which country might claim jurisdiction over activities. Iran’s fatwa calling for the death of Salman Rushdie due to his publication of The Satanic Verses is a prominent example. More relevant are the effects of territorial uncertainty where conduct might legitimately affect a certain country. To be sure, uncertainty about the laws that might apply to an activity is not helpful guidance.241 But here, too, such uncertainty relates to the nature of overlapping sovereignty rather than the specifics of virtual worlds. To the extent that jurisdictional uncertainty weighs against the rule of law, it does so generally with respect to any medium that crosses borders, and such uncertainty existed long before the internet. Indeed, this is why there are treaties about service of process, intellectual property, child custody disputes, and a myriad of other potentially border-crossing disputes. It may be that certainty about which sovereign will claim jurisdiction over a matter is unattainable, but people live with this type of uncertainty every day. Blue-sky laws in securities regulation are a good example, but even making a telephone call or sending a letter by international mail can implicate the laws of several states and even countries. To be sure, users will likely not know which country other avatars come from, but anecdotal evidence suggests that most users congregate in locally provided worlds—language barriers alone encourage locality. Even if they did not, people often do not know when their activities might implicate the laws of other countries, especially in modern commerce.

6. CONCLUSION This chapter’s analysis shows that virtual worlds exhibit—in practice, at least—few of the indicators of the rule of law. The reasons for the failure depend on the source of regulation. Market-based regulations, such as contracts, lack neutral and consistent enforcement mechanisms. Code-based constraints are often implemented arbitrarily and without notice. Community norms are often vague, unwritten, and enforced by mob rule. Autonomous self-regulation is too complex and costly. Real-world laws, no matter how clear and impartial in real-space, do not have a history that gives any confidence about how they might apply to virtual activity. The chapter also shows that academically popular sources of regulation—community norms and autonomous self-regulation—are the least likely to achieve the rule of law. Whether providers, users, or lawmakers should do anything about these failings depends on normative judgments that this chapter expressly avoids. If, however, providers want to attract paying customers, they might choose to do so by enhancing the rule of law as it applies to their users. Of course, this is easier said than done; what they might want to do and what they can actually achieve are two different things. The following are a few potentially achievable reforms. 241

 Post, supra note 4, at 183–5.

BARFIELD PRINT.indd 100

13/11/2018 11:52

Virtual rule of law  101 First, providers should provide more specificity about allowable activity in their agreements, with defined penalties that match the offense.242 Further, community norms—especially those enforced by the provider’s agreement—should be written and published.243 Many sites already do this, but most postings are vague. Second, this specificity might explicitly incorporate which real-world laws apply (or not) to in-world behavior. Common torts and crimes have fairly well-defined elements, and these can apply consistently to in-game behavior. Such incorporation will more precisely clarify what game rules allow and what they do not. Of course, real-world laws may still preempt the agreement, because (1) some in-game activity may be intolerable even if the parties agree, and (2) some in-game activity may cause real-world harm that is preempted by real-world law. Third, providers could incorporate explicit third party beneficiary clauses into their agreements to allow users to seek relief against other users for breaches of the rules. While such provisions are arguably unnecessary, they would provide clarity about actionable claims.244 This is especially important to providing enforcement where the provider has no desire to enforce its own agreement. Furthermore, to the extent that providers want to provide rule of law without losing central enforcement responsibility, the provider can enumerate exactly which types of claims are actionable while expressly reserving breaches that only the provider can enforce. Fourth, providers could institute some neutral arbiter of fact, especially when they resist third party beneficiary claims.245 For example, avatar arbitrators that hear argument and make factual findings could judge in-world wrongdoing.246 Nonvirtual wrongdoing, such as exploiting a system bug in breach of contract, could involve real-world arbitrators (hopefully cheaply) making factual findings.247 Neither option need be a full-blown 242  Stoup, supra note 13, at 337–8; Fairfield, Magic Circle, supra note 106 (worlds can selfgovern by creating penalties for breach of game rules); cf. Grimmelmann, Comparative, supra note 54, at 169 (“In deterrence terms, what a virtual world needs, in some sense, is a properly graded scheme of punishments”); Kerr, Criminal, supra note 16, at 428 (“virtual crimes should trigger virtual remedies”). 243  Grimmelmann, Comparative, supra note 54, at 181–3 (comparing real world rules with potential virtual world rules: “For similar reasons, administrative agencies issue guidance documents not only to put others on notice of the applicable law, but also to send a signal that they are constraining their own discretion with respect to particular issues”). 244   Compare Risch, supra note 141 with Fairfield, Anti-Social, supra note 89. 245  Grimmelmann, Comparative, supra note 54, at 181–3 (describing the beneficial implementation of certain rule of law features in virtual worlds: “I suspect that the ‘best practices’ of good games, ones which players think are basically fair, will closely resemble some of the ‘best practices’ of good governments”). 246   Cf. Grimmelmann, Comparative, supra note 54, at 169 (“The key is that the community as a whole needs sanctions not available to individuals”); Noveck, supra note 216, at 260 (virtual worlds provide mechanisms for large-scale deliberations); Goldsmith & Wu, supra note 12, at 132 (discussing early dispute resolution on eBay by “Uncle Griff ”); Lastowka & Hunter, Laws, supra note 6, at 50 (“For instance, in LambdaMOO and other textual MUDs, the ethos is one of sharing and community, and property disputes seem capable of resolution within the confines of the virtual world”); Zittrain, supra note 97, at 142–3 (describing dispute resolution mechanisms by the Wikipedia community). 247  Arias, supra note 4, at 1340–41 (discussing community based dispute resolution); cf. Goldsmith & Wu, supra note 12, at 137 (discussing SquareTrade online mediation service).

BARFIELD PRINT.indd 101

13/11/2018 11:52

102  Research handbook on the law of virtual and augmented reality arbitration so long as a neutral party can determine whether the claimed offense occurred before the contractual penalty is issued by the provider.248 These are just a few of the ways to implement a more robust rule of law. There are surely other methods that can better satisfy the ten indicators. Further, these methods can and should be targeted at the specific shortcomings that this chapter identifies. However, some problems, such as uncertainty about real-world laws, can only be improved with time and experience.

SOURCES Cases Bragg v. Linden Research, Inc., 487 F. Supp. 2d 593 (E.D. Pa. 2007). United States v. Lowson, No. CRIM. 10-114 KSH, 2010 WL 9552416, at *6 (D.N.J. Oct. 12, 2010). United States v. Nosal, 676 F.3d 854, 863 (9th Cir. 2012). United States v. Nosal 844 F.3d. 1024 (2016) (access without permission can constitute violation), https://cdn. ca9.uscourts.gov/datastore/opinions/2016/12/08/14-10037.pdf.

Statutes U.S. Const. Amends I–X. 47 U.S.C. § 230.

248   In the alternative, the avatar death penalty could be barred absent real adjudication by some neutral party.

BARFIELD PRINT.indd 102

13/11/2018 11:52

4.  Mixed reality: how the laws of virtual worlds govern everyday life Josh A.T. Fairfield*

1. INTRODUCTION Imagine a museum. The museum has statuary pictures, and a fountain located in the atrium. As you enter the museum, you adjust your glasses that offer you four different experiences. You select the first channel. A virtual docent appears and begins to provide a lecture accompanied by multimedia presentations of each of the exhibits.1 You can also see wiki-style notes and comments on each note written and rated for usefulness by former museum visitors.2 Curious, you flip to channel two—intended for children. There, you experience conversations with the artworks themselves, which animate and engage you in entertaining banter. The third channel contains entirely virtual exhibits that occupy physical spaces in and around the museum. The fourth and last channel displays an avantgarde mash-up in which artists from across the city have experimented with and built on the exhibits, altering pictures in subtle ways, adding lighting to statues to change the effect, or outrightly modifying the exhibits so that they are half real, half virtual.3 Although this type of technology may sound like science fiction, New York’s Museum of Modern Art (“MoMA”) has already featured such an installation.4 Now, let us interject reality. The scene described above will never be the future. In the future rather than four channels there will be thousands of channels offering experiences in the museum from the fantastical to the statistical, and everything in between. The merging of the Internet with the physical world around us—“realspace”—is called Mixed Reality. Mixed Reality technology is already in use and its adoption is only accelerating.5

*  Originally published in the Berkeley Technology Law Journal (citation: 27 Berkeley Tech. L.J. 55 (2012)). 1   See Alexander Fidel, Art Gets Unmasked in the Palm of Your Hand, N.Y. TIMES (Dec. 1, 2010), http://www.nytimes.com/2010/12/02/arts/02iht-rartsmart.html (discussing the use of smartphones, overlaying digital content onto real spaces that effectively connects the content to a realspace anchor, creating “augmented reality”—for example, a museum patron who points his or her smartphone at a sculpture, and the artist appears on the screen ready to be interviewed). 2   Id. (discussing the use of Layar, an augmented reality app that can tap into multiple layers of reality tied to realspace locations, like the MoMA). 3   Id. 4   Id. (identifying the MoMA’s application as one of the most popular). 5   See Woodrow Barfield, Commercial Speech, Intellectual Property Rights, and Advertising Using Virtual Images Inserted in TV, Film, and the Real World, 13 UCLA Ent. L. Rev. 153, 158–59 (2006) (discussing the use of a mobile computer that is used in conjunction with a wireless network resulting in the use of information from the Internet to mediate reality, such as virtual advertising on real objects); see also 3 Janna Quitney Anderson & Lee Raine, Ubiquity, Mobility, Security: The Future of the Internet 297–99 (2009).

103

BARFIELD PRINT.indd 103

13/11/2018 11:52

104  Research handbook on the law of virtual and augmented reality Imagine a world in which your clothes are free but your clothes carry shifting advertisements on smart fabric. Imagine a world in which your communication and interaction with others are so extensively digital that you can choose to edit your ex-husband out of your existence—you won’t have to see him, hear him, see anything he has written, hear any phone calls, nothing. The future of Mixed Reality, where virtual world technologies govern our everyday life, is already here.6 Through mobile technology, computing has finally come out from behind a desk and into the street.7 As a result, the laws that govern virtual worlds have a greater and greater impact on our everyday lives.8 However, the law is playing a desperate game of catch-up in order to adapt as disputes and lawsuits over Mixed Reality arise.9 Most scholarship to date has assumed that modern society is increasingly virtualized.10 It is more accurate to note that virtual data is increasingly realized as it becomes tied to realspace features and geography.11 Yet while virtual experiences are entering real life at an ever-increasing pace, the legal literature on virtualization technologies lags badly. The bulk of virtual worlds research focuses on the impact that real world regulatory regimes have on online spaces and communities.12 This chapter proposes that the traditional focus

 6   Brian X. Chen, Always On 4–7 (2011) (discussing the far-reaching impact of the iPhone and its role in weaving data with physical reality).  7   Sometimes, literally into the street, as in the example of augmented reality windows in upcoming Toyota vehicles. See Toyota’s “Window to the World” Offers a Taste of Driving Technology to Come, Independent (London) (July 28, 2011), http://www.independent.co.uk/life-style/motoring/ toyotas-window-to-the-world-offers-a-taste-of-driving-technology-to-come-2327504.html (“In the future . . . drivers can expect windshields to act in a similar manner, able to overlay digital information for practical, rather than educational or entertainment purposes.”).  8   See Bragg v. Linden Research, Inc., 487 F. Supp. 2d 593 (E.D. Pa. 2007) (detailing the use of a Term of Service, or “TOS”). TOSs—or End User License Agreements (“EULAs”) as they are often called—are the dominant form of legal relationship in virtual worlds. See Joshua A.T. Fairfield, Anti-social Contracts: The Contractual Governance of Virtual Worlds, 53 McGill L. J. 427, 429 (2008) (discussing the prevalence of EULAs in virtual worlds). EULAs, as contracts, define the terms of the relationship between the company and the user. Not unlike Linden Research’s game Second Life, mobile phone carriers and the creators of mixed reality applications use EULAs and TOSs to control their software. This is already evident in any app downloaded from the Apple Store. See Legal Information & Notices, Apple.com, http://www.apple.com/legal/ terms/site.html (last updated Nov. 20, 2009).  9   See, e.g., Rosenberg v. Harwood, No. 100916536, 2011 WL 3153314 (D. Utah May 27, 2011) (awarding Google’s motion to dismiss because Google was found not negligent). The Plaintiff sued Google after she used the Google Maps direction feature and was struck by a car. See also Kirit Radia, Google Nearly Starts a War. Seriously., ABC News: Note (Nov. 11, 2010, 12:43 PM), http:// blogs.abcnews.com/thenote/2010/11/google-nearly-starts-a-war-seriously.html (describing how two nation-states in Central America almost started a war after Google Earth showed a border in the wrong location). 10   See, e.g., Jack M. Balkin, Virtual Liberty: Freedom to Design and Freedom to Play in Virtual Worlds, 90 Va. L. Rev. 2043 (2004); Gregory Lastowka & Dan Hunter, The Laws of the Virtual Worlds, 92 Calif. L. Rev. 1 (2004); Juliet M. Moringiello, What Virtual Worlds Can Do for Property Law, 62 Fla. L. Rev. 159 (2010). 11   See Doug Gross, New Wave of Location-Based Apps Mark a “Paradigm Shift,” CNN Tech (July 29, 2011), http://www.cnn.com/2011/TECH/mobile/07/29/discovery. apps/index.html; see also Fidel, supra note 1. 12   See Lastowka & Hunter, supra note 10, at 1, 11, 29 (discussing the permeable nature

BARFIELD PRINT.indd 104

13/11/2018 11:52

Mixed reality  105 is backwards—that the legal regimes governing virtual worlds are increasingly coming to govern real world day-to-day life. This trend is accelerating as Mixed Reality applications integrate virtual objects and experiences into the real world.13 The growing application of online law to realspace is a problem because offline and online law have significantly diverged. Consider the simple act of purchasing a book. If you purchase a book offline, you own the book. If you purchase an e-book, you own nothing.14 As Mixed Reality technologies merge real and cyberspace, the critical question is whether online or offline law will determine consumers’ rights over property and data. There is a very real risk that courts will continue to reason from online analogies for offline issues, rather than turning to offline common law rules to determine consumers’ rights. This chapter proposes rebalancing the law governing Mixed Reality by using analogies to real world situations rather than limiting legal analysis to intellectual property and online licensing law. Because the common law proceeds by reasoned progression based on the closest available analogy, it seems more in line with the American legal tradition to look to “real world” law for Mixed Reality, despite the increasing virtual enhancement enabled by Mixed Reality applications. For example, imagine that a disgruntled neighbor defaces your home with an obscene word that appears when your house is viewed using a Mixed Reality application. The most appropriate analogy may be to the law of property and trespass rather than referring to the online licensing agreements of the application creator. In proposing a rebalancing, this chapter bridges serious gaps in two sets of legal literature. First, there is a gap in the legal literature with respect to the impact of Mixed Reality applications. Over 200 articles have been published on law and virtual worlds or virtual reality in recent years.15 To date, none have focused on the legal impact of Mixed between virtual worlds and the real world and the role of “real-world” law such as property); Leandra Lederman, “Stranger Than Fiction”: Taxing Virtual Worlds, 82 N.Y.U. L. Rev. 1620, 13   See Jason W. Croft, Antitrust and Communications Policy: There’s an App for Just About Anything, Except Google Voice, 14 SMU Sci. & Tech. L. Rev. 1, 1–4 (2010) (discussing the widespread growth in smartphones, in particular the iPhone, and the extensive offerings of the app store); see also Dan Fletcher, 10 Tech Trends for 2010, TIME (Mar. 22, 2010), http://ti.me/AfOUk4 (detailing the rise in augmented reality particularly among iPhone apps). 14   Gregory K. Laughlin, Digitization and Democracy: The Conflict Between the Amazon Kindle License Agreement and the Role of Libraries in a Free Society, 40 U. Balt. L. Rev. 3, 5 (2010) (“Amazon . . . retains ownership of the ‘Digital Content’ (i.e., the e-book) and imposes a number of restrictions that are inconsistent with transfer of ownership to the purchaser, including prohibiting redistribution.”). 15   See, e.g., Balkin, supra note 10; Bryan T. Camp, The Play’s the Thing: A Theory of Taxing Virtual Worlds, 59 Hastings L. J. 1 (2007); Jack L. Goldsmith, Against Cyberanarchy, 65 U. Chi. L. Rev. 1199 (1998); I. Trotter Hardy, The Proper Legal Regime for “Cyberspace,” 55 U. Pitt. L. Rev. 993 (1994); Steven Hetcher, User-Generated Content and the Future of Copyright: Part Two— Agreements Between Users and Mega-sites, 24 Santa Clara Computer & High Tech L. J. 829 (2008); Andrew E. Jankowich, Property and Democracy in Virtual Worlds, 11 B.U. J. Sci. & Tech. L. 173, (2005); Sarah K. Jezairian, Lost in the Virtual Mall: Is Traditional Personal Jurisdiction Analysis Applicable to e-Commerce Cases?, 42 Ariz. L. Rev. 965 (2000); Lastowka & Hunter, supra note 10; Juliet M. Moringiello, What Virtual Worlds Can Do for Property Law, 62 Fla. L. Rev. 159 (2010); Michael H. Passman, Transactions of Virtual Items in Virtual Worlds, 18 Alb. L. J. Sci. & Tech. 259 (2008); David G. Post, Against “Against Cyberanarchy,” 17 Berkeley Tech L. J. 1365 (2002); Steven R. Salbu, Who Should Govern the Internet?: Monitoring and Supporting a New Frontier, 11

BARFIELD PRINT.indd 105

13/11/2018 11:52

106  Research handbook on the law of virtual and augmented reality Reality applications16 even though Mixed Reality is far more common and commercially important than are pure virtual worlds.17 This gap is all the more important because computing has begun a great migration away from desktop computers and towards laptops, tablets, and most of all, smartphones.18 Mobile computing has led to the augmentation of real people, places, and things with virtual experiences and data.19 For example, when you compare prices on eBay or Amazon from your desktop, there is not a pressing need to tie the data to a specific location. But if you use a smartphone barcode scanner to compare prices as you shop in the local supermarket, the actual physical location of competing products at lower prices matters. Stores that augment their brick-and-mortar locations with virtual data to assist tech-savvy shoppers will gain a competitive advantage.20 With the advent of smartphone technology, the importance and depth of adoption of Mixed Reality applications far outstrips that of pure virtual reality applications. The second gap lies in the legal literature of pervasive computing (“PerC”). PerC is the predicted future embedding of computer chips into the physical environment. In the future, PerC theorists predict that there will be microprocessors—such as radio frequency identification (“RFID”) chips—in credit cards, shoes, toasters, walls, ceilings, and refrigerators.21 But the PerC literature has missed the mark because data tagging of

Harv. J. L. & Tech. 429 (1998); Andrew D. Schwarz & Robert Bullis, Rivalrous Consumption and the Boundaries of Copyright Law: Intellectual Property Lessons from Online Games, 10 Intell. Prop. L. Bull. 13 (2005); Allan R. Stein, The Unexceptional Problem of Jurisdiction in Cyberspace, 32 Int’l L. 1167 (1998); Richard S. Zembek, Jurisdiction and the Internet: Fundamental Fairness in the Networked World of Cyberspace, 6 Alb. L.J. Sci. & Tech. 339 (1996). 16   See Barfield, supra note 5 (discussing the use of augmented reality exclusively in the virtual advertising context). 17   Compare Greg Lastowka, Virtual Justice: The New Laws of Online Worlds 9 (1st ed. 2010) (“[I]n 2009, by conservative estimates, about 100 million people were interacting in some sort of virtual world . . . [and] about 10 percent of adults in the United States have participated in some kind of virtual world.”), with Michael K. Cheng, iPhone Jailbreaking Under the DMCA: Towards a Functionalist Approach in Anti-circumvention, 25 Berkeley Tech. L. J. 215 (2010) (discussing the rapid growth to date, predicted growth, and impact of smartphones along with their associated app stores), and Elaine Glusac, Travel Apps 2.0, N.Y. Times: In Transit (Jan. 4, 2010), http:// intransit.blogs.nytimes.com/ 2010/01/04/travel-apps-20/ (discussing how augmented reality is “the hot new thing” for smartphones, and specifically the targeting of augmented reality travel apps to everyday travelers), and Jane L. Levere, Penney Sells Back-to-School Clothes the Digital Way, N.Y. Times (Aug. 2, 2010), http://nyti.ms/y41xjU (detailing the use of augmented reality to sell clothes to teenage girls). 18   See Evelyn M. Rusli, Google’s Big Bet on the Mobile Future, N.Y. Times: Dealb%k (Aug. 15, 2011), http://nyti.ms/z2PB9d (“Google made a $12.5 billion bet on Monday that its future—and the future of big Internet companies—lies in mobile computing, and moved aggressively to take on its arch rival Apple in the mobile market.”). 19   See Gross, supra note 11. 20   See Chris Crum, Is Augmented Reality the Future of E-Commerce? eBay Lets You Virtually Try On Sunglasses, Webpronews/Tech. (Jan. 7, 2011), http://www.webpronews.com/is-augmentedreality-the-future-of-e-commerce-2011-01 (“Augmented reality could potentially be the biggest thing in e-commerce since the search engine.”). 21   See Jerry Kang & Dana Cuff, Pervasive Computing: Embedding the Public Sphere, 62 Wash. & Lee L. Rev. 93, 98 n.10, 99 (2005) (“What we can expect [if active RFIDs stood in our shoes], then, are networks of miniaturized, wirelessly interconnected, sensing, processing, and actuating computing elements kneaded into the physical world.”).

BARFIELD PRINT.indd 106

13/11/2018 11:52

Mixed reality  107 realspace has preceded PerC by at least twenty years.22 While implantation of chips into people and the environment is still in its infancy, Yelp, Google Latitude, Hidden Park, Parallel Kingdom, and other mixed reality applications are already here, and from these applications will come the next wave of great internet successes.23 Thus, the PerC literature assumes a legal environment based on the prevalence of physical objects that transfer information embedded throughout our everyday lives, when the reality is that technology has developed in a different way, and much more quickly. This reality—that of Mixed Reality here and now—has not been addressed in the legal literature discussing PerC. While the World Wide Web revolutionized human knowledge by linking it together, indexing it, and making it searchable, today a far greater revolution is underway: the real world itself is becoming hyperlinked and indexed. One example of hyperlinking the real world are quick response codes (“QR-codes”)—tags in the real world that can either link to a website or contain information about the location or object near which they are found. Once a smartphone recognizes the QR-code, it provides the user with access to websites, free e-books, streaming videos, or even three-dimensional (“3-D”) overlays onto the physical reality perceived by the smartphone user through the device. Thus, the Mixed Reality revolution is already happening. The real world is already alive and crawling with attached data. Data is routinely attached to real-world people, places, and things, and mobile devices permit users to experience this local data in the place to which it is attached. Well in advance of the advent of the pervasive computing world, data tagging has already hyperlinked and virtualized the real world—our world today. The remainder of this chapter proceeds in four Parts. Part II explores and defines Mixed Reality technologies and demonstrates the gaps in the legal literature on virtual worlds and the legal literature on pervasive computing. Part III analyzes the legal implications of the ongoing extension of virtual governance regimes into realspace and projects future trends. Part III also anticipates several legal problems including a Mixed Reality land rush24 similar to the domain name rush of the late 1990s,25 the advent of new forms of cyberdefamation or reputation poisoning, and dignitary harms based on false information propagated through Mixed Reality applications.26 Part IV modestly 22   See Nancy J. King, When Mobile Phones Are RFID-Equipped—Finding E.U.-U.S. Solutions To Protect Consumer Privacy and Facilitate Mobile Commerce, 15 Mich. Telecomm. & Tech. L. Rev. 107, 112 (2008) (discussing the use of RFID tags in mobile phones and RFID readers while also discussing location-based services). RFID chips are present but are being used as support tools for mixed reality apps. 23   Id.; see also Lesley Fair, Fed. Trade Comm’n, Information Technology Law Institute 2011: Navigating the New Risks in Mobile Technology, Social Media, Electronic Records and Privacy, 1043 PLI/Pat 417, 483 (Apr.–May 2011) (noting that the FTC convened to explore the emergence of RFID only in 2004). 24   See CHEN, supra note 6, at 20, 31 (discussing the mobile app store as a digital gold rush, which is a strong indication that the Apple App Store is still in its infancy and is the sequel to the dot-com boom). 25   See Jacqueline D. Lipton, Bad Faith in Cyberspace: Grounding Domain Theory in Trademark, Property, and Restitution, 23 Harv. J. L. & Tech. 447, 448–9 (2010) (detailing the origins of the domain name rush and the negatives that ensued such as cybersquatting); see also Anupam Chander, The New, New Property, 81 Tex. L. Rev. 715, 724 (2003) (discussing the domain name “land rush”). 26   See Reit v. Yelp!, Inc., 907 N.Y.S.2d 411 (Sup. Ct. 2010) (discussing claim by dentist for defamation and deceptive acts and practices against Yelp!).

BARFIELD PRINT.indd 107

13/11/2018 11:52

108  Research handbook on the law of virtual and augmented reality proposes that as real and virtual worlds converge, the best available analogy for governing Mixed Reality are the background principles of the common law, not the law of online intellectual property licensing. Part V offers a brief conclusion.

2.  MIXED REALITY This Part places Mixed Reality experiences on a continuum between virtual and real worlds. It first describes the various technologies and techniques for creating Mixed Reality experiences and then offers a more developed taxonomy for describing the various types of experiences that the technology can create. 2.1  The Technology Mixed Reality is exactly what it sounds like—the mixing of “virtual” and “actual” reality.27 The core of Mixed Reality is not new. The central element of Mixed Reality is the tying of data to an anchor in the real world, be it a person, geographic location, or structure. Some early examples of data tying are gossip circles in medieval villages or land records that indicated property ownership. In gossip circles, the act of gossiping “tagged” a person with information about that person. Land records—although much less accessible in medieval times—similarly linked a person to an ownership interest in property. Today the same type of data can be tied to a person or property through virtual technology. For example, circles in Google+28 now give information about a person, much like a medieval gossip circle.29 Mobile applications that list land ownership and property values when the user snaps a photo of a house with a smartphone now link that information to the property.30 In both of these examples, information that had always been tied to an object (a person’s reputation through gossip, or real estate ownership through land records) is now being made available seamlessly through technology. What makes Mixed Reality significant is the scale of this new data-enriched realspace. One new aspect that Mixed Reality introduces is the combination of mobile computers with geotagged data and the extent to which this combination is a part of our everyday lives. Through mobile devices, users see data that is tied to particular places, objects, or 27   See Paul Milgram & Fumio Kishino, Taxonomy of Mixed Reality Visual Displays, E77-D IEICE Transactions on Info. & Sys. 1321, 1322–29 (1994), available at http://ci. nii.ac.jp/ naid/110003209335. 28   See A Quick Look at Google+, GOOGLE, http://www.google.com/+/learnmore/ (last visited Feb. 18, 2012). 29   Google+ profiles show other users whom the profile owner has placed in his “circles” and facilitates the sharing of the owner’s daily life (depending on how frequently he uses Google+). Medieval gossip circles would have similarly indicated whom someone knew, and would have also revealed the goings-on of the person’s life. Thus, both have the same fundamental function; the difference now is that this information is readily accessible through a smartphone, whereas one would have to actually sit in a gossip circle to gain this information. 30   See, e.g., ZipRealty, ZipRealty Real Estate, Apple iTunes, http://itunes.apple.com/ us/app/ ziprealty-real-estate/id340513671 (last updated Jan. 30, 2012).

BARFIELD PRINT.indd 108

13/11/2018 11:52

Mixed reality  109 people that they encounter.31 Smartphone technology and other miniaturized computers permit a more mobile and interactive experience with our surroundings.32 Coupled with the growth in mobile computing is the growth in Mixed Reality applications. Now, a husband who goes shopping can peer through his smartphone camera at a product and immediately see tagged locations of local competing stores with better prices.33 A lost tourist in London can look through her smartphone and see virtual arrows overlaid on top of the real world that guide her to the nearest Underground station.34 A potential Boston bar-crawler can use his cell-phone to examine the virtual tags that other patrons have left behind describing the best drinks served at a given hotspot.35 Parents can install geolocation devices in cars that mentor overzealous teenage drivers and provide parents with instant information about their teen’s driving.36 In short, Mixed Reality takes computing out from behind the desk and into the real world.37 As we do so, what matters is not that computers are everywhere, but that they are with people. Given that a person can carry a smartphone in her pocket and access data that other people have tied to a given location, object, or person, people now move through a world augmented with data tags.38 In the same way that people can click the “like” button on a Facebook comment, they can now click the “like” button for a restaurant, or a colleague, or a neighborhood. Popular apps like Yelp and Foursquare have already turned this practice into a runaway business model. Thus, data tagging—the tying of information to a specific geographical location or realspace anchor—not embedded computing, is driving the virtualization of realspace.39 Data tagging can be done in a number of different ways. Global Positioning System (“GPS”)

31   See In re Implementation of Section 6002(b) of the Omnibus Budget Reconciliation Act of 1993 (Data Traffic Growth), 25 FCC Rcd. 11407, 11412–25 (2010); id. ¶ 4 (“Data traffic has grown significantly, due to the increased adoption of smartphones and data consumption per device.”); see also The State of Mobile Apps, Nielsenwire (June 1, 2010), http://blog.nielsen.com/ nielsenwire/online_mobile/the-state-of-mobile-apps/ (“21% of American wireless subscribers have a smartphone at Q4 2009, up from 19% in the previous quarter and significantly higher than the 14% at the end of 2008.”). 32   See Data Traffic Growth, 25 FCC Rcd. 11407, ¶ 4 (“As of the end of 2008, 90 percent of Americans had a mobile wireless device.”). 33   See, e.g., SHAPE Servs., Barcode Reader, Apple iTunes, http://itunes.apple.com/us/app/ barcode-reader/id340825499 (last updated June 9, 2011). 34   See acrossair, Nearest Tube, Apple iTunes, http://itunes.apple.com/app/nearest-tube/id32​ 2436683 (last updated July 16, 2010). 35   See GoTime, Happy Hours, Apple iTunes, http://itunes.apple.com/us/app/happy-hours/ id303814652 (last updated Oct. 17, 2011). 36   See In-Car Teen Mentoring Device, Am. Nat’l Prop. & Cas. Co., http://www.anpac.com/ DriveSmart/WhatIsDriveSmart/Mentoring/default.aspx (last visited Feb. 18, 2012). 37   See Data Traffic Growth, 25 FCC Rcd. 11407, ¶ 4 (“Data traffic has grown significantly, due to the increased adoption of smartphones and data consumption per device.”). 38   See CHEN, supra note 6, at 4 (“Data has become so intimately woven into our lives that it’s enhancing the way we engage with physical reality.”); see also, e.g., SHAPE Servs., supra note 33 (describing the Barcode Reader app as permitting the user to instantly compare prices by scanning items in a physical store). 39   See Kang & Cuff, supra note 21; see also Mark Weiser, The Computer for the Twenty-First Century, 265 Sci. Am. 94, 104 (1991) (“Already computers in light switches, thermostats, stereos and ovens help to activate the world. These machines and more will be interconnected in a ­ubiquitous

BARFIELD PRINT.indd 109

13/11/2018 11:52

110  Research handbook on the law of virtual and augmented reality tagging, other Location Based Services, tagging through what will be called Identification Services (“IDS”), mobile tagging, and Near Field Communication (“NFC”) are all forms of data tagging. Each of these data tagging methods add to Mixed Reality in a unique way. By far the most common data tagging method is GPS data tagging (colloquially, “geotagging”).40 A GPS-enabled smartphone knows to overlay a given reputation bar on top of the local eatery because a global positioning system has identified the location of the restaurant.41 When the smartphone knows its longitude and latitude, it can display information relevant to that location. A simple example and common application of GPS tagging comes from the world of outdoor hiking.42 Geocaching has become an international phenomenon.43 A geocacher is a hiker who hides a small object for other hikers to find by using a GPS tag left by the original geocacher. With their smartphones or GPS device, hikers find these objects in the real world by relying on the data that is tagged to the physical location of the geocache. They can then log their visits online with the rest of the tech-savvy hikers that located the geocache before. Geocaching takes a simple activity like hiking, augments it with data tagged to real-world physical locations, and transforms it into a hidden world of treasure hunting.44 Due in part to the success of GPS, it has been coupled recently with another type of data tagging, sometimes called Identification Services.45 The difference between GPS and IDS is that, while GPS relies on GPS coordinates, IDS relies on some visual or audio cue within the local environment, be it a corporate logo, a face (for facial recognition software), or even a fragment of a musical tune (as in the case of the popular Shazam music-identification app). The real and virtual worlds connect through the smartphone’s lens or audio pickup.46 This requires the user to be in front of the real world cue. For network.”). This literature is either behind the times, or has managed to identify the new trend, that data tagging is driving the virtualization of realspace. 40   See Ian Austen, Pictures, with Map and Pushpin Included, N.Y. Times (Nov. 2, 2006), http:// www.nytimes.com/2006/11/02/technology/02basics.html (defining geotagging in the photography context as a technology “which, broadly speaking, is the practice of posting photos online that are linked to Web-based maps, showing just where in the world the shutter was pressed”); see also Andrew Adam Newman, Appearing Virtually at a Store Near You . . ., N.Y. Times, Jan. 19, 2011, at B9. 41   See AB InBev, Stella Artois—Le Bar Guide, Apple iTunes, http://itunes.apple.com/us/app/ id335624129 (last updated June 29, 2010). 42   See Groundspeak, Inc., Geocaching—The Official Global GPS Cache Hunt Site, GEOCHACHING, http://www.geocaching.com/ (last visited Feb. 18, 2012) (detailing the user base at “1,648,021 active geocaches and over 5 million geocachers worldwide”). 43   See Mark Couhig, Geocaching Is Catching On, Sequim Gazette (Dec. 15, 2010), http:// www.sequimgazette.com/news/article.exm/2010-12-15_geocaching_is_catching_on. 44   See Groundspeak, Inc., supra note 42; see also bulpadok, The Hidden Park, Apple iTunes, http://itunes.apple.com/us/app/the-hidden-park/id314518306 (last updated Mar. 5, 2010). 45   See Bryan Pardo, Finding Structure in Audio for Music Information Retrieval, 23 Signal Processing Mag., May 2006, at 126, 127 (referring to Shazam and similar products as “identification services”); Chris Crum, Augmented Reality + Location = The Holy Grail for Marketers?, Webpronews (Feb. 28, 2011), http://www.webpronews.com/augmented-reality-plus-locationthe-holy-grail-for-marketers-2011-02 (discussing how adding a visual element to GPS-based services makes consumer engagement much stronger than simple GPS-based applications). 46   See Shazam, http://www.shazam.com (last visited Dec. 8, 2011) (featuring a music identification service).

BARFIELD PRINT.indd 110

13/11/2018 11:52

Mixed reality  111 instance, a user must be in front of a Starbucks and view the Starbucks logo through the camera lens before the application will “check in” using an IDS. Mobile tagging also hyperlinks reality. While closely related to IDS it relies on barcodes or other machine-readable codes to retrieve virtual information. Mobile tagging, such as QR-codes, is more involved than IDS because information is not tagged to a visual or audio cue but rather embedded within the barcode or image itself. The mobile tag contains the code that creates the virtual experience, whereas IDS merely identifies the point where information is tagged. Mobile tag data can be a web address, a connection to a wireless network, a free e-book, a Sudoku puzzle, or even an animated graphic of a tank bursting through the wall.47 The most prevalent examples of mobile tagging are QR-codes that act as a link to an online web presence, but there are numerous other applications.48 Near Field Communication49 is yet another form of data tagging. With NFC, the application is closer to the vision of the pervasive computing literature. NFC relies on computer chips embedded in the environment or objects that are able to communicate information to one another via extremely short range radio fields (separated by mere meters). An example of NFC technology might be an application that allows a mobile device to function as a credit card and that could be waved near a receptor in a store in order to be “swiped.” However, because NFC requires two sets of embedded chips in order to function, it is a far less commonly used method of geolocating data than GPS, IDS, or mobile tagging. 2.2  A Taxonomy of Experiences along the Reality–Virtuality Continuum As will be seen, infra, understanding where a given communication lies on the continuum between virtual and real will be of some help in understanding what legal analogy should apply to a given phenomenon. This Section provides a practical set of terms for discussing the technology. As a baseline, the chapter uses Milgram’s Reality-Virtuality continuum (“RV continuum”) as one potentially useful scheme for measuring out the steps from the virtual to the real.50 As can be seen below, the term “Mixed Reality” sometimes refers to the various experiences between fully virtual and fully real. This Section therefore builds out the continuum to provide a more complete picture of the range of experiences offered by virtualization technologies and explains how Mixed Reality—as used in this chapter— is used much more narrowly than the broad concept of Mixed Reality as a description of experiences on the RV continuum, as seen in Figure 4.1.

47   See Andy Vuong, Wanna Read That QR Code, Get the Smart Phone App, Denver Post (Apr. 18, 2011), http://www.denverpost.com/business/ci_17868932. 48   Other types of two-dimensional barcodes developed include DataMatrix, Cool-DataMatrix, Aztec, Upcode, Trillcode, Quickmark, Shotcode, mCode, Beetagg, and Microsoft’s new Microsoft tag. 49   See About NFC, NFC Forum, http://www.nfc-forum.org/aboutnfc/ (last visited Aug. 15, 2011). 50   See Fumio Kishino et al., Augmented Reality: A Class of Displays on the Reality-Virtuality Continuum, 2351 Proc. Spie 282, 283 (1994), available at http://spiedigitallibrary.org/proceedings/ resource/2/psisdg/2351/1/282_1 (identifying Milgram’s Reality-Virtuality Continuum).

BARFIELD PRINT.indd 111

13/11/2018 11:52

112  Research handbook on the law of virtual and augmented reality Mixed Reality (MR)

Real Environment

Augmented Reality (AR)

Augmented Virtuality (AV)

Virtual Environment

Figure 4.1  Milgram’s Reality-Virtuality (RV) Continuum51 Mixed Reality occupies the space between virtual worlds and realspace.52 Like any inbetween technology, Mixed Reality is defined both by what it is and what it is not. The first problem in defining mixed reality is how to keep a definition from spilling over into generalized network technologies. Overuse of the word “virtual” exacerbates this challenge.53 Virtual has come to mean anything electronic.54 Thus, without care, it is easy to expand the definition of Mixed Reality to include almost all data used by people in the real world—that is, all data. A definition that broad is unlikely to be of much use. A more accurate definition characterizes a virtual object or experience as a digital representation of something that we would typically expect to find in the real world.55 Mixed Reality then re-injects or repositions that virtual object back into our real world experience. For example, consider a table. One can build a virtual table in a video game or virtual world, but it does not appear in the real world. But with Mixed Reality technologies, one can experience a virtual table in the real world; one can see an image of it manifested in the real world through mobile computing technologies and perhaps decorate it with virtual flowers as well. A table is an overly simple example but the point remains clear: Mixed Reality involves the injection of virtual places, objects, experiences, or other data into real-world contexts. The second problem with defining Mixed Reality is how to locate Mixed Reality in the range of technologies from virtual worlds to pervasive computing. Over-breadth is again a real risk. Simply defining Mixed Reality as any application of mobile or pervasive computing, when coupled with the “virtual” fallacy above, would mean that one might classify nearly any mobile phone app as a Mixed Reality experience. In fact, the term means something quite specific: it means the projection of virtual objects and experiences into our physical lives.56 51

  Milgram & Kishino, supra note 27, at 1321.   See Robin Fretwell Wilson, Sex Play in Virtual Worlds, 66 Wash. & Lee L. Rev. 1127, 1131–32 (2009) (“These ‘augmented reality’ technologies push virtual experiences and object down into real space, erasing the boundary between the virtual world and the real world.”); see also Marc Jonathan Blitz, The Freedom of 3D Thought: The First Amendment in Virtual Reality, 30 Cardozo L. Rev. 1141, 1144 (2008) (noting that engineers “erase the perceptual barriers” with mixed reality by “mak[ing] illusory three-dimensional people and objects spring up in the more familiar settings in front of us”). 53   See M. Scott Boone, Ubiquitous Computing, Virtual Worlds, and the Displacement of Property Rights, 4 I/S: J.L. & Pol’y Info. Soc’y 91, 108–9 (2008). 54   Id. 55   See Milgram & Kishino, supra note 27, at 1324–25; see also Boone, supra note 53, at 109. 56   See Milgram & Kishino, supra note 27, at 1322 (“[T]he most straightforward way to view a Mixed Reality environment, therefore, is one in which real world and virtual world objects are presented together within a single display . . . .”). 51 52

BARFIELD PRINT.indd 112

13/11/2018 11:52

Mixed reality  113 Mixed Reality and PerC fundamentally differ with respect to where data is stored and processed.57 Pervasive computing implies that the processing power is embedded in objects all around the user. Instead of tagging data on the cloud to virtual points in the real world, PerC stores and processes this information in computers in the physical environment around you. Unlike PerC, Mixed Reality utilizes data stored on the cloud globally but accessed locally; it is tied to real places, people, or objects through wireless connectivity and location based services. Mixed Reality and cloud computing go hand-inhand.58 With these definitions in mind, the following Sections explore the broader range of experiences that virtualization technologies offer and attempt to locate Mixed Reality technologies within that spectrum. 2.2.1  Virtual reality One end of the virtualization spectrum is marked by pure virtual reality. Virtual reality is virtualization at its most profound because the goal is to immerse the user in a virtual environment as completely as possible.59 Virtual reality is the “goggles and gloves” technology that attempts to capture every sensation possible.60 Due to bandwidth and processor constraints, as well as the required gear that tends to be expensive, cumbersome, and complex, the technology has not progressed far past the experimental stage or the occasional appearance in movies like Tron, The Matrix, or Lawnmower Man.61 The technology is visually interesting and full virtual reality is often the first thing to spring to the layman’s mind when contemplating virtual experiences. Yet full goggles-and-gloves reality does not capture the current flowering of virtual experiences—known in somewhat passé technological parlance as “Web 2.0.” The reason is simple: the current digital revolution is social, not technological.62 Virtual experiences matter because they are shared,63 not because they utilize exceptionally rendered 3-D

57   See Eric Taub, Storing Your Files Inside the Cloud, N.Y. Times (Mar. 2, 2011), http://www. nytimes.com/2011/03/03/technology/personaltech/03basics.html (“Cloud backups are appealing for another reason: as computing becomes more mobile—on laptops, tablets and smartphones—you need to have reliable access to the data anywhere over an Internet connection.”). 58   See id; see also Edward Lee, Warming Up to User-Generated Content, 2008 U. Ill. L. Rev. 1459, 1500–501 (discussing that cloud computing is a major component of Web 3.0 in which the Internet converts traditional desktop-based applications into web-based applications that run off of massive amounts of data on remote servers). 59   See Milgram & Kishino, supra note 27, at 1321 (“The conventionally held view of a Virtual Reality . . . environment is one in which the participant-observer is totally immersed in, and able to interact with, a completely synthetic world.”). 60   See Jonathon W. Penney, Privacy and the New Virtualism, 10 Yale J.L. & Tech. 194, 220 (2008) (discussing virtual reality and the two-way interaction between virtual technology and the user). 61   See id. at 220 (“The amount of information processing power necessary for such seamless interaction has not been developed and might never be.”). 62   See Jacqueline D. Lipton, Mapping Online Privacy, 104 NW. U. L. Rev. 477, 480–81 (2010) (referencing the social nature of Web 2.0, for example the expanded options for people to magnify their voice through blogs, wikis, social networks, and MMOGs). 63   See Benjamin Duranske, Virtual Law: Navigating the Legal Landscape of Virtual Worlds 12 (2008) (“Most people who enter virtual worlds do so to interact with other users. This makes virtual worlds highly social spaces . . . .”).

BARFIELD PRINT.indd 113

13/11/2018 11:52

114  Research handbook on the law of virtual and augmented reality computer graphics.64 There exists a sweet spot where the technology is simple enough to be widely adopted, yet complex enough to offer a compelling virtual experience.65 This explains why smartphones—not virtual reality goggles and gloves—are the current carriers of Mixed Reality experiences. The reality is that more people will have these shared experiences if they are enabled by readily available and relatively inexpensive mobile devices. Shared experiences, not completely immersive experiences, are driving the current push into the most successful mobile apps.66 Consumers have clearly indicated that they seek socially rich virtual experiences. But mobile computing depends on smaller computers.67 Thus, virtual worlds have become simpler, isometric social spaces that can be effectively accessed through a smartphone, rather than the fully-rendered immersive 3-D spaces that require bleeding edge (and large) computers.68 Mobile means smaller, and smaller means more social and less graphically intensive.69 Thus, in considering virtual worlds technologies in the following Section, this analysis includes those highly social but graphically simple virtual worlds that are based on mobile and browser technology. 2.2.2  Virtual worlds A virtual world is a persistent, interactive, avatar-mediated, simulated 3-D space (much like virtual reality), but with more social and fewer immersive features than pure virtual reality.70 Virtual worlds facilitate social interaction and enrich it with a shared graphical context.71 At the nexus between social networking technology and 3-D game environments, virtual worlds follow a very different aesthetic from virtual reality. Virtual worlds do not place the user directly “inside” a virtual environment. Rather, virtual worlds often make use of avatars—characters viewed in the third person that represent the players in the virtual world. Avatars permit an increased range of interaction within the world by

64   See Edward Castronova, Virtual Worlds: A First-Hand Account of Market and Society on the Cyberian Frontier 6 (CESifo, Working Paper Series No. 618, 2001), available at http:// papers.ssrn. com/sol3/papers.cfm?abstract_id=294828 (noting that successful virtual worlds combine 3D computer graphics with “chat-based social interaction systems”). The graphics in the games mentioned are not part of complex, full-immersion virtual reality experience, but rather they are experienced merely through one’s computer. 65   To see how simple graphics interfaces in highly social games can be more appealing than less socially-oriented games with high end graphics, compare Zynga, Farmville, Facebook, http://www.facebook.com/FarmVille (last visited Aug. 3, 2011) (identifying 34,070,983 monthly active users), with Press Release, Blizzard Entm’t, World of Warcraft Subscriber Base Reaches 12 Million Worldwide (Oct. 7, 2010), http://us.blizzard.com/en-us/company/ press/pressreleases. html?id=2847881 (stating a 12 million subscriber population). 66   See The State of Mobile Apps, supra note 31 (highlighting that Facebook is the most popular app on the iPhone and BlackBerry, and the second most popular on the Android platform). 67   See Reza B’far, Mobile Computing Principles: Designing and Developing Mobile Applications with UML and XML 5, 12–13 (1st ed. 2005) (describing the evolution of mobile computing). 68   See Zynga, supra note 65 (describing a popular virtual world game, Farmville, which is browser-based and does not require high-performance graphics hardware and computing power). 69   See B’far, supra note 67, at 6, 13. 70   See Castronova, supra note 64, at 5–6; see also Duranske, supra note 63, at 2. 71   See Duranske, supra note 63, at 12.

BARFIELD PRINT.indd 114

13/11/2018 11:52

Mixed reality  115 indicating a user’s focus of attention and that of other players.72 Judge Posner illustrated the unique ability for avatars to convey user focus when he conducted an interview in the virtual world of Second Life entirely through his virtual avatar.73 Although graphically engaging and immersive, virtual worlds continue to utilize the lower end of graphics capabilities in order to capture as many users as possible. Some virtual worlds remain graphically rich and only run on high-end computers, but the number of players in such worlds has been rapidly outstripped by browser-based games, such as those running on Adobe’s Flash platform. Thus tension exists between the immersiveness of the environment and accessibility to large numbers of users, many of whom may not have high-end computers.74 The trend towards ease of access has led to the popularity of several flash games, in particular, Zynga Networks’ Cityville and Farmville, which run on Facebook.75 These games have exploded both in the traditional computer setting and in mobile computing with smartphones and tablet PCs. These games demonstrate the principles outlined in the following Sections: they are graphically simple and run over social networks. 2.2.3  Augmented virtuality Augmented virtuality is the point at which realspace begins to enter virtual worlds. Virtual worlds—such as Second Life—include the capacity to import real events into the virtual space for virtual world denizens to view. A real-world presidential debate can be imported and streamed live to those virtual world denizens.76 Avatars can sit in an auditorium within an entirely virtual environment and watch events unfolding in the real world. Google Maps and Microsoft’s Bing Maps provide other examples. Both now include a “drill down to reality” function.77 Whereas before a Google Maps or Bing Maps user 72   See Penney, supra note 60, at 221 (detailing that avatars are not only a visual representation of the user in the virtual world where the user has full control over the avatar’s appearance and actions, but often, the avatar becomes the person in the virtual world). An example is a recent lecture given by Professor Lastowka and this author at the Governance in Virtual Worlds Conference at the ASU Sandra Day O’Connor Law School. Participants attended with their own avatars and were able to interact with the professors’ avatars. The avatars served as a means to convey social information: social focus, gaze, proximity, experience, and engagement are all conveyed via the avatar. Although avatar-mediated discourse is not as immediate as person-to-person conversation, the use of avatars as markers for social discourse permits a feeling of increased social engagement despite the limitations of the virtual environment. 73   See UChicagoLaw, Judge Posner (or at Least His Avatar) Talks to Second Life, U. Chi L. Sch.: Faculty Blog (Nov. 29, 2006), http://uchicagolaw.typepad.com/faculty/2006/11/ judge_posner_or.html. 74   See Joshua A.T. Fairfield, The Magic Circle, 11 Vand. J. Ent. & Tech. L. 823, 838 (2009). 75   See Douglas Macmillan, Zynga and Facebook. It’s Complicated, Bloomberg Businessweek (Apr. 22, 2010), http://www.businessweek.com/magazine/content/10_18/b4176047938855.htm (detailing the close relationship between Zynga and Facebook). 76   See Presidential Debate Festivities in Second Life, Game Pol. (Sep. 29, 2008), http:// www. gamepolitics.com/2008/09/29/presidential-debate-festivities-second-life. 77   See John D. Sutter, Bing Wows Crowd with Live-Video Maps, CNN Scitechblog (Feb. 12, 2010), http://scitech.blogs.cnn.com/2010/02/12/bing-wows-crowd-with-live-video-maps/ (­discussing the live-feed feature in Bing Maps); Viewing Layers—Maps Help, Google, http://support.google. com/maps/bin/answer.py?hl=en&answer=144359 (last visited Feb. 18, 2012) (stating that one of Google Map’s viewing layers contains live images from webcams around the world).

BARFIELD PRINT.indd 115

13/11/2018 11:52

116  Research handbook on the law of virtual and augmented reality might have ended her journey with a real world photograph of her destination (taken by the Google Streetview cars or geographically tagged photos taken by passers-by), the currently evolving functionality augments the virtual world with a drill down to a live camera view. Thus the drill down of a motorist using Bing Maps might be to a traffic camera, or the drill down of a remote viewer might be to a handheld camera that is currently active in the location. One example of this technology is the subject of Blaise Agüera y Arcas’s TED talk, in which he demonstrated the ability to drill down all the way from a virtual world into realspace real-time live handheld cameras.78 These technologies permit users of virtual worlds access to the real world. In so doing, they begin to mix even more reality into the virtual environment. 2.2.4  Mixed/Augmented Reality This leads to the narrow definition of Mixed Reality within the RV continuum. Although the term “Mixed Reality” can broadly encompass all stages of data-enriched real or virtual environments (i.e., the entire RV continuum), for purposes of this chapter Mixed Reality represents a narrow point in the spectrum where near-field technology,79 geolocation services,80 identification services, mobile tagging, and other data tagging techniques enrich the real world with virtual data through the use of technology. This technology is also sometimes called “Augmented Reality.” The November 2009 issue of Esquire Magazine contains one example of Mixed Reality. It contained Mixed Reality tags such that the magazine cover and several internal advertisements contained virtual elements that only appeared when the magazine was viewed through a computer or smartphone camera.81 Another Mixed Reality application called Hidden Park permits children to see fantastic dragons, trolls, and fairies when specific areas of the park are viewed through a smartphone.82 A child might look at a tree through the smartphone camera and see a goblin face peering out, or a child might look over a field to see elves dancing. Beyond opening up fantastical opportunities for play, these Mixed Reality apps have tremendous potential as educational tools.83 The combination of an app like wikitude84—one of the most popular augmented reality applications—and Hidden Park may fulfill a goal of many parents and teachers: getting 78   See Blaise Aguera y Arcas Demos Augmented-Reality Maps, Ted (Feb. 2010), http:// www. ted.com/talks/blaise_aguera.html (discussing the work of Microsoft with Bing Maps, the integration of cartography, imagery, and user content to augment realspace). 79   See King, supra note 22, at 211 (discussing Near Field Communications (“NFC”) technologies and how they rely on RFID chips in mobile handsets, the software on the mobile handsets, and how NFC will deliver mobile advertising and other location-based services). 80   See Wendy A. Adams, Intellectual Property Infringement in Global Networks: The Implications of Protection Ahead of the Curve, 10 Int’l J.L. & Info. Tech. 71, 89 (2002) (identifying geolocation services as referential databases that are arguably inferior to GPS technologies). 81   See Shira Ovide, Esquire Tries Out Digital Reality, Wall St. J., Oct. 29, 2009, at B10. 82   See bulpadok, supra note 44. 83   See Mark Sutton, Soar Valley College: Augmented Reality in the Classroom, Guardian (London) (Dec. 2, 2010), http://www.guardian.co.uk/classroom-innovation/video/soar-valley-college (discussing a professor’s successful effort to interact with underachieving students by using an augmented reality experience to engage students with the solar system). 84   See Mobilizy GmbH, Wikitude, Apple iTunes, http://itunes.apple.com/us/app/wikitude/ id329731243 (last updated June 29, 2011).

BARFIELD PRINT.indd 116

13/11/2018 11:52

Mixed reality  117 kids to enjoy learning. With technology that provides an interactive experience, otherwise boring topics like math or history may become more social, fun, and educational. Children might be more likely to learn math if it involved counting virtual dinosaurs in the park and might be more likely to learn American history with a virtual world overlay of revolutionary-era Boston life.85 This blending of education with fun using technology is not new. The difference now is that mobile computing combined with data tagging and the resulting virtually enriched realspaces are far more social and dynamic than these earlier media. Mixed Reality applications sit at the midpoint of the RV continuum. They are grounded in real objects and space but augment those objects or places with computer-generated data. For example, some greeting cards now contain a virtual enhancement.86 The card includes a code that the sender can customize with an animated message and the receiver can scan it with a cell phone or web-cam and see the cartoon. The computer layers the 3-D representation onto the object. Often identifying which image to display does not even rely on server architecture. Rather, in many cases, the augmented reality object contains a mobile tag that provides sufficient data for the computer to render a three dimensional image.87 In these instances the real world mobile tag provides the data for the virtual application; the mobile tag augments realspace with data that links directly to that physical printed code. 2.2.5 “Reality+” The final point on this continuum is Reality+. I borrow and use the term reality plus— rather than simple reality—because realspace has always been enriched by information ever since the first fisherman told the second which fishing holes were especially good.88 What is worth noting about realspace is that we have always augmented reality with crude data tagging. Maps and charts have served as crude data tagging devices tied to latitude and longitude. The revolution is in the accuracy, availability, and accessibility of such markers89 and of the propagation of information on a global scale. 85   See, e.g., Dinosaur Games, PBS Kids, http://pbskids.org/games/dinosaurs.html (last visited Oct. 15, 2011) (providing dinosaur-themed games for children to learn various skills and subjects). 86   See Webcam Greetings, Hallmark, http://www.hallmark.com/online/webcam-greetings. aspx (last visited Aug. 8, 2011). 87   See id. The example above includes a feature where you can print out a free sample from Hallmark complete with the mobile tag. If you simply hold the card up to a webcam, your image pops to life. 88   See Mark Burdon, Privacy Invasive Geo-mashups: Privacy 2.0 and the Limits of First Generation Information Privacy Laws, 2010 U. Ill. J.L. Tech. & Pol’y 1, 4 n.62 (explaining Fishing Lake Map, an app that provides geotagged updates on fishing holes). 89   See Nick Bilton, Augmented Reality on Your Phone, N.Y. Times Bits (Dec. 20, 2010), http://bits.blogs.nytimes.com/2010/12/20/augmented-reality-on-your-phone/ (identifying, based on a recent report from Forrester research, that augmented reality apps will become an integral, and common, part of using a mobile phone); see also Thomas Husson, Mobile Augmented Reality: Beyond the Hype, a Glimpse into the Mobile Future, Forrester: Thomas Husson’s Blog BLOG (Dec. 20, 2010), http://blogs.forrester.com/thomas_husson/10-12-20 mobile_augmented_reality_ beyond_the_hype_a_glimpse_into_the_mobile_future (stating that while augmented reality is not new, it is moving to mobile platforms). Although augmented reality is currently overhyped due to unrealistic expectations, it is growing rapidly and drivers for growth are in place.

BARFIELD PRINT.indd 117

13/11/2018 11:52

118  Research handbook on the law of virtual and augmented reality It is important to remark on what reality has always shared with information-enriched environments because of how law works. Much of law is a primitive form of augmenting real spaces and objects with data tags. Think about the title recording system for land. Land is not naturally divided into three-acre parcels. An entry in a paper or (increasingly) an electronic database tags specific land as “yours.” Property is, therefore, a form of information-enriched geotagging.90 Like all of law, property is a consensual fiction based on information-enriched reality. 2.3  The Gap between the Legal Literatures of Virtual Worlds and Pervasive Computing Having discussed the technology and terminology of Mixed Reality, the chapter now turns to the two closest legal literatures: virtual worlds and pervasive computing. There is an extensive legal literature on virtual worlds and a less extensive, but still fascinating, legal literature on pervasive computing. This Section examines the gaps within and between these literatures and then demonstrates that a developed legal theory of Mixed Reality fills those gaps. 2.3.1  The legal literature of virtual worlds Legal academics have written several hundred articles focusing on virtual worlds in past years.91 This rich literature has addressed issues including virtual property,92 democracy,93 control over land,94 the use of contracts to govern virtual worlds,95 the impact of policing and surveillance in virtual worlds,96 the taxation of virtual currency,97 and the sales of virtual goods.98 The articles share an intuition that virtual worlds are not only an interesting and novel technology, but that they also represent a compelling example of the law’s development through the common law process.99 As new communities encounter new technologies, they first develop norms, then they develop practices that are adopted by courts and eventually the practices are codified by statute.100 The

 90   See id.; Burdon, supra note 88, at 7–9 (discussing the expanded use of GPS, locationoriented, and function-oriented geo-mashups which overlay information onto a map of the real world). This use of software to tag information, such as a new cycling or running route, is substantially the same as the overlay of property boundaries upon realspace.  91   For a nonexhaustive list of virtual worlds literature, see sources cited supra note 15.  92   See Joshua A.T. Fairfield, Virtual Property, 85 B.U. L. Rev. 1048 (2005).  93   See Beth Simone Noveck, Introduction: The State of Play, 49 N.Y.L. Sch. L. Rev. 1 (2005); see also Jankowich, supra note 15  94   See Jankowich, supra note 15, at 207–8 (discussing the use of licensing by Linden Labs and Sony to control property in virtual worlds).  95   See Fairfield, supra note 8.  96   Joshua A.T. Fairfield, Escape into the Panopticon: Virtual Worlds and the Surveillance Society, 118 Yale L.J. Pocket Part 131 (2009).  97   See Camp, supra note 15.  98   See Michael H. Passman, supra note 15.  99   See Jankowich, supra note 15, at 189 nn.85–86 (discussing open source virtual worlds and the norms generated in them and how this is comparable to the common law process). 100   See Joshua A.T. Fairfield, Castles in the Air: Greg Lastowka’s Virtual Justice, 51 Jurimetrics J. 89, 90 (2010) (“In so doing, Lastowka frees the field of virtual law from niche status and demonstrates that virtual worlds are participating in the core processes of the common law—they

BARFIELD PRINT.indd 118

13/11/2018 11:52

Mixed reality  119 way that communities respond to emerging technologies drives, in significant part, the development of the law.101 However, one notable lack of the otherwise extremely successful virtual worlds literature is the failure to address issues of mobile computing and Mixed Reality.102 There are several reasons for this. First, virtual worlds have been traditionally defined as graphically rich 3-D persistent spaces in which social groups can gather. But a hidden criterion of virtual worlds is that they be both synchronous103 (interaction within the world occurs among users who are logged in continuously and at the same time) and persistent104 (the world exists without the user’s presence). Most Mixed Reality applications do not seem at first blush to map to the synchronous or persistent nature of virtual worlds and they use many fewer processor-intensive graphics. The failure to fully address Mixed Reality leaves a significant gap in the virtual worlds literature. Synchronicity and persistence are, in fact, traits of Mixed Reality experiences, although the “world” that provides the experience in Mixed Reality is the real one. Thus, although the technology itself may not create synchronicity or persistence, Mixed Reality does share these characteristics of a virtual world. This is unsurprising; a virtual world draws its characteristics from the real one.105 But what has eluded commentators to date is that virtualized applications, even though they themselves are asynchronous or impermanent, or do not include avatar-based interaction, in fact do form part of a virtual world experience—one that straddles the divide between the virtual and the real. This chapter broadens the virtual worlds literature by examining a new use of virtualization technologies: the augmentation of the real world with rich sets of virtual objects and data. This immeasurably widens the subject matter. Virtual worlds articles are most often about games. Mixed or augmented reality applications can be games, but they are just as commonly shopping, travel, health, or fitness apps. This results in an enormous number of applications. The offerings in the Apple App Store are growing by leaps and bounds and Mixed Reality applications are among the most popular offerings.106 Services are jurisgenerative spaces. When courts apply law to the new technologies of virtual worlds, they incrementally adapt traditional concepts to a burgeoning technological world. In short, Lastowka demonstrates that virtual law is common law.”). 101   See id. 102   See, e.g., Barfield, supra note 5, at 159–60 (discussing virtually-enriched advertising but failing to address issues surrounding mobile applications of such advertising); Burdon, supra note 88, at 8 (discussing GPS and RFID technologies and their use in mobile phones to record a new wealth of geographic information and turning humans into geographical sensors but not delving deeper into mixed reality); King, supra note 22, at 125–27 (detailing the growth of mobile advertising but only in the context of RFID chipped mobile phones); see also Kang & Cuff, supra note 21, at 109 (discussing augmented reality in the context of embedded computing, not mobile computing, and arguing that augmented realities will occur through pervasive computing). 103   See Mark W. Bell, Toward a Definition of “Virtual Worlds,” 1 Virtual Worlds Res. 1, 2–3 (2008) (requiring synchronous communication in the definition of a virtual world). 104   See Castronova, supra note 64, at 5–6. 105   See Milgram & Kishino, supra note 27, at 1324–25; see also Boone, supra note 53. 106   See Shan Li, Businesses Quickly Adopting Augmented Reality Apps for Consumers, L.A. Times (Oct. 13, 2011), http://lat.ms/GSPDEh (describing the current proliferation of augmented reality apps).

BARFIELD PRINT.indd 119

13/11/2018 11:52

120  Research handbook on the law of virtual and augmented reality like Foursquare and Yelp, which offer game-like rewards for providing information about locations, goods, and services, have transformed nightlife and fine dining.107 Finally, it is possible that Mixed Reality applications may realize certain goals of virtualization technologies before virtual worlds or virtual reality do. For example, although many elements of virtual worlds—including badges, ranks, experience points, and layered fantasy elements—have entered the real world through Mixed Reality, the real world sense of touch has struggled to enter the virtual. Because it is difficult to import touch into virtual worlds, it may be possible to build fantasy worlds on top of real world physicality long before we will import kinetics (i.e., the sensation of touch and balance) into a virtual world. The former is cheap and just beginning to proliferate; the latter is still the stuff of science fiction—imagine the Star Trek holodeck, which provides kinesthetic sensation through force fields. A more practical approach would be to take real objects, such as a blank-faced robot or a moving floor panel (both of which have already been the subject of fascinating demonstrations) and use them as the underlying surface on which to layer augmented reality experiences.108 So the blank-faced robot becomes the kinetic element of a knight or fair maiden; a larger robot provides the kinetic surfaces for a dragon, and in fact whole different landscapes could be layered onto the real one.109 Or, another example: one company has explored a Mixed Reality “floor” consisting of tiles that constantly move under the user’s feet, giving the user the perception that she can keep walking infinitely in any given direction.110 Mechanical malfunctions of such kinetic interfaces will lead to broken limbs. From the legal perspective, one simple reason to care that kinetics will come sooner to augmented reality than previously thought is that kinetics lead to personal injury lawsuits. Even today’s Mixed Reality applications have implications for physical harm and personal injury suits. By integrating virtuality into the real world, Mixed Reality applications create the threat that people will inevitably ignore some of the real-life aspects of Mixed Reality experiences. Consider the recent unsuccessful case against Google for harm to a pedestrian who followed Google Maps’ driving directions and was struck by a car.111 The information provided was accessible on the Internet and the directions came up in other services as well. The cause was the driver and not the directions themselves. Google did not interact with the end user in a one-to-one manner.112 Neither realspace nor virtual reality caused the harm.113 Rather, it was the underlying reality of the car and the road that impacted the pedestrian. While the court rejected the claim in this instance, it is clear 107   See Yelp for Mobile, Yelp, http://www.yelp.com/yelpmobile (last visited Aug. 4, 2011); Foursquare, https://foursquare.com/ (last visited Feb. 18, 2012). 108   See Utsushiomi, U-Tsu-Shi-O-Mi at Asiagraph 2007, Youtube (Oct. 16, 2007), http:// www.youtube.com/watch?v=htkVlCfCV2M (demonstrating virtual reality overlays on robotic substrates); Joseph L. Flatley, CirculaFloor Robot Floor Tiles Keep You Moving in Virtual Reality, Engadget (Feb. 26, 2009), http://www.engadget.com/2009/02/26/circula 109   See Vernor Vinge, Rainbows End (2006) (exploring a possible future in which mixed reality would layer 3D virtual experiences on the real world). 110   See Flatley, supra note 108. 111   See Rosenberg v. Harwood, No. 100916536, 2011 WL 3153314 (D. Utah, May 27, 2011) (granting Google’s motion to dismiss because Google was not negligent). 112   Id. 113   Id.

BARFIELD PRINT.indd 120

13/11/2018 11:52

Mixed reality  121 that Mixed Reality can have very real legal consequences if users sue for physical harm caused while using these Mixed Reality applications. For this reason alone, a developed legal literature of Mixed Reality will have serious salutary effects for law that is currently in the process of being developed in the courts. 2.3.2  Pervasive computing The fewer than five legal articles that have discussed augmented or Mixed Reality have done so in the context of discussions of pervasive or ubiquitous computing.114 The literature on law and PerC is limited to a few articles that share a common definition. PerC envisions the virtualization of realspace through the actual embedding of small computer processors placed ubiquitously throughout the environment.115 Computers would be embedded in walls, floors, ceilings, and toasters. Your refrigerator would automatically update your shopping list, which would be sent to the supermarket for just-in-time delivery to your house. This embedded pervasive computing presence would run constantly and invisibly, providing computing everywhere. Objects would be linked to the network.116 RFID technology would permit objects to communicate with the rest of the embedded and pervasive computing environment to form the “Internet of things.”117 But pervasive computing is developing unevenly.118 Some elements of pervasive computing have advanced rapidly, others not at all, or only slightly.119 The processors that 114   See Kang & Cuff, supra note 21; King, supra note 22; see also Boone, supra note 53, at 104–5 (discussing two traits of pervasive computing: embeddedness and mobility). The first trait has not come to pass but the second characteristic is closer to the mark. 115   See E. Casey Lide, Balancing Benefits and Privacy Concerns of Municipal Broadband Applications, 11 N.Y.U. J. Legis. & Pub. Pol’y 467, 472 (“‘[T]he Internet of Things,’ in which tiny, inexpensive radio transceivers are installed in various everyday items, ‘enabling new forms of communication between people and things, and between things themselves.’”). 116   See Kang & Cuff, supra note 21, at 112 (detailing that PerC is a digital nervous system grafted into the real world space around us, resulting in a networked system). 117   See King, supra note 22, at 109 n.1 (discussing an International Telecommunications Union report on different technologies, in particular RFID chips, that will lead to an “Internet of things”). 118   See Justin M. Schmidt, RFID and Privacy: Living in Perfect Harmony, 34 Rutgers Computer & Tech. L.J. 247, 250–52 (2007) (discussing the disparity between active and passive RFID tags: passive tags are small and cheap but have fewer applications while active tags are significantly larger and more expensive and have different uses). 119   Compare Derek E. Bambauer & Oliver Day, The Hacker’s Aegis, 60 Emory L.J. 1051, 1069–71 (2011) (discussing the use of RFID chips ranging from access cards to buildings, toll payment systems, and passport readers), with Adam Powell, Benchmark Legislation: A Measured Approach in the Fight Against Counterfeit Pharmaceuticals, 61 Hastings L.J. 749, 759–60 (2010) (discussing the use of RFID chips in drugs that are counterfeited and varied benefits such as being written on and the speed of scanning, but also the lack of widespread use due to high and variable costs along with privacy and accuracy concerns), and Boone, supra note 53, at 10 (identifying ubiquitous computing as “still relatively new and still developing” and the mix of terminology which can include “mobile computing”). Boone’s article continues with other terminology that is often included in the ubiquitous computing literature, specifically, “wearable computing, augmented reality” and “near-field communications.” Id. at 101; see also King, supra note 22. RFID chips are largely an industry tool for tracking. See Schmidt, supra note 118. They are present to an extent in mobile computing and in other tools that are consumer driven, for example credit cards, but they act more as a support system for the more widespread and commercially viable mixed reality systems in smartphones. See King, supra note 22.

BARFIELD PRINT.indd 121

13/11/2018 11:52

122  Research handbook on the law of virtual and augmented reality support mobile phones and the growing stream of user data are in the cloud; they are not present ubiquitously in the local environment.120 Actual computer processors are more remote than ever, rather than ubiquitous and embedded all around.121 On the other hand, the ability to richly augment the real world with data has grown quickly.122 Information is accessed locally but stored remotely. The phenomenon of hyperlocal data has gone hand-in-hand with the development of remote cloud computing. Thus, there is a non-trivial gap in the pervasive and ubiquitous computing literature: computer processors must not necessarily be located locally in order to provide rich hyperlocal data.123 Examples are easy to provide: Google Maps augments realspace with significant virtual data and pushes that data out to mobile phones. But the Google Maps data itself is managed and maintained remotely. And this is necessarily so. Massive data sets still require massive amounts of computing power somewhere. The consumer carries the light client program on a mobile device while remote servers perform the heavy computational lifting elsewhere. Pervasive wireless connectivity replaces pervasive computing power. Conversely, the nanotechnology or micro-microprocessing technology envisioned by pervasive computing has not, largely speaking, come to pass.124 Computing surrounds us today because it is mobile and moves with users, not because it is ubiquitous, already waiting for users wherever they may go. The difference between mobile and ubiquitous computing carries non-trivial legal implications. Property law would likely govern a hardware-focused regime like PerC, with the property owner as dominant legal entity.125 An example would be a PerC-enabled mall,

120   See Kevin Werbach, The Network Utility, 60 Duke L.J. 1761, 1812–13 (2011) (“Most experts participating in a 2010 Pew Foundation Future of the Internet Survey expected that within a decade, remote servers would be the primary means of accessing applications and sharing information, rather than local applications.”). Mobile phones also continue to grow and aid in the growth of cloud computing. Id. at 1814. 121   See Konstantinos K. Stylianou, An Evolutionary Study of Cloud Computing Services Privacy Terms, 27 J. Marshall J. Computer & Info. L. 593, 604 (2010) (“Web 2.0 may have made the Internet more interactive, but it is cloud computing that signifies the transition to ubiquitous always-on networking which has the potentials to substitute part of the desktop computer.”). 122   See Croft, supra note 13; Fletcher, supra note 13. 123   See Werbach, supra note 120. 124   See Kang & Cuff, supra note 21, at 109–12 (presenting the idea of computing in the air, walls, and in our sunglasses, but later identifying the augmentation of experiences with realspace with layers of contextually relevant information). But see Jesse Hicks, DARPA’s Next-Gen Wearable Display: Augmented Reality, Holographic Sunglasses, Engadget (Apr. 12, 2011), http:// www.engadget.com/2011/04/12/darpas-next-gen-wearable-display-augmented-reality-holographi/ (reporting that sunglasses are driven by AR technologies, meaning they are not driven by embedded chips). Kang & Cuff envision “software [that] will manage our datasense and constantly seek out and filter information . . . .” Kang & Cuff, supra note 21, at 110; see also Chen, supra note 6, at 20, 35 (writing that “[t]he iPhone took Apple’s core belief—that software is the key ingredient to hardware’s success—and expanded it” and, further, that software is now pervasive). Smartphones are not just one device, but literally hundreds of thousands of things due to apps. Chen, supra note 6, at 9–-10. Software is what has become truly pervasive and not computing; instead, computing has become mobile. 125   See eBay, Inc. v. Bidder’s Edge, Inc., 100 F. Supp. 2d 1058, 1067 (N.D. Cal. 2000) (analyzing an intrusion of hardware under a property regime with the finding that an owner of a computer system has a property right and can exclude others from it); CompuServe Inc.v. Cyber Promotions,

BARFIELD PRINT.indd 122

13/11/2018 11:52

Mixed reality  123 in which RFID chips embedded throughout the mall communicate to the shopper.126 In contrast, intellectual property and attendant licenses govern software-focused regimes like Mixed Reality,127 and the application provider or developer maintains control over the servers. The current mobile computing regime is closer to the latter structure. The remote servers that produce Mixed Reality experiences are owned and controlled by the corporate entities128 that also own the intellectual property rights in the virtual objects or experiences.129 Use of developers’ networks, apps, or programs subjects the users to a license regime that can severely restrict users’ rights.130 In short, the literature on pervasive computing has discussed two distinct topics: the first Mixed Reality,131 and the second nanotechnology and infrastructure.132 The literature has focused overwhelmingly on the latter. As such, the current phenomenon of data tagging and hyperlinking realspace (Mixed Reality) remains under-examined. This chapter fills this significant gap and provides a legal foundation for a world dominated not by PerC, but by Mixed Reality. 2.3.3  Mixed reality: patching the gap In place of the converging trends predicted by PerC—that both processors and the data processed will become hyperlocal—Mixed Reality is characterized by divergent trends in the location and access of data. As data becomes available hyperlocally, information is increasingly processed and maintained globally. Data is and will progressively be accessed locally from remote and distributed networks. Whether nanotechnology or pervasive computing ever takes off is of secondary importance. Currently, it is clear that a migration

Inc., 962 F. Supp. 1015, 1021 (S.D. Ohio 1997) (finding that an owner of a computer system has a possessory interest and that electronic signals are “sufficiently physically tangible to support a trespass cause of action”); see also Richard A. Epstein, Cybertrespass, 70 U. Chi. L. Rev. 73, 79–80 (2003) (identifying servers as physical property thereby allowing for a trespass to chattels as the server can functionally be touched). 126   See generally Kang & Cuff, supra note 21 (discussing the idea of a mall that makes full use of embedded PerC technologies). 127   See generally Bradford L. Smith & Susan O. Mann, Innovation and Intellectual Property Protection in the Software industry: An Emerging Role for Patents?, 71 U. Chi. L. Rev. 241 (2004) (finding a strong link between software, its emergence as a vital part of the U.S. economy, and the protections provided by intellectual property laws). 128   See John Markoff, Data Center’s Power Use Less Than Was Expected, N.Y. Times (July 31, 2011), http://nyti.ms/wilXuU (identifying that Google not only rents servers but also “generally builds custom computer servers for its data centers”). 129   See Google Terms of Service, Google (Apr. 16, 2007), http://www.google.com/accounts/ TOS (outlining the relationship between Google and the user with regards to “Google’s products, software, services, and websites”). Google is overhauling its terms of service as of March 1, 2012. Id. 130   Id.; see also Vernor v. Autodesk, Inc., 621 F.3d 1102, 1111 (9th Cir. 2010) (outlining how a software vendor can phrase its license agreement to avoid characterization of the transaction as a sale). 131   See Kang & Cuff, supra note 21, at 110 (“Preliminary implementations of such augmented reality already exist. For instance, contractors can walk through construction sites with a visor that paints a digital overlay of the approved architectural drawings on the building in progress.”). 132   Id. at 98–99 n.14 (detailing the use of micromotors, and the reliance on nanotechnology, in addition to varying sizes and forms of devices for pervasive computing). Kang and Cuff detail the infrastructure of pervasive computing under the idea of embeddedness, where computers are embedded everywhere and are capable of wireless communications. Id. at 97.

BARFIELD PRINT.indd 123

13/11/2018 11:52

124  Research handbook on the law of virtual and augmented reality to remote computing, with increased reliance on broadband wireless connectivity, is what lies on the computing horizon. Since developers will continue to create tools that permit consumers to use data maintained and processed on the cloud in hyperlocal applications, this chapter focuses on the legal significance and implications of this technology virtualizing realspace.133 This different approach requires attention to different technologies.134 Prior discussions of pervasive computing have focused heavily on RFID technology: short-range radio that will permit objects to interact with the environment. RFID will play a role, as it currently does, but only in a support role for Mixed Reality apps.135 The relevant technology is the expanding reach of mobile telecommunications broadband networks—Evolution Data Optimized (“EVDO” or “3G”) and Long-Term Evolution (“LTE” or “4G”)—that deliver broadband technology to smartphones and tablets.136 This chapter also fills an important gap in the virtual worlds literature. The virtual worlds literature has not addressed Mixed Reality technologies. This is a significant oversight given that browser and mobile delivery are the fastest growing methods of delivery of virtual experiences. Further, the virtual worlds literature has been characterized by a willingness to treat virtual worlds as a separate reality governed by distinct rules separate from realspace (i.e., the rules of intellectual property).137 Mixed Reality necessarily puts an end to that distinction. Virtual worlds cannot be regulated independently from realspace when virtual objects and places increasingly are a part of realspace itself. In the Part that follows, this chapter grapples with some of the problems that Mixed Reality applications raise for law, both broadly and as a matter of specific challenges that will arise within individual legal contexts. In so doing, this chapter highlights that the law that governs virtual worlds—mostly intellectual property and licensing law—increasingly supplants or subverts the legal regimes that traditionally govern everyday life.138 What 133   See also Jamais Cascio, Filtering Reality, How an Emerging Technology Could Threaten Civility, Atlantic Mag. (Nov. 2009), http://www.theatlantic.com/magazine/archive/2009/11/ filtering-reality/7713/ (detailing augmented reality and current apps, such as Layar, that allow users to “see location-specific data superimposed over their surroundings” in addition to upcoming technologies planned by Sony, for example wearable AR devices like sunglasses). 134   See Chen, supra note 6, at 194–99 (detailing different AR technologies, such as smartphones, headwear, eyewear, and sensory specific options that go beyond just visual, such as audio cues from earpieces). 135   See King, supra note 22. 136   See Jeffrey Paul Jarosch, Reassessing Tying Arrangements at the End of AT&T’s iPhone Exclusivity, 2011 Colum. Bus. L. Rev. 297, 330–31 (detailing the growth in wireless networks from 2000 to 2009). 137   See, e.g., Jack M. Balkin, Law and Liberty in Virtual Worlds, in State of Play: Law, Games, and Virtual Worlds 86, 94 (Jack M. Balkin & Beth Simone Noveck eds., 2006) (arguing that real world laws and legal bodies should allow virtual worlds to construct their own standards for internal needs); Edward Castronova, The Right To Play, 49 N.Y.L. Sch. L. Rev. 185, 204 (2005) (arguing for a law of iteration where, for example, virtual economies would be governed by a body of law that was completely separate from real world economies). 138   See Vernor v. Autodesk, Inc., 621 F.3d 1102, 1111–12 (9th Cir. 2010) (finding that the software owner was not the owner and therefore could not sell it to other users); see also Bowers v. Baystate Techs., Inc., 320 F.3d 1317, 1325–26 (Fed. Cir. 2003) (detailing that private parties are able to contract out of the limited ability to reverse engineer software, a fair use under the exemptions of the Copyright Act); Davidson & Associates, Inc. v. Internet Gateway, Inc., 334 F. Supp. 2d 1164,

BARFIELD PRINT.indd 124

13/11/2018 11:52

Mixed reality  125 we once owned, we will in the future only license.139 What was once a simple breach of contract may now be a hacking crime or potential copyright infringement.140 The following Sections chronicle the replacement of legal systems designed to secure citizens’ reputational, property, and privacy interests with intellectual property licenses that endanger all of these interests.141

3.  THE LAW OF MIXED REALITY The technological revolution of Mixed Reality is well underway. Reality is being hyperlinked, data tagged, indexed, and made searchable in the same way that the bulk of human knowledge was made accessible suddenly and surprisingly through the hyperlinked Internet. The coming legal revolution in response to Mixed Reality will both resolve existing legal debates and raise new, and potentially troubling, questions. For example, a developed theory of Mixed Reality finally puts an end to the enduring and erroneous theoretical idea of the Magic Circle, a metaphorical legal boundary that commentators have supposed separates the real world from virtual ones.142 Virtual worlds cannot be deemed legally separate from the real one.143 All virtual worlds are to some extent mixed: they are experienced by real world people, who interject elements of reality into the virtual world.144 The world may be virtual, but the economic, artistic, and even romantic lives of the participants are quite real.145 Although the Magic Circle is now broken, the legal effects of the Mixed Reality revolution are uncertain. Laws that govern the real world will apply to both elements of the Mixed Reality experience: intellectual property and e-commercial contracts will continue 1181 (E.D. Mo. 2004), aff’d sub nom. Davidson & Associates v. Jung, 422 F.3d 630 (8th Cir. 2005) (“The defendants in this case waived their ‘fair use’ right to reverse engineer by agreeing to the licensing agreement.”). 139   See Vernor, 621 F.3d at 1111; see also Mark A. Lemley, Terms of Use, 91 Minn. L. Rev. 459 (2006) (finding that contracts, such as license agreements, “clickwrap” and “browsewrap,” and Terms of Use, have grown in popularity and, critically, are increasingly enforced by courts). 140   See MDY Indus., LLC v. Blizzard Entm’t, Inc., 629 F.3d 928, 939 (9th Cir. 2010) (setting out the contractual terms that limit the scope of a license as a “condition” and all other license terms as “covenants”). A user can still violate a covenant and thereby breach a contractual term. If a user were to violate a condition of a license agreement, copyright would be implicated. See also Digital Millennium Copyright Act (DMCA), Pub. L. No. 105-304, 112 Stat. 2860 (1998) (codified as amended in scattered sections of 17 U.S.C.). The DMCA contains three provisions that create a framework to address circumvention of technological measures that protect copyrighted works. See 17 U.S.C. § 1201(a)(1)–(2), (b)(1) (2006); ProCD, Inc. v. Zeidenberg, 86 F.3d 1447, 1452 (7th Cir. 1996) (describing that UCC § 2-204(1) provides for different formations of contracts, such as a prompt on a computer screen, which can prevent access). ProCD proposed a contract that a buyer would accept by using the software after having an opportunity to read the license at leisure. 141   See MDY, 629 F.3d at 938 (detailing a particular Term of Use and the prohibition of cheats, hacks, or other third party software, essentially requiring fair play). 142   See Castronova, supra note 137, at 200–205. 143   See Fairfield, supra note 74. 144   Id. at 825. 145   See Bragg v. Linden Research, Inc., 487 F. Supp. 2d 593 (E.D. Pa. 2007) (“While the property and the world where it is found are ‘virtual,’ the dispute is real.”).

BARFIELD PRINT.indd 125

13/11/2018 11:52

126  Research handbook on the law of virtual and augmented reality to govern the software and firmware in the devices; real-world property and tort law will continue to govern where users can go and what they can do in the real world. However, the principal issue between these two spheres is the encroachment of IP and e-commercial contracts into realspace via the virtualization of realspace. Will intellectual property law come to govern the extent to which consumers can use their own real and chattel property,146 just as it now governs what applications consumers may use on their own devices? Without a developed theory of Mixed Reality, IP and e-commercial contracts will overtake property and torts in realspace. To supplement these basic conclusions about the current trajectory of the law, the following Sections analyze the biggest areas of shift by category: contract law, tort law, property law, and privacy law. There are of course other important legal shifts that this chapter must necessarily leave out, if only for space reasons. An example would be free speech: a shift will occur here too, when most of human discourse switches from telephone or email to corporate-controlled social networks, like Twitter, Google+, or Facebook.147 However, the chapter presents selected examples that highlight a pattern in the legal shift: an accelerating trend away from private property and consumer choice in a free market,148 and toward corporate hegemonic control over consumers,149 exercised by the threat of intellectual property lawsuits based on mass-market consumer application End User License Agreements (“EULAs”) or website Terms of Use (“TOUs”).150 146   See Bowers v. Baystate Techs., Inc., 320 F.3d 1317, 1323–34 (Fed. Cir. 2003) (evaluating the issues using copyright and patent law); Davidson & Associates, Inc. v. Internet Gateway, Inc., 334 F. Supp. 2d 1164, 1187 (E.D. Mo. 2004), aff’d sub nom. Davidson & Associates v. Jung, 422 F.3d 630 (8th Cir. 2005) (finding in favor of licensor); see also MDY, 629 F.3d at 938–43 (framing the claim of improper use of software in copyright); Lemley, supra note 139, at 460 (placing shrinkwrap, clickwrap, and browsewrap licenses under the umbrella “terms of use” because they all seek to control the extent to which buyers of software, or visitors to a site, can use that software or site). 147   See Miguel Helft, Facebook Wrestles with Free Speech and Civility, N.Y. Times (Dec. 12, 2010), http://www.nytimes.com/2010/12/13/technology/13facebook.html (“‘Facebook has more power in determining who can speak and who can be heard around the globe than any Supreme Court justice, any king or any president,’ said Jeffrey Rosen, a law professor at George Washington University who has written about free speech on the Internet. ‘It is important that Facebook is exercising its power carefully and protecting more speech rather than less.’”); Ashlee Vance & Miguel Helft, Hackers Give Web Companies a Test of Free Speech, N.Y. Times (Dec. 8, 2010), http:// www.nytimes.com/2010/12/09/technology/09net.html (detailing the tension between the high praise received by Twitter and Facebook as outlets of free speech and their corporate aspirations of both because both rely so heavily on advertising). 148   See Chen, supra note 6, at 92 (discussing the negative feedback to Apple’s App Store and its legal agreements with developers). “[I]f a person makes an app for the iPhone, he has to make it Apple’s way or it won’t be offered in the App Store. He has to play by Apple’s strict rules.” Id. “Apple must approve every iPhone app before it goes up for sale in the App Store, and this means that the corporation can regulate and censor content however it wishes.” Id.; see also Lemley, supra note 139, at 470–72 (“The problem is that the shift from property law to contract law takes the job of defining the Web site owner’s rights out of the hands of the law and into the hands of the site owner.”). 149   See Brian X. Chen, Programmer Raises Concerns About Phone-Monitoring Software, N.Y. Times Bits (Dec. 1, 2011), http://bits.blogs.nytimes.com/2011/12/01/programmer-raises-concernsabout-phone-monitoring-software/ (discussing the newly discovered data-collection software, Carrier IQ, that major cellular phone carriers have installed and that collects data such as users’ locations and telephone activity). 150   See Lemley, supra note 139, at 470–72; see also MDY, 629 F.3d at 938–39 (discussing the

BARFIELD PRINT.indd 126

13/11/2018 11:52

Mixed reality  127 3.1 Contract Law: EULAs and Intellectual Property Licensing Will Govern Everyday Life Mixed Reality technologies on mobile computing devices augment our daily lives with rich data. But these technologies also bring the dangerously flawed law of copyright to the real world. The laws governing the licensing of intellectual property were intended to govern intangibles, not the tangible world.151 The basic economic assumption underlying intellectual property law is that expression is costly to produce and cheap to copy.152 This is the foundation of the copyright system.153 Since expressions are so cheap to copy, fewer people would invest time and money in creation. And indeed, many of the early internet fights were about copying.154 The Recording Industry Association of America (“RIAA”) famously sued teens and grandmothers across the nation for copying MP3 files.155 Copying—the driving concern of copyright law—turned out to be a bad paradigm for internet technologies. Internet technologies incentivize users to stream content rather than copy content because it is more expensive to maintain a local copy of a file on your own computer than it is to stream it.156 As a result, the copyright system does little to inform the circumstances surrounding internet technologies. Copyright also serves as a bad paradigm because copyright holders wanted to do far more than restrict their customers from rote copying of copyrighted materials—they wanted control over their customers.157 ability of a company to write a contract such that the buyer owns or licenses the software and further that restrictions can be placed within such a license agreement that would allow a company to later sue for violations of that agreement based on a breach of its terms resulting in a contract breach or even copyright infringement); Lemley, supra note 139, at 466–68 (discussing the rise in enforcement of license agreements and TOSs by courts). 151   Michael Grynberg, The Judicial Role in Trademark Law, 52 B.C. L. Rev. 1283, 1335 (2011) (“Intellectual property gives functional property rights to the creators of intangible goods.”). 152   See Gideon Parchomovsky & Peter Siegelman, Towards an Integrated Theory of Intellectual Property, 88 Va. L. Rev. 1455, 1466–67 (2002) (stating that absent legal protections, competitors would be able to copy expressive works and inventions without incurring the initial costs of production which, in turn, would drive down the market price leaving original authors and inventors without the ability to recover their initial production costs). 153   See Smith & Mann, supra note 127, at 241–42 (discussing the role of IP protection in incentivizing developers to invest in their programming as it prevents copying by protecting the original expression and aids in preventing copies that diminish the return on development costs). 154   See, e.g., Recording Indus. Ass’n of Am., Inc. v. Verizon Internet Servs., Inc., 351 F.3d 1229, 1231 (D.C. Cir. 2003); Recording Indus. Ass’n of Am. v. Diamond Multimedia Sys., Inc., 180 F.3d 1072 (9th Cir. 1999). 155   See John Schwartz, Record Industry May Not Subpoena Online Providers, N.Y. Times (Dec. 19, 2003), http://nyti.ms/GSNpVq. 156   See Kier Thomas, Cloud Computing: The Executive Summary, PC WORLD (Dec. 29, 2010), http://www.pcworld.com/businesscenter/article/215134/cloud_computing_the_ executive_­summary. html (noting that cloud computing is cheaper than maintaining local files for businesses). 157   See Chamberlain Grp., Inc. v. Skylink Techs., Inc., 381 F.3d 1178, 1201 (Fed. Cir. 2004) (discussing The Chamberlain Group’s arguments as seeking to control consumers options); Walter S. Mossberg, Media Companies Go Too Far in Curbing Consumers’ Activities, Wall St. J., Oct. 20, 2005, at B1, available at http://on.wsj.com/yiv0Sv (explaining that DRM comes in several forms, is widely used, and controls not just whether something can be copied, but also whether it can even be accessed, such as with TiVo and a given TV program expiring after a certain period of time).

BARFIELD PRINT.indd 127

13/11/2018 11:52

128  Research handbook on the law of virtual and augmented reality Beyond the right to control copying, copyright holders often asserted the ability to prevent customers from doing business with the copyright holder’s competitors.158 Unfortunately, copyright law has indulged copyright holders in this regard.159 Early case law in the Ninth Circuit (a critical circuit for online technologies) held that copyright could be used to control behavior in an entirely new set of cases to which it had been previously inapplicable.160 According to these cases, merely loading a computer program for a purpose outside of the software license agreement may constitute copyright infringement.161 Thus, while flipping through a book at a bookstore never implicated copyright law (because no copy was being made), accessing the same material on a Kindle or iPad does implicate copyright law due to the Ninth Circuit’s reading of the copyright statute.162 158   Aaron Perzanowski & Jason Schultz, Digital Exhaustion, 58 UCLA L. Rev. 889, 901 (2011) (“Today, device makers and content distributors can easily introduce barriers to compatibility . . .. [S]hifting legal and technological landscapes, marked by the introduction of digital works and technological measures designed to restrict lawful access, have created serious concerns over lock-in.”); see also Chamberlain, 381 F.3d at 1201 (finding that the copyright holder, Chamberlain, sought “to leverage its sales into aftermarket monopolies—a practice that both the antitrust laws . . . and the doctrine of copyright misuse . . . normally prohibit”); Chen, supra note 6, at 6 (discussing the emerging use of app stores by TV makers and car companies such as Ford, “all with the common goal of trapping consumers inside their product lines”). 159   See, e.g., Phillip A. Harris Jr., Mod Chips and Homebrew: A Recipe for Their Continued Use in the Wake of Sony v. Divineo, 9 N.C. J.L. & Tech. 113, 134 (2007) (“Prior to the DMCA, courts took a very liberal view on reverse engineering of video game protections and allowed it for ‘intermediate copying.’ After the creation and implementation of the DMCA, however, courts showed a stricter approach to copying and held that the interest in protecting copyright holders’ security measures is greater than the interest of fair users that may attempt to use the functional components of intellectual property to create new platforms and software” (citation omitted)); Joseph E. Van Tassel, Remote Deletion Technology, License Agreements, and the Distribution of Copyrighted Works, 97 Va. L. Rev. 1223, 1236 (2011) (“Furthermore, [the] balance of intellectual property rights arguably already skews in favor of the copyright holder, so courts should be wary of further curtailment of users’ rights through the use of license agreements.”). 160   See Wall Data, Inc. v. Los Angeles Cnty. Sheriff’s Dep’t, 447 F.3d 769 (9th Cir. 2006) (finding that because defendant was a licensee and not an owner, it therefore infringed the plaintiff’s copyright by copying the software and installing it on multiple computers in violation of the license agreement); Triad Sys. Corp. v. Se. Express Co., 64 F.3d 1330 (9th Cir. 1995); MAI Sys. Corp. v. Peak Computer, Inc., 991 F.2d 511, 519 (9th Cir. 1993) (creating the random access memory (“RAM”) copy doctrine: “[S]ince we find that the copy created in the RAM can be ‘perceived, reproduced, or otherwise communicated,’ we hold that the loading of software into the RAM creates a copy under the Copyright Act.”). 161   The RAM copy doctrine makes it a copyright violation to violate any term of a license agreement where the software is copied into the computer’s RAM. See MDY Indus., LLC v. Blizzard Entm’t, Inc., 629 F.3d 928, 941 (9th Cir. 2010) (“The rationale would be that because the conduct occurs while the player’s computer is copying the software code into RAM in order for it to run, the violation is copyright infringement.”). 162   See MAI, 991 F.2d at 519–20; MDY, 629 F.3d at 941. But see Cartoon Network LP v. CSC Holding, Inc., 536 F.3d 121 (2d Cir. 2008) (finding that an embodiment and a durational requirement needed to be met in order for a data stream to be “fixed”). MAI did not require a durational requirement, only an embodiment. See MAI, 991 F.2d at 519–20 (finding temporary storage on RAM to constitute a copyright violation). Circuits have varied in their adoption of the MAI holding, and the Second Circuit’s decision presented a circuit split concerning, for example, whether streaming data results in a momentary copy. This apparent split caused the parties to Cartoon Network to seek further review. See Kate M. Manuel, Cong. Research

BARFIELD PRINT.indd 128

13/11/2018 11:52

Mixed reality  129 Shifting a simple breach of contract claim into a claim for intellectual property infringement has several immediate effects. Breach of contract generally generates expectation damages. Copyright infringement, however, entails a statutory remedy system as detailed in 17 U.S.C. § 504(c).163 This statutory regime is the mechanism through which the RIAA can seek millions of dollars in damages from teenagers.164 Each download of a separate registered work constitutes a separate infraction.165 This shift alone significantly increases corporate control over consumer behavior. And the shift from the traditional remedy of expectation damages to statutory damages also changes consumers’ incentives to breach abusive and overreaching online contracts. Where the company’s expectation damages from a consumer’s breach of an EULA are vanishingly small, the statutory damages regime of copyright can turn litigation against one’s own customers from a losing strategy to a profit center.166 The shift away from consumer rights and toward corporate control over consumers’ daily lives via EULAs and TOUs is one of the largest unheralded shifts in law of our generation. Threats of copyright infringement suits167 require consumers to comply with a wide range of restrictions utterly unrelated to the making of copies.168 Facebook provides one example of how copyright law significantly shifts the balance of power to producers

Serv., Rl34719, Cartoon Network Lp V. Csc Holdings, Inc.: Remote-Storage Digital Video Recorders And Copyright Law 10 N.90 (2009), available at http://ipmall.info/hosted_resources/ crs/RL34719_090706.pdf (noting that in June 2009 the Supreme Court denied certiorari for review of the apparent circuit split created by the Second Circuit’s ruling). The Supreme Court’s denial of certiorari suggests that the circuit split is not significant enough to justify review. The circuit split has not affected the large bulk of client-server architecture applications, in which there is undoubtedly a copy of the creator’s content made on the local client. 163   See 17 U.S.C. § 504(c) (2010) (creating a framework of awards for infringements of one particular work, instead of multiple copies of one work, no less than $750 and no more than $30,000 based on the court’s determination). Where the infringement was committed willfully, and the copyright owner sustains the burden of proving as such, the damage award jumps to no more than $150,000 but where the infringer is able to demonstrate they were not aware and had no reason to believe they were infringing a copyright, the damage is reduced to no less than $200 at the court’s discretion. Id. 164   See Pamela Samuelson & Tara Wheatland, Statutory Damages in Copyright Law: A Remedy in Need of Reform, 51 Wm. & Mary L. Rev. 439, 441 nn.4 –5 (2009) (“Although Congress intended this designation to apply only in ‘exceptional cases,’ courts have interpreted willfulness so broadly that those who merely should have known their conduct was infringing are often treated as willful infringers.”). There have been several cases in which courts awarded amounts as large as $80,000 per infringed song and a final award as large as $1.92 million, even where the actual damages determined were near $50. Id. at 442– 43 nn.13–14. 165   See § 504(c) (providing for an award of “statutory damages for all infringements involved in the action, with respect to any one work”). 166   See id. (outlining statutory damages for copyright infringement). 167   See Viva R. Moffat, Super-Copyright: Contracts, Preemption, and the Structure of Copyright Policymaking, 41 U.C. Davis L. Rev. 45, 64 (2007) (“Although these terms may rarely be enforced, at least for now, their consistent inclusion and their consistent, but not uniform, language indicates that the lawyers or website developers who are including these terms seek to reserve their rights to bring breach of contract actions (or to send cease-and-desist letters), possibly coupled with copyright infringement claims seeking copyright’s statutory damages.”). 168   See, e.g., MDY Indus., LLC v. Blizzard Entm’t, Inc., 629 F.3d 928, 938 (9th Cir. 2010) (requiring users to use software only in the ways allowed by the agreement).

BARFIELD PRINT.indd 129

13/11/2018 11:52

130  Research handbook on the law of virtual and augmented reality from consumers of internet technology. Facebook asserts a perpetual license in all of its users’ private information.169 Facebook also ferociously limits what users can say. For example, a recent academic conference focusing on the use of internet kill switches in stifling speech was itself ironically stifled when it tried to advertise via a Facebook page because Facebook does not permit use of the term “Internet kill switch.”170 One might not be bothered by such decisions were it not that Facebook surpasses email as the means preferred for communication by many Americans.171 Threats of copyright liability—like the threat by Facebook—attach any time someone purchases software, visits a website, or uses a social media site. As intrusive as copyright licensing is for purely online computing, it is far more so for the next generation of internet technologies—Mixed Reality and mobile computing. Now, there exists the danger that the copyright law dominating online interactions will flow into Mixed Reality and govern its users in realspace. This new breed of online contracts impacts legal regimes across the board because our current system of law permits parties to alter almost any background legal arrangement via consent.172 To enter a digital store a consumer must agree to the store’s terms.173 By remaining on a website, a consumer ostensibly signs a contract. Engaging in online transactions also requires a consumer to agree to the contractual terms and conditions of the site. For example, on smartphones Google Maps is an indispensable application that permits use of the telephone as a GPS device. But Google Maps, like other Mixed Reality apps, also utilizes GPS technology to track,174 collect, package, and resell the real world physical

169   See Statement of Rights and Responsibilities, Facebook, § 2(1), http://www.facebook.com/ terms.php (last updated Apr. 26, 2011) (“[Y]ou specifically give us the following permission, subject to your privacy and application settings: you grant us a non-exclusive, transferable, sub-licensable, royalty-free, worldwide license to use any IP content that you post on or in connection with Facebook (IP License). This IP License ends when you delete your IP content or your account unless your content has been shared with others, and they have not deleted it.”). 170   See CLIP Roundtable: Internet Switch—National Security or Public Repression?, Facebook (Feb. 10, 2011), http://www.facebook.com/event.php?eid=194151337269200 (missing the “kill” in the title after being edited). 171   See Matt Richtel, Email Gets an Instant Makeover, N.Y. Times (Dec. 20, 2010), http://www. nytimes.com/2010/12/21/technology/21email.html (noting how many people in the younger generations prefer other communications media, such as Facebook, to email). 172   See Davidson & Associates, Inc. v. Internet Gateway, Inc., 334 F. Supp. 2d 1164, 1184 (E.D. Mo. 2004), aff’d sub nom. Davidson & Associates v. Jung, 422 F.3d 630 (8th Cir. 2005) (finding consent to the software agreement); see also Fairfield, supra note 74, at 831–35 (noting how games allow individuals to alter even the rules of society with regards to one another through consent). 173   One of the most notable examples of this is the iTunes splash screen, displaying the iTunes terms of service. See Terms and Conditions, Apple, http://www.apple.com/legal/itunes/us/terms. html#SALE (last updated Oct. 12, 2011). 174   There are multiple layers of tracking. Even if app providers did not collect real-world location information, telephone companies themselves record the browsing habits and IP addresses assigned to their smartphone customers, as well as cell-site location information, which is the information on where the telephone user has traveled in real life as indicated by the cell towers that the mobile phone contacts. And even if mobile phone companies did not record this information, internet advertising giants leverage their vast consumer information databases to track wherever their customers browse online, all without anything more than the figleaf of consent of a buried clause in an electronic contract that no consumer ever sees or reads. (And what would consumers

BARFIELD PRINT.indd 130

13/11/2018 11:52

Mixed reality  131 location of smartphone users to a broad array of third parties.175 To use Google Maps, consumers must consent to Google’s terms, which also effectuates consent to the tracking process just mentioned.176 Under the current legal regime, these tracking activities are protected under the law governing online contracts (EULAs and TOUs) because users have given their consent in order to use the services. But this is a poor legal framework for Mixed Reality because it fails to recognize the “reality” aspect of Mixed Reality and, most importantly, how Mixed Reality applications are coming to affect the real world. A comparison may clarify: consider the different reactions that Google’s surveillance of its customers received online and offline. Online, Google retains all of its users’ searches, ties them to specific user profiles, and further ties all users’ online browsing habits (whether tracked by Google cookies on the company’s own sites, or tracked through the Google advertising service on third-party sites, or some combination of the two).177 While the practice is met with some criticism, it is generally accepted that users of their services consent to these practices. But a similar practice of data collection, when injected into the real world, got Google into serious legal trouble. Google Streetview cars accessed individuals’ open home wireless networks as the cars roamed around taking pictures for Google Maps, collecting data from those networks.178 Even though the result was basically the same—collection of user data—the fact that the activity took place in a tangible way made a significant difference in the way the practice was perceived. As a result, lawsuits were filed across the United States, and state attorneys general began to investigate the search giant for possible illegal wiretapping and invasions of communication privacy.179 The real difference in this comparison is not between “online” and “offline” collection of data—after all, the Streetview cars tapped into wifi connections—but whether there was the barest figleaf of contractual consent in place. Google ostensibly secures consent for an enormous amount of intrusive surveillance on its customers as soon as users surf to its web page.180 Streetview cars did not have any such contractual figleaf.

do if they did read it and objected? Use the Internet without Google?) The confluence of these technologies means that all of a user’s activity—online and off—is tracked and recorded. 175   See Julia Angwin & Jennifer Valentino-Devries, Apple, Google Collect User Data, Wall St. J. (Apr. 22, 2011), http://on.wsj.com/zp2Euo (“Google and Apple are gathering location information as part of their race to build massive databases capable of pinpointing people’s locations via their cellphones. These databases could help them tap the $2.9 billion market for location-based services—expected to rise to $8.3 billion in 2014.”). “[S]ome of the most popular smartphone apps use location data and other personal information even more aggressively than this—in some cases sharing it with third-party companies without the user’s consent or knowledge.” Id. 176   See Privacy Center—Privacy Policy, Google, http://www.google.com/intl/en/privacy/ privacy-policy.html (last modified Oct. 20, 2011) (outlining Google’s use of personal information, which includes using location data to “improve” its services for users). 177   See Anne Klinefelter, When To Research Is To Reveal: The Growing Threat to Attorney and Client Confidentiality from Online Tracking, 16 Va. J.L. & Tech. 1, 6–9 (2011) (detailing how web browsing, searching, and online activities in general—including the use of Google—give rise to attorney-client confidentiality concerns due to the data being saved and indexed). 178   See Elec. Privacy Info. Ctr. (EPIC), Investigations of Google Street View, Epic.org, http:// epic.org/privacy/streetview/ (last visited Oct. 31, 2011) (summarizing the various investigations around the world into Google’s practice of collecting data from unencrypted wireless networks). 179   Id. 180   See Google Terms of Service, supra note 129.

BARFIELD PRINT.indd 131

13/11/2018 11:52

132  Research handbook on the law of virtual and augmented reality The fighting question for Mixed Reality applications will be whether such online contracts of adhesion will finally be pushed down into the real world, such that courts will protect intrusions—like those of the Streetview car—under the theory that consumers have consented to the surveillance.181 Google already engages in online surveillance operations gathering data far more comprehensive than any of the data gathered by Streetview, but it is privileged to do so under a strained reading of contract law.182 We have come full circle. The special set of rules that were originally intended to govern intangible, intellectual property now govern the everyday, walkabout lives of U.S. citizens. American citizens do not functionally own their private information and cannot stop the indiscriminate recording of data about their everyday lives short of refusing to use cell phones and the Internet. What is needed is a robust path forward based on existing, established contract, tort, property, and privacy law. A legal regime not muddled by a strained reading of intellectual property law will protect consumers, scale back untrammeled corporate control of consumer information, and return copyright to its original role of protecting copying of creativity, rather than controlling the economic and intimate lives of citizens. The following Sections explore the issues in other areas of law, before offering proposals for re-balancing the law in Part IV. 3.2  Tort Law: Cyberdefamation and Mixed Reality Reputation Systems The Mixed Reality future will include facial recognition software that is able to access reputational ratings of people the user runs across in her everyday life.183 Tagging real people with data raises obvious issues relating to the law of reputational interests, which acts to protect individuals against the publication of false statements made against their image.184 Where employers now Google applicants, in the future they will merely check the person’s online reputation with a range of online reputation providers and social networks. Once personally-tagged reputation and personal information becomes ubiqui-

181   See Elinor Mills, Carrier IQ Faces Lawsuits, Lawmaker Seeks FTC Probe, CNET (Dec. 2, 2011), http://cnet.co/xgWjcX (describing the lawsuit filed against Carrier IQ as performing surveillance without consumer consent). Carrier IQ responded to the criticisms by claiming that it was assisting carriers in gathering data, and at least one carrier stated that its practices of using Carrier IQ did not violate its privacy policy. David Sarno & Tiffany Hsu, Carrier IQ Defends Itself in Furor over Smartphone Users’ Privacy, L.A. Times (Dec. 2, 2011), http://lat.ms/GSNVmp. At the time of this writing, Carrier IQ has not issued a formal response to the suits. 182   See Lemley, supra note 139, at 468–70 (citing ProCD’s questionable legal reasoning based on incomplete reliance upon the UCC, in particular §§ 2-204, -207, and -209, with the subsequent legal reality that these rigid contracts are typically upheld in favor of their corporate authors). 183   See John Biggs, iOS 5 To Have Powerful Face Detection, Techcrunch (July 27, 2011), http://techcrunch.com/2011/07/27/ios-5-to-have-powerful-face-detection/ (reporting on Apple’s purchase of a facial recognition software company, Polar Rose, and the plan to incorporate it into Apple’s iPhone operating system); Ben Parr, Top 6 Augmented Reality Mobile Apps, Mashable (Aug. 19, 2009), http://mashable.com/2009/08/19/augmented-reality-apps/ (describing a mobile app called Augmented ID that recognizes a person’s face and pulls up information about that person). 184   See RESTATEMENT (SECOND) OF TORTS § 569 (1977) (“One who falsely publishes matter defamatory of another in such a manner as to make the publication a libel is subject to liability to the other although no special harm results from the publication.”).

BARFIELD PRINT.indd 132

13/11/2018 11:52

Mixed reality  133 tously available to everyone with a smartphone, the temptation to manipulate or poison that information to cause reputational harms will inevitably arise. Such reputational harms can already be found in the purely online context.185 For example, an early Google bomb—using search engine optimization to prioritize the results of the “bomber” on Google—targeted Senator Rick Santorum.186 Political detractors of the senator created an off-color definition of the senator’s name and then raised the search rank of the alternative result by crosslinking and referencing the neologism.187 The end result was that searches on Google for the senator’s name would find the alternative definition in the first page of results.188 Because the Internet has become the primary purveyor of both personal and professional reputational information, the risk of harm is magnified. Employers Google prospective applicants.189 Social networks like LinkedIn manage professional connections.190 eBay maintains reputation systems for third party vendors, facilitating transactions between parties that otherwise would not trust each other.191 Under the current legal and statutory regime, however, companies that create and maintain reputational networks lack incentive to keep reputational data accurate. This is because § 230 of the Communications Decency Act (“CDA”)192 generally immunizes interactive computing services providers from tort lawsuits stemming from inaccurate data supplied by users of the service.193 Defamation law governs reputational harm offline—where there is no corresponding immunity for providers of reputational information. Thus, online service providers generally avoid liability where offline providers incur liability. This duality makes the existence and proliferation of false or misleading

185   Kai Ma, Dan Savage Threatens To ‘Google Bomb’ Rick Santorum, Yet Again, Time Newsfeed (July 28, 2011), http://newsfeed.time.com/2011/07/28/dan-savage-threatens-to-goog​ le-bomb-rick-santorum-yet-again/. 186   Tom McNichol, Your Message Here, N.Y. Times (Jan. 22, 2004), http://www.nytimes. com/2004/01/22/technology/your-message-here.html (explaining the phenomenon of “google bombing”). 187   See Marziah Karch, Google Bombs Explained, About.com, http://google.about.com/od/ socialtoolsfromgoogle/a/googlebombatcl.htm (last visited Dec. 12, 2011) (noting that Santorum’s name was linked to the definition of a lewd phrase through a Google bomb). 188   See id. 189   See Employers Google Applicants, ABC News (Apr. 28, 2007), http://abcnews.go.com/ Business/video?id=3083837. 190   See What Is LinkedIn?, LinkedIn, http://www.linkedin.com/static?key=what_is_ linkedin&t​ rk=hb_what (last visited Oct. 31, 2011) (describing the professional networking qualities of LinkedIn). 191   See Detailed Seller Ratings, eBay, http://pages.ebay.com/help/feedback/detailed-sellerratings.html (last visited Oct. 31, 2011) (describing eBay’s seller rating system and how it is used to determine seller quality). 192   47 U.S.C. § 230 (2010). 193   See Barnes v. Yahoo!, Inc., 570 F.3d 1096 (9th Cir. 2009); Doe v. MySpace, Inc., 528 F.3d 413 (5th Cir. 2008); Fair Hous. Council v. Roommates.com, LLC, 521 F.3d 1157 (9th Cir. 2008); Chi. Lawyers’ Comm. for Civil Rights v. Craigslist, 519 F.3d 666 (7th Cir. 2008); Mazur v. eBay, No. C 07-03967 MHP, 2008 WL 618988, at *1 (N.D. Cal. Mar. 3, 2008); Doe v. SexSearch.com, 502 F. Supp. 2d 719 (N.D. Ohio 2007) (dismissing claim against online site SexSearch on grounds that plaintiff’s claim based on SexSearch’s promise that all users were over 18 was barred by CDA § 230 when a minor entered false data as to age), aff’d, 551 F.3d 412 (6th Cir. 2008) (affirming on grounds of failure to state a claim, but declining to adopt district court’s reading of CDA § 230).

BARFIELD PRINT.indd 133

13/11/2018 11:52

134  Research handbook on the law of virtual and augmented reality reputational information that much more appealing on the Internet. This Section will explore these problems with online reputation in turn. An example may help to clarify the current state of the law and the problems Mixed Reality raises. Imagine an online dating website that assures users that its users are unmarried and have not committed a felony. User A lies about her marriage status and criminal record. User B dates user A, and is harmed as a result. Let us also assume that the site knowingly or willfully failed to implement measures that would easily have detected A’s falsehoods. On these facts, the caselaw as to the website’s liability is split.194 Section 230 clearly seems to bar any lawsuit based on the false information that user A entered.195 However, courts are split over whether the site can be held liable for its failure to live up to promises regarding the data added by users.196 Some courts seem to lean toward immunity: since the inaccuracy in the site’s representation was caused by the third party’s false data entry, the site would be immune to any lawsuits for failing to remove false data.197 Other interpretive approaches might lean in the opposite direction, reasoning that the claim that the dates were “safe” was itself a representation by the company, not a representation by a third party user of the site.198 This system of online content management leads to very strange incentives. The corporate curator of a reputation network has immunity from suit based on third party false representations of trustworthiness,199 but also directly profits from a high overall reputation average within the network. For example, new apps in the Android app market appear to receive a five-star rating at the outset.200 This rating is then modified by third party reviews of the software. The overall sense that this generates is that Android apps

194   Joshua Dubnow, Ensuring Innovation as the Internet Matures: Competing Interpretations of the Intellectual Property Exception to the Communications Decency Act Immunity, 9 NW. J. Tech. & Intell. Prop. 297, 307 (2010) (“While the courts have reached two competing interpretations of § 230(c)(1) and (e)(2) of the Communications Decency Act, this split must ultimately be resolved because of the vastly different outcomes to which each interpretation leads.”). 195   See § 230(c)(1) (“No provider or user of an interactive computer service shall be treated as the publisher or speaker of any information provided by another information content provider.”). 196   See Barnes, 570 F.3d at 1109 (denying Yahoo! § 230 liability where a Yahoo! associate made a direct promise to remove nude pictures of plaintiff posted by a third party); Fair Hous. Council, 521 F.3d at 1176–77 (denying Roommates.com § 230 immunity where it exercised such control over the statements of the users that it functionally became the source of their illegal housing advertisements); Mazur, 2008 WL 618988, at *14 (denying eBay § 230 immunity where eBay itself made representations about the nature of certain auctions). 197   See SexSearch.com, 502 F. Supp. 2d at 727–28. 198   See David S. Ardia, Free Speech Savior or Shield for Scoundrels: An Empirical Study of Intermediary Immunity Under Section 230 of the Communications Decency Act, 43 Loy. L.A. L. Rev. 373, 397, 411, 479 (2010) (explaining that the first empirical study of § 230 reveals that the statute has been haphazardly applied by courts and has led to mixed—but generally positive—­ outcomes for providers). 199   Id. at 379 (citing § 230(c)(1) and stating that it effectively grants “operators of Web sites and other interactive computer services broad protection from claims based on the speech of third parties”). 200   Android’s application programming interface (“API”) allows developers to use a code feature called RatingBar which allows them to assign a default star rating for an app. See RatingBar, Android Developers, http://developer.android.com/reference/android/widget/RatingBar.html (last updated Oct. 27, 2011).

BARFIELD PRINT.indd 134

13/11/2018 11:52

Mixed reality  135 are high quality and safe, when the reality is that many are merely new. In fact, due to Android’s popularity, dangerous and fraudulent apps are at an all-time high201 and benefit disproportionately from the appearance of trustworthiness that the Android market creates. Legal precedent appears to incentivize the network to make untrue statements about the high level of trustworthiness of the network. This exacerbates the tension between the network’s financial stake in a good reputation and the very point of such a network (to help users detect bad actors). Even if a bad actor’s false inputs into the reputation network render the reputation provider’s statements untrue, there is a high likelihood that network provider will be immune from liability.202 This leads back to the problem of online licensing and increasing control over consumers. At the same time that copyright law has given online service providers unprecedented power over consumers, courts have also granted providers unprecedented immunity against even claims based on the companies’ own promises. Consider a standard online EULA or Terms of Use contract. That contract can impose strict controls on the consumer, on pain of copyright infringement and statutory damages. But the return promises of the company to keep the network safe or to provide accurate reputational information regarding other users of the network may well be largely unenforceable under CDA § 230.203 The advent of Mixed Reality technologies will aggravate this liability imbalance significantly. Again, the core example is mobile technology that can recognize another person and then report reputational data to the user. Facial recognition technology is already being built into mobile devices.204 And such technology does not merely recognize the device’s user, but it can also recognize people in photographs that the user takes. Technologies like Face.com’s facial recognition software are already combing online photo albums and identifying anyone who appears in the photographs.205 Google+’s picture and video uploads are particularly aggressive—if the user is not cautious with permissions, Google+ will automatically upload pictures and movies from the user’s telephone, and all future pictures and videos will be auto-uploaded.206 Facial recognition is a standard mixed reality application, in that it takes indicia from the environment (here, the target face) and augments it with data (here, the person’s reputation). The confluence of facial recognition, reputation, and mobile technologies will push problems of online reputation down to the personal level. Where once an employer had to be at a desk to Google your online

201   See, e.g., Google Moves To Delete ‘RuFraud’ Scam Android Apps, BBC News (Dec. 14, 2011), http://bbc.in/GSOqNl. 202   See Ardia, supra note 198, at 481 (analyzing statistics of decisions under § 230, concluding that “overall, defendants won dismissal in 76% of the cases studied”). 203   Id. at 493 (“[D]efendants won dismissal on section 230 or other grounds in more than threequarters of the cases studied.”). 204  Biggs, supra note 183; Nick Bilton, Facebook Changes Privacy Settings To Enable  Facial Recognition, N.Y. Times Bits (June 7, 2011), http://bits.blogs.nytimes.com/2011/ 06/07/facebook-chan​ ges-privacy-settings-to-enable-facial-recognition/. 205   See Face.com, http://face.com/ (last visited Feb. 19, 2012); see also The Facial Recognition Software That Will Put a Name to Every Photograph in the Internet, Daily Mail Rep. (Aug. 23, 2010), http://www.dailymail.co.uk/sciencetech/article-1305191/Facial-recognition-software-allowability-identify-people-photographs-internet.html. 206   Instant Upload Settings, Google, http://www.google.com/support/mobile/bin/answer.py?​ answer=1304820 (last visited Aug. 14, 2011).

BARFIELD PRINT.indd 135

13/11/2018 11:52

136  Research handbook on the law of virtual and augmented reality reputation or check your social networking sites, now facial recognition will seamlessly integrate the process of online reputation into real life. Without progress in the law, the current legal framework governing reputation networks will replicate the same perverse incentives for Mixed Reality reputation systems that it has generated for online reputation networks. Consumers will bear a disproportionate amount of liability pursuant to the EULAs and TOUs of Mixed Reality applications, while their creators will largely be immune from liability. 3.3  Property Law: The Digital Land Wars Mixed Reality augments real world objects, places, and people with virtual experiences. The augmentation of objects and places necessarily implicates property law. Imagine that someone “augments” your house with a virtual tag that contains an obscene word viewable through a Mixed Reality application. Can you assert rights as a property owner to remove the offensive virtual sign?207 This Section tracks property shifts in response to technology and predicts shifts based on emerging Mixed Reality applications. One perennial feature of the digital landscape is that of the digital land war.208 The cycle goes as follows. First a range of options for the location of information is proposed. There is a divergence, and many different locations, applications, or networks are considered candidates for the “best” piece of internet real estate. There is then convergence once one address, application, or network becomes the “best,” and people shift attention to it. Once attention shifts to only one address, application, or network, legal battles then follow as the owners of pre-existing property rights try to take the prime pieces of internet real estate away from the people who bet on the right technology.209 For example, early in the Internet’s development, there were a few different top-level domain names. Some of these were restricted, like .mil and .edu. Some were general, like .net, .org, and .com. It was not immediately apparent that a .com domain name would become the most valuable piece of land on the Internet. It was only after the cycle of divergence—multiple top-level domain names existed—and convergence—to the .com domain name as the first choice of the searching consumer—that the legal wars over the .com domain names began. When they did, they did so in earnest, with Congress enacting legislation in support of the rights of trademark owners,210 allowing them to take domain names from people who had registered them. The Anticybersquatting Consumer Protection Act permits the owner of a registered trademark to take a domain name that references the mark away from the registering party.211 This permits trademark owners

207   See John C. Havens, Who Owns the Advertising Space in an Augmented Reality World?, Mashable (June 6, 2011), http://mashable.com/2011/06/06/virtual-air-rights-augmented-reality/ (noting that “[m]ultiple apps feature the ability for ads to appear on your mobile screen as miniature virtual billboards assigned to GPS coordinates”). 208   See supra note 25. 209   See Lipton, supra note 25, at 448. 210   See Trademark Act of 1946, Pub. L. No. 79-489, 60 Stat. 427 (codified as amended in scattered sections of 15 U.S.C.). 211   Pub. L. No. 106-113, § 1000(a)(9), 113 Stat. 1501, 1536 (1999) (enacting into law § 3002 of the Intellectual Property and Communications Omnibus Reform Act of 1999, S. 1948, 106th

BARFIELD PRINT.indd 136

13/11/2018 11:52

Mixed reality  137 to wipe out free riding by parties who wish to use the trademark to sell goods, but it also gives the trademark owner another tool to quell critics of the trademark owner, or critics of the goods and services that the owner sells.212 These land wars are far from settled. Another wave of land battles occurred over the use of metadata—mechanisms to drive customers to one site or another via search engine optimization.213 Another modern-day wave is “Twitterjacking.”214 As Twitter became a social networking phenomenon, certain Twitter handles became valuable property. Immediately following the BP Gulf oil spill in 2010, some enterprising individual registered the handle “BPGlobalPR” and began a series of sardonic, self-involved, and hilarious tweets supposedly on behalf of BP.215 It is not immediately clear that BP has the right to any Twitter handle that contains its name, especially ones that are being used for parody, or to convey truthful critical consumer information to the market. The land wars continue in the sphere of Mixed Reality. For example, land wars are currently being waged over geolocated data tags. Yelp, a company that places GPS-located tags on businesses, includes reviews from ostensible customers. Litigation is now pending in New York against Yelp. The plaintiffs’ goal is to force Yelp to remove negative reviews and stop removing positive reviews that are geotagged to the plaintiffs’ businesses.216 Furthermore, some European countries have voiced unrest because of the lack of control over the virtual representations of houses and property in Google Earth and through Google Maps.217 A coming wave of digital land wars will likely involve mirror worlds. Mirror worlds are virtual worlds that mirror the real world. The full 3-D version of Google Earth is a good example. With the latest software, consumers can see 3-D representations of buildings and view real-time relays from cameras at specific locations.218 Fights over data tagged mirror Cong., reprinted at 113 Stat. 1501A-545 to -548 (1999) and codified as amended at 15 U.S.C. § 1125(d)). 212   See Susan Thomas Johnson, Internet Domain Name and Trademark Disputes: Shifting Paradigms in Intellectual Property, 43 Ariz. L. Rev. 465, 476 (2001) (listing various types of cybersquatters, including one “who registers a domain name using the same or a very similar version of another entity’s name to harass or criticize that entity” and “one who intentionally appropriates a famous trademark or tradename as a domain name for financial gain”). 213   See Eric Goldman, Search Engine Bias and the Demise of Search Engine Utopianism, 9 Yale J.L. & Tech. 111 (2006); see also David Segal, The Dirty Little Secrets of Search, N.Y. Times (Feb. 12, 2011), http://www.nytimes.com/2011/02/13/business/13search.html. 214   See William Sloan Coats & Jennifer L. Co, Kaye Scholer LLP, The Right of Publicity & Celebrity Licensing, 1065 PLI/PAT 277, 298 (2011) (describing Twitterjacking as the phenomenon of someone creating a Twitter feed and pretending to be a famous individual). 215   See Noam Cohen, For Dueling BP Feeds on Twitter, Biting Trumps Earnest, N.Y. Times (June 7, 2010), http://www.nytimes.com/2010/06/07/business/media/07link.html. 216   See Reit v. Yelp!, Inc., 907 N.Y.S.2d 411 (Sup. Ct. 2010). 217   See, e.g., Catherine Bolsover, German Foreign Minister Joins Criticism of Google’s Mapping Program, Deutsche Welle (Aug. 14, 2010), http://www.dw.de/dw/article/0,,5910738,00. html (describing complaints against Google Maps in Germany); Google Street View Blacked Out in Greece, CNN (May 13, 2009), http://articles.cnn.com/2009-05-13/world/greece.google.street. view.blocked_1_google-earth-search-giant-google-maps (describing the same in Greece); Call To ‘Shut Down’ Street View, BBC News (Mar. 24, 2009), http://news.bbc.co.uk/2/hi/7959362.stm (describing the same in the United Kingdom). 218   See Google Maps, Google, http://maps.google.com (last visited Oct. 31, 2011) (containing a drop-down menu that enables the viewing of various live camera feeds in areas all over the world).

BARFIELD PRINT.indd 137

13/11/2018 11:52

138  Research handbook on the law of virtual and augmented reality worlds will be intense. Imagine if a global anarchist protest movement grabbed the mirror world location of local Wal-Mart stores and targeted them for mirror world protests by tagging the GPS location with anti-Wal-Mart facts and slogans; or imagine if members of the Occupy Wall Street movement tagged the locations of Wall Street firms with accusations and criticisms.219 Consider the virtual defacing of a political headquarters in lieu of the more traditional brick through the window. The land wars leave open a number of legal questions. The first question is whether owners of intellectual property rights—here, generally trademark owners—should be permitted to take prime internet locations away from first movers. Second, and more significantly, the land wars leave open the question of whether intellectual property law is itself the correct legal framework to apply. The law of intellectual property tends here, as elsewhere, to exacerbate the trend towards increasing corporate control at the expense of protecting individuals. For example, imagine that a user Twitterjacks@Fairfield begins to tweet as this chapter’s author. The author does not have the kind of celebrity that would give rise to a misappropriation of likeness claim, nor trademark or other IP ground on which to assert a claim to ownership of this new internet real estate. Yet Fairfield Inn & Suites would have a reasonable expectation of success in seizing the Twitter designation from a new registrant if someone were to register @Fairfield and begin tweeting hotel deals.220 Thus, while I must register @Fairfield preemptively to protect my online persona, intellectual property owners often have the luxury of waiting to see which emergent technologies become dominant and then moving to secure the most valuable digital real estate. This gives IP holders a significant advantage. Mixed Reality will only intensify the trend towards corporate control. As Mixed Reality causes real and virtual experiences to converge, there is a serious risk that the “virtual” rights holders (IP owners) will prevail and that “real” rights holders (real people and owners of physical property) will lose out. Further, Mixed Reality is inherently a multi-channel exercise: which applications and which channels within those applications will become dominant is anyone’s guess.221 However, once an application or channel does become dominant, those users who first adopted a technology run the risk of being sidelined in favor of IP holders. And interestingly enough, real property owners—the owner of the hypothetical defaced house in the example above, for instance—do not have any such strengthened rights regarding their real world property.222 3.4  Privacy Law: Privacy’s Death and Resurrection The advent of mobile computing has enabled the totalitarian dream (or nightmare) of tracking citizens at all times. For the most part, however, it is not the government that

219   See Occupy Movement (Occupy Wall Street), N.Y. Times, http://nyti.ms/AmZh25 (last updated Feb. 13, 2012) (containing articles chronicling the anti-Wall Street movement and the accompanying protests). 220   See Trademark Policy, Twitter, https://support.twitter.com/groups/33-report-a-violation/ topics/148-policy-information/articles/18367-trademark-policy (last visited Aug. 14, 2011). 221   See Havens, supra note 207 (suggesting that Google Goggles will be a dominant player given Google’s current dominant market position). 222   See id. (“Google will own the virtual air rights within Goggles.”).

BARFIELD PRINT.indd 138

13/11/2018 11:52

Mixed reality  139 tracks citizens. Tracking is largely accomplished through the technology consumers themselves use.223 Tracking technology is rampant and widespread. Google Streetview cars captured unencrypted personal data as the cars passed private homes.224 Facebook initiated an open architecture for its developers that permits almost anyone to capture large amounts of information through an app installed by a friend of a friend.225 GPSenabled cell phones constantly record the real-world locations of their users.226 Internet service providers do the same, tracking their customers across the digital landscape.227 It follows that mobile broadband providers can not only track users’ physical locations, but also correlate those locations with the users’ online browsing habits. Likewise, even our friends and family can track us using widely available technology. As soon as one person takes a photograph and uploads it to Facebook, facial recognition technology can recognize and tag the people in the photograph with metadata (often including date and real-world physical location).228 The government need not do much more than ask for this information from the mass of third parties who have already ­collected and indexed it.229 A discussion of online privacy is necessary in the context of Mixed Reality because Mixed Reality technology permits companies and governments to know not only a person’s digital profile, but also his real-world habits.230 There is nowhere to hide. Offline, real-world activity is now coded and recorded, parsed, and re-sold—thanks to the integration of Mixed Reality applications in our everyday lives—just as online activity has 223   See CHEN, supra note 6, at 47 (“[A]rmed with a camera-equipped smartphone and live streaming-video software, every citizen will have the power to broadcast anything to the world in real time, thus creating a collectively omniscient society of watching eyes.”). 224   See Kevin J. O’Brien, Germany Asks Apple About iPhone’s Data Gathering, N.Y. Times (June 28, 2010), http://www.nytimes.com/2010/06/29/technology/29apple.html (“[I]t had improperly collected 600 gigabytes of personal data, including fragments of e-mail messages and unencrypted passwords, on individuals around the world as it scanned home Wi-Fi networks while it gathered information for its Street View map archive.”). 225   See Emily Steel & Geoffrey A. Fowler, Facebook in Privacy Breach, Wall St. J. (Oct. 18, 2010), http://on.wsj.com/xVVPKF (describing how Facebook apps violate user privacy). 226   Noam Cohen, It’s Tracking Your Every Move and You May Not Even Know, N.Y. Times (Mar. 26, 2011), http://www.nytimes.com/2011/03/26/business/media/26privacy. html (discussing the practice of how cell phone providers track users’ latitude and longitude). 227   See Peter Whoriskey, Every Click You Make, Wash. Post (Apr. 4, 2008), http:// wapo.st/ xpbudQ (describing the growing phenomenon of internet service providers tracking individuals’ online activity). 228   See Lauren Effron, Facebook in Your Face: New Facial Recognition Feature Raises a Few Eyebrows, ABC News (June 10, 2011), http://abcn.ws/wFWrBG (describing Facebook’s facial recognition feature). 229   Cf. Kevin Werbach, Sensors and Sensibilities, 28 Cardozo L. Rev. 2321, 2325 (2008) (explaining the massive surveillance power of individuals resulting from the proliferation of cameraphones coupled with widespread mobile phone usage, and noting that cameraphones function as “sensors hooked into end-user devices”); see Mark Milian, U.S. Sent Google 8,888 Requests for User-Data in 2010, CNN Tech (June 27, 2011), http://www.cnn.com/ 2011/TECH/web/06/27/ google.data.requests/index.html. 230   See Cohen, supra note 226 (“One product, CitySense, makes recommendations about local nightlife to customers who choose to participate based on their cellphone usage. Many smartphone apps already on the market are based on location but that’s with the consent of the user and through GPS, not the cellphone company’s records.”).

BARFIELD PRINT.indd 139

13/11/2018 11:52

140  Research handbook on the law of virtual and augmented reality been. Where I drive every day can be cross-compared to my web surfing habits.231 Where a consumer walks during the day is as marketable as which websites she has visited—and a combination of the two is more potent still. Mixed Reality makes privacy increasingly elusive and unattainable. In fact, some urge that those who care about privacy should give up networked technologies. Former Google CEO Eric Schmidt implied that users who do not want to be tracked by Google all across the Internet, including any site that serves Google ads reporting back to Google, should simply not use the Internet.232 Thus, in more recent years the move by privacy advocates has been to move some activity off the grid, or at least out of the reach of datamining corporations.233 The advent of Mixed Reality technologies forecloses even this option. 3.4.1  Privacy is dead, long live privacy The ubiquity of technology that constantly tracks consumers’ realspace movements and cross-references them with online activity has caused government and corporate actors to declare that “privacy is dead.”234 This Section explores the questions of whether privacy is in fact dead, whether Mixed Reality and mobile computing killed it, and what can be done about the current bleak situation. This chapter takes the position that privacy is not an end state, but rather a point on a sliding continuum between secrecy and disclosure. Because privacy is a tension point, rather than an absolute category, it is inaccurate to state that privacy is dead.235 Rather, the effects of mobile computing on privacy are a side effect of the nature of information systems used to locate, retain, and distribute information. Since people will always seek to keep some information confidential and other people will always seek to discover or disclose it, we can quickly dispense with the “privacy is dead” paradigm. “Privacy is dead” is simply the battle cry of consumer disempowerment. To the extent privacy is dead, it is dead because the law has prevented consumers from getting and using the tools necessary to protect their personal privacy.236 For example, it

231   See Andrew Munchbach, Apple Sued over iPhone Location Tracking Scandal, BGR (Apr. 25, 2011), http://www.bgr.com/2011/04/25/apple-sued-over-iphone-location-tracking-scandal/. 232   See Jared Newman, Google’s Schmidt Roasted for Privacy Comments, PCWorld (Dec. 11, 2009), http://www.pcworld.com/article/184446/googles_schmidt_roasted_for_privacy_comments. html (quoting Schmidt as stating, “[i]f you have something that you don’t want anyone to know, maybe you shouldn’t be doing it in the first place, but if you really need that kind of privacy, the reality is that search engines including Google do retain this information for some time . . . .”). 233   See Helen A.S. Popkin, Privacy Is Dead on Facebook. Get Over It, MSNBC (Jan. 13, 2010), http://on.msnbc.com/xUnaGG. 234   See Newman, supra note 232; see also O’Brien, supra note 224 (“60 percent of households in Germany use a retail bonus card . . .. By participating, consumers give the company the right to collect and market data on their purchasing habits, as well as send them advertising.”); Popkin, supra note 233. 235   See Chen, supra note 6, at 188 (discussing how the conception of privacy has been forced to change and how the focus should be on developing new technologies to combat privacy concerns rather than simply decrying the existing framework); see also Nick Bilton, Privacy Isn’t Dead. Just Ask Google+, N.Y. Times Bits (July 18, 2011, 12:59 PM), http://nyti.ms/ GSOUTE (outlining how Google benefited by learning from privacy issues on Facebook and focusing on privacy concerns in Google+). 236   See Tanzina Vega, Industry Tries To Streamline Privacy Policies for Mobile Users, N.Y. Times Mobile (Aug. 15, 2011), http://nyti.ms/yPYIj7 (describing positive developments on the

BARFIELD PRINT.indd 140

13/11/2018 11:52

Mixed reality  141 still is nearly impossible to surf the Internet securely,237 or to make use of a cell phone without constantly revealing physical location or personal and financial information.238 The goal, then, should be to provide consumers with the simple, built-in tools necessary to protect privacy. 3.4.2  Privacy by design The FTC has made much of “privacy by design.”239 This is an important meme to explore because it is both widespread and ineffective in securing privacy. Technologies that have been designed from the ground up to collect, package, and sell information cannot “by design” keep that information private. This Section explores the privacy by design meme, critiques it, and then discusses some more viable alternatives in the following Sections. Privacy by design is an incorrect approach for two reasons.240 First, the idea that privacy needs to be designed complicates a very simple problem. Corporations do not need to design for privacy because corporations do not need to record their customers’ information in the first place. The need for privacy design arises only because the existing technologies have already been designed to gather and sell customers’ information. Once one rephrases “privacy by design” as “designing systems and services with the purpose of collecting and disseminating information to not collect or disseminate information,”241 the futility of the approach becomes apparent.242 Thus, the first reason that privacy by design has not produced privacy online is because the technologies have been designed not to allow for privacy. The second, and related, reason is that the privacy options that consumers do have are privacy front as including “one company [that] is trying to make privacy policies that are both easy for consumers to read and easy for mobile application developers to create”). 237   See John Markoff, Do We Need a New Internet?, N.Y. Times (Feb. 14, 2009), http://nyti.ms/ GSOVHg; Kate Murphy, New Hacking Tools Pose Bigger Threats to Wi-Fi Users, N.Y. Times (Feb. 17, 2011), http://nyti.ms/GSOY5D. 238   Angwin & Valentino-Devries, supra note 175; Susan Freiwald, Cell Phone Location Data and the Fourth Amendment: A Question of Law, Not Fact, 70 MD. L. REV. 681, 705–6 (2011) (stating that if cell phone data of one individual were recorded, “it could create a . . . virtual map of all the places the person went and how much time he spent at each place along the way”); Ki Mae Heussner, Apple Tracks Location with iPhone, iPad Data, ABC News (Apr. 20, 2011), http://abcn. ws/wghaeh (noting that the “Apple iPhone and iPad 3G record the device’s geographic position and corresponding time stamp in a hidden file”). 239   See Fed. Trade Comm’n, Protecting Consumer Privacy in an Era of Rapid Change (2010), available at http://www.ftc.gov/os/2010/12/101201privacyreport.pdf (advocating that companies adopt “privacy by design” as a means of protecting consumer privacy by limiting disclosure of consumer data through product design). 240   For background on privacy by design, see generally Ira S. Rubinstein, Regulating Privacy by Design, 26 Berkeley Tech. L.J. 1409, 1431–43 (2011) (describing various reasons why privacy by design has not enjoyed the amount of success anticipated). 241   Simply put, the technology today has been developed with collection of user data in mind. The privacy by design concept is seemingly contradictory because it would be used to enhance privacy in systems that have been designed specifically to gather consumer information. 242   See Klinefelter, supra note 177, at 18 (identifying major concerns, in particular for the legal community stemming from confidentiality, and the concerns raised by viruses, third parties, and other bad actors with regard to online research and data saved by third party tracking).

BARFIELD PRINT.indd 141

13/11/2018 11:52

142  Research handbook on the law of virtual and augmented reality designed to be too expensive in terms of the time and attention required to use them.243 For example, industry advocates continue to argue against regulatory enforcement of a browser “do not track” flag that would follow the model of the quite successful federal “do not call” list.244 Instead, consumers must navigate a different privacy architecture for each application provider, online service provider, and software developer. Drawing an analogy to the telephone context is instructive. Prior to the do- not-call list, each telemarketer was required to maintain and honor lists of people who did not wish to be contacted. Yet the sheer weight of informing each telemarketer (never mind the telemarketers who ignored the rules) made it such that telemarketing was hardly impacted. A unified, simple do-not-call list permitted consumers to express their preferences just once, rather than serially on the phone serially with each one of a thousand different callers. Privacy by design is to some degree contradictory because the current commercial data architectures are in fact mechanisms for collecting, packaging, and reselling consumers’ private and personally identifiable data.245 Further, the privacy options that online users do have are designed to exhaust and confuse the user by requiring them to understand and address their privacy concerns with each vendor separately.246 Privacy by design is a system designed not to work. 3.4.3  Privacy as control The solution to the privacy problem is simple, default, universal, and legally enforceable consumer controls for privacy. Consumer control is not only necessary; it is an effective solution in light of changing consumer conceptions with regard to privacy. Consumers are coming to treat privacy as a matter of control rather than an absolute prohibition on disclosure.247 A consumer control regime would, as it should, allow consumers to sell their personal information and allow internet companies to use it.248 If implemented, effective consumer-side privacy controls can provide a true market in information. Additionally, the decision as to whether or not to permit online or offline tracking can more easily and effectively reside in the customer’s hands. This is clear in light of the 243   See Vega, supra note 236 (discussing the difficulty with online privacy policies, the importance of privacy policies on data collection, and the growing concern over data collection). 244   See David Goldman, FTC ‘Do Not Track’ Plan Would Cripple Some Web Giants, CNN Money (Dec. 3, 2010), http://money.cnn.com/2010/12/02/technology/ftc_do_not_track/index.htm (identifying several industry leaders, such as Google, who are against “do not track” due to unforeseen security problems and loss in e-commerce and advertising revenues). 245   See Rubinstein, supra note 240, at 1412 (noting that the profits derived from online advertising make firms reluctant to voluntarily impose systems that will increase consumer privacy to the detriment of their ability to collect consumer information). 246   See Vega, supra note 236. 247   See CHEN, supra note 6, at 188; Bilton, supra note 235 (praising Google+ for its default privacy settings). 248   See Vega, supra note 236 (harmonizing the needs of users with the needs of companies to create a balance on the privacy front, but noting how online advertising reduces the costs of mobile applications). Mobile apps are free or cheap largely because of mobile advertising. See also Jim Harper, The Great Privacy Debate—It’s Modern Trade: Web Users Get as Much as They Give, Wall St. J. (Aug. 7, 2010), http://on.wsj.com/ArHG25 (“The reason why a company like Google can spend millions and millions of dollars on free services like its search engine, Gmail, mapping tools, Google Groups and more is because of online advertising that trades in personal information.”).

BARFIELD PRINT.indd 142

13/11/2018 11:52

Mixed reality  143 unworkable alternatives to consumer control of privacy. The current regime of privacy policies (that contain no privacy protections) and EULAs (that bury invasive privacy terms twenty pages deep in electronic documents) has proved unworkable. Similarly, leaving privacy controls in individual companies’ hands has proven to be a longstanding fox-in-the-henhouse type failure. Implementing a consumer control regime would be relatively easy. A simple, expedient solution such as legally enforcing the “do not track” flag already available in browsers would do the trick. A consumer-side “do not track” option would test the economic arguments of privacy naysayers. These naysayers argue that there is no privacy because consumers do not want it, or at least that consumers want products more than privacy.249 For example, Eric Schmidt has stated in relation to Google Streetview that “[i]f you have something that you don’t want anyone to know, maybe you shouldn’t be doing it in the first place.”250 Milder versions of the same strange argument include the assertion that consumers who do not wish to be tracked are free to not use Google, or are free not to use the Internet, or are free not to use the telephone. And with the advent of mobile computing technology, we might say that a consumer who does not wish to be tracked and recorded is free not to leave her house. The market for consumer privacy has yet to be tested because “privacy by design” policies shift all of the transaction costs of privacy onto consumers. To discover what consumers make of privacy online, the transaction costs of privacy should be shifted from consumers to the owners of internet technology. Shifting the transaction costs from consumers and offering consumers simple and legally enforceable control over online and offline privacy would also test the argument advanced by some in the internet technology industry that citizens do not want privacy.251 And if internet technology companies like Google refuse to respect consumers’ privacy settings, companies will have the choice to not offer service to those consumers. A true market for privacy requires customers to have market choices that are not overwhelmingly burdened by transaction costs. The self-regulating approach to the private information market in the United States characterized by decentralized, complex, and unenforceable privacy controls has resulted in full market failure. Simple, legally enforceable, default consumer-side browser-level protections for consumers will remedy this problem by centralizing decision-making in consumers, rather than in corporations or the government. Privacy as control represents an alternative to privacy by design.252 Privacy as control assumes that consumers have effective, unitary, and legally enforceable controls in their own hands,253 rather than scattered, complex controls that vary according to each service

249   See Harper, supra note 248 (arguing the same point on behalf of consumers in that protections for consumers would invite them to abandon personal responsibility). 250   See Newman, supra note 232. 251   See Rubinstein, supra note 240, at 1412 (listing various reasons why consumers may not want privacy, such as lack of knowledge, behavioral biases, or simply not caring about the issue). 252   See Chen, supra note 6, at 188–89 (“Perhaps we have already given up our digital privacy, but we still have control over boundaries . . .. In a modern online context a violation of privacy may only occur when we are manipulated into sharing more than we were told we would be sharing.”). 253   See Lawrence Lessig, Code Version 2.0, at 88–111, 228–30 (2006) (arguing for a property model to protect privacy).

BARFIELD PRINT.indd 143

13/11/2018 11:52

144  Research handbook on the law of virtual and augmented reality provider.254 This vision of true consumer control over privacy is particularly important in the Mixed Reality context. Without real control over information, consumers will be every bit as subject to constant tracking in their real lives as they are now in their online habits.

4.  BALANCED LAW FOR MIXED REALITY Mixed Reality merges the real world and cyberspace. It presents exciting opportunities for consumers to augment realspace with rich virtual experiences. But the merging of real and virtual worlds also presents a basic legal problem. Law online has ventured far away from its offline roots. There is a very real risk that as virtual and real merge, the law intended to govern intangible assets will come to govern everyday life.255 The law governing intangible assets was not designed to apply to the real world. If and when it is applied to the real world, the result will be extremely problematic. The specialized law of intellectual property and online contracting is not the best rule set from which to draw rules about everyday human life. Real world analogies, not online analogies, are the best source for legal rules governing the convergent technology of Mixed Reality.256 Solutions that have proven useful in law historically should be applied to the emerging legal problems generated by Mixed Reality.257 The law has long-evolved internal checks and balances. For example, the common law has long imposed restrictions on how much control an intellectual property owner may assert once she has sold a product.258 Similarly, the law has long set basic limits on contracts—limits that should be given new life in the online context generally, but also particularly in the context of Mixed Reality. Very little is needed to solve one of the major problems, that of protecting consumers’ data. Here the law need only enforce consumers’ expressed preferences to maintain their privacy and reject the pernicious myth of consent to the sale of personal information. Simple, unitary, default, and legally enforceable privacy controls will generate a much better market in consumer information.259 The following Sections demonstrate through three examples 254   See Chen, supra note 6, at 189 (“Online privacy advocates criticize online services when they are unclear or dishonest about what they are doing with our data, not when they are using our data—because, of course, they are.”). 255   See Boone, supra note 53, at 114–15 (detailing that underlying code is what controls a virtual world). See generally Chen, supra note 6 (describing how the iPhone collapsed the physical and virtual world); Barfield, supra note 5, at 161 (describing advertising in augmented reality); Gross, supra note 11 (describing an augmented reality mobile application); Wilson, supra note 52, at 1133 (describing the potential of near field communications). 256   See Lyria Bennet Moses, Recurring Dilemmas: The Law’s Race To Keep Up with Technological Change, 7 U. Ill. J.L. Tech. & Pol’y 239, 279–80 (2007); see also Perzanowski & Schultz, supra note 158, at 892. 257   See Fairfield, supra note 8, at 475–76 (“For online communities to thrive, courts must recognize that private property, torts, and other community-critical rights and obligations can be adapted from the familiar rules that already govern communities in the real world to suit the realities of the virtual world.”). 258   See Perzanowski & Schultz, supra note 158, at 892 (“[C]opyright exhaustion, like many principles recognized in the Copyright Act, was created by and should continue to develop through common law judicial reasoning.”). 259   Cf. Eric J. Feigin, Architecture of Consent: Internet Protocols and Their Legal Implications,

BARFIELD PRINT.indd 144

13/11/2018 11:52

Mixed reality  145 that the law of intangible assets applied online should not be applied to Mixed Reality applications. 4.1  Constraining Intellectual Property The first and largest problem that this chapter has identified is that as virtual and real legal interests merge, the law appears set to grant far greater rights to intellectual property holders than it does to holders of other legal rights, like personal dignity or real property.260 Copyright law is the main culprit, but other areas of intellectual property are also at fault. For example, as seen in Section 3.3, supra, owners of trademarks have a decided advantage in the race for prime online real estate.261 The law has already developed checks on the ability of an intellectual property holder to exert continuing control over its customers once it has sold its product, but courts have not applied these checks to online law.262 Copyright is meant to prevent copying. Once a copy is sold however, a copyright owner no longer has the power to control the copy. This is the common law doctrine of “first sale,” which has been enshrined in the Copyright Act.263 The doctrine hinges on whether a copyrighted work has been “sold” or merely “licensed.” The answer to this question is complex, and courts rarely get it right. Again, online law has diverged strongly from offline law. For example, although Netflix can buy a physical DVD and rent it out to any customer (thus ensuring that almost any program or movie is available through Netflix’s mail service), Netflix must seek individual license deals on a per-provider basis in order to stream the very same content. Thus, while Netflix’s ability to use the physical copy of the same movie is not restricted, its ability to utilize an online copy is curbed by licensing and copyright law—in this case, the IP holder has much more power over the online version. There is, however, one interesting development. Courts seem to draw a distinction between electronic data that is recorded or embedded in a physical medium, and data that is merely free flowing.264 This is an improper distinction from a functional point of view: there is no relevant distinction between a song encoded on a CD and a song downloaded as an MP3. Yet courts continue to apply real world physical analogies to digital goods embedded in physical objects and apply the law of intangible assets to digital goods that are

56 Stan. L. Rev. 901, 902 (2004) (“Higher-level protocols, such as those utilized in most web interactions, involve exchanges that should be considered express consent: the formation of a legally binding contract.”). 260   See Lemley, supra note 139; see also sources cited supra note 255. 261   See text accompanying supra notes 211–12. 262   See 17 U.S.C. § 106(1) (2010) (giving the exclusive right to reproduce work that is copyrighted). But see id. § 117(a)(1) (providing for the essential step defense for the software context where an owner of a lawful copy does not infringe the reproduction right of the copyright owner if the reproduction is an essential step in the utilization of the software); id. § 109(a) (providing for the first sale doctrine where a lawful owner of a copy of a copyrighted work is able to sell or otherwise dispose of the possession of the relevant copy at the owner’s discretion). Both of these affirmative defenses are limited to owners of lawful copies of copyrighted works. 263   § 109(a). 264   See RealNetworks, Inc. v. Streambox, Inc., No. C99-2070P, 2000 U.S. Dist. LEXIS 1889, at *1, *3 (W.D. Wash. Jan. 18, 2000) (“Streaming is to be contrasted with ‘downloading.’”).

BARFIELD PRINT.indd 145

13/11/2018 11:52

146  Research handbook on the law of virtual and augmented reality not so embedded. Thus, while a consumer may record movies on her TiVo, there is a serious question as to whether she may record a streaming movie with a virtual VCR.265 Similarly, recent cases in the Ninth Circuit indicate that although a seller may not resell copies of a computer program on e-Bay (since the original license agreement purported to prohibit such resale),266 she would be free to sell music CDs that purported to have the self-same restriction.267 The presence of some physical element—here the TiVo physical box or the music CD—seems to provide courts with some comfort that analogies to the law of the sale of physical objects is a better analogy than the law of licensing of intellectual property. These cases may provide a ray of hope for Mixed Reality. Although the data is tied and not embedded, there is hope that the link to the real world may make courts more likely to use analogies drawn from the full range of law, rather than analogies drawn solely from online intellectual property law.268 4.2  Limiting Online Contractual Control The second step to returning balance to the law as cyberspace and realspace merge is to restore balance to online contract. Reducing the application of intellectual property law and increasing the application of traditional principals of contract and property law can largely restore balance in online contracts.269 This will return contract law to its normal place within the constellation of legal tools as the tool of bargained-for exchange and expectation damages. As things stand, online contracts often implicate intellectual property statutory damages that have no relation to actual damages. For example, when a user modifies her physical gaming console in violation of an EULA, the console seller can pursue statutory damages for direct and vicarious copyright infringement. However, if traditional principles of contract law applied, the console seller would be limited to contract damages. The latter approach follows logically because the gamer owns her console and should be able to do with it what she will. A return to basic principles of contract law has also generated promising trends elsewhere in the law. A related area in which intellectual property control over contracts is

265   Id. at *34–35 (halting Streambox’s continuance of its product, a virtual VCR). But see Cartoon Network LP v. CSC Holdings, Inc., 536 F.3d 121, 128 (2d Cir. 2008) (“Accordingly, we construe MAI Systems and its progeny as holding that loading a program into a computer’s RAM can result in copying that program. We do not read MAI Systems as holding that, as a matter of law, loading a program into a form of RAM always results in copying.”). 266   See Vernor v. Autodesk, Inc., 621 F.3d 1102, 1110–12 (9th Cir. 2010) (establishing a three point framework to determine if a purchaser of software is an owner or licensee). 267   See UMG Recordings, Inc. v. Augusto, 628 F.3d 1175, 1183 (9th Cir. 2011) (finding resale of a music CD not to constitute copyright infringement despite a label claiming a license limitation restricting such a sale). 268   See Moses, supra note 256. 269   See Lyria Bennet Moses, Toward a General Theory of Law and Technology: Why Have a Theory of Law and Technological Change?, 8 Minn. J.L. Sci. & Tech. 589, 595–96 (2007) (“Overemphasis on the technological angle in discussing legal and social problems is evident in various contexts. . .. Judges occasionally fall into the same trap of assuming that because events took place on the Internet, the law must be different.”).

BARFIELD PRINT.indd 146

13/11/2018 11:52

Mixed reality  147 being scaled back is in circuit courts’ interpretations of the Digital Millennium Copyright Act (“DMCA”).270 Here, the intersection between contract and copyright occurs in the anti-circumvention systems used to protect the copyrighted materials. A click-through contract (the classic “click I Agree or exit”) can serve as both a binding contract and as a technological protection measure,271 since the content governed by the contract cannot be accessed without going through the contract. If the DMCA’s prohibition on circumvention of such protective measures is relaxed, users can make use of their programs on their own devices despite overreaching contract terms. Interestingly, here too courts have been persuaded by analogies to physicality to relax the prohibitions of the DMCA.272 Whereas hacking into a purely electronic software program seems to be a clear violation of the DMCA,273 bypassing protections embedded into physical objects receives more lenient treatment. Indeed, the Library of Congress recently added anti-circumvention exceptions to the DMCA that would allow users to alter their smartphones to use unofficial but legally-obtained software.274 The Court of Appeals for the Sixth Circuit held that circumventing controls that limited the number of times that a user could refill printer cartridges that she had purchased was not a violation of the DMCA.275 Similarly, the Federal Circuit held that a universal garage door opener that bypassed the garage door manufacturer’s rolling numeric access code did not trigger the sanctions of the DMCA.276 It may seem obvious that universal remotes do not violate anti-hacking laws, but legally speaking the issue is an extraordinarily close one, since software is embedded in the remote. Courts’ willingness to give weight to the consumer’s expectation that a garage door would be compatible with universal remotes over the strict letter of the DMCA affirms the importance of consumers’ property rights and ­expectations with respect to their own property. Two ancient but basic limits on overreaching contractual control are slowly coming back into fashion online: consideration and its cousin, illusoriness. The common law has long declined to look into the value of a particular bargained-for exchange, but has instead used the doctrines of consideration and illusoriness to ensure that some bargain was indeed struck—that promises were made on both sides.277 But in the online context it is not clear that 270   DMCA, Pub. L. No. 105-304, 112 Stat. 2860 (1998) (codified as amended in scattered sections of 17 U.S.C.). The DMCA contains three provisions that create a framework to address circumvention of technological measures that protect copyrighted works. See 17 U.S.C. § 1201(a) (1)–(2), (b)(1) (2010). 271   See ProCD, Inc. v. Zeidenberg, 86 F.3d 1447, 1452 (7th Cir. 1996); see also 17 U.S.C. §  1201(a)(1)(A) (“No person shall circumvent a technological measure that effectively controls access to a work protected under this title.”). 272   See Lexmark Int’l, Inc. v. Static Control Components, Inc., 387 F.3d 522, 529 (6th Cir. 2004); Chamberlain Grp., Inc. v. Skylink Techs., Inc., 381 F.3d 1178, 1203–04 (Fed. Cir. 2004). 273   § 1201(a)(1)(A). 274   See Paul Miller, Library of Congress Adds DMCA Exception for Jailbreaking or Rooting Your Phone, ENGADGET (July 26, 2010), http://www.engadget.com/2010/07/26/library-of-con​ gress-adds-dmca-exception-for-jailbreaking-or-root/ (noting the Library of Congress’s exception to the DMCA). 275   Lexmark, 387 F.3d at 529. 276   Chamberlain, 381 F.3d at 1182. 277   See Harris v. Blockbuster Inc., 622 F. Supp. 2d 396, 397–400 (N.D. Tex. 2009) (holding that an arbitration clause was illusory because the drafter could alter it at will).

BARFIELD PRINT.indd 147

13/11/2018 11:52

148  Research handbook on the law of virtual and augmented reality enforceable promises are being made on both sides. EULAs and TOUs are lists of promises that the user makes. Ostensibly, the return promise by the corporation is that it will permit the user to access a valued service. But courts are increasingly questioning contracts that contain unlimited modification clauses.278 These are a staple of online contracts, but they are becoming more and more dangerous for companies. Judges have begun to reason that if a company is free to change the EULA or TOU at any time and in any way, then the company has not made any true return promise.279 This is an important legal development since it increases the odds that the contract to which a consumer agrees will state the actual responsibilities that may eventually be enforced against the consumer. Similarly, such legal rulings increase consumers’ confidence that the return promises of the company are equally enforceable. Constraining overreaching contracts—especially those contracts that forbid the user to customize or accessorize her own property—is essential for Mixed Reality. Mixed Reality devices will increasingly control how users view the world around them. They are the access point for users’ ability to see the data-enriched experiences that augment real places, people, and things.280 The devices that control Mixed Reality experiences must be firmly in citizens’ hands. Owners of Mixed Reality devices must be able to modify those devices in order to control the reality that they experience. The alternative is imaginable but unthinkable: just as Sony now claims that it has the sole right to control what players experience via its gaming console, on pain of criminal sanction and intellectual property infringement, so Mixed Reality providers would claim the ability to control the very reality that citizens experience and share. 4.3  Returning Control over Privacy to Consumers In the privacy context, a very simple but fundamental rebalancing of contract law as applied online will resolve many problems. Privacy is handled as a matter of contract under U.S. law.281 That is not the problem. The problem is that the law of contract as applied online has denied consumers the power to draft contracts. Even in the strangest circumstances, courts enforce contracts written by corporations, including legal language contained on pages that the user has not even seen. Yet that reasoning ought to cut both ways. Citizens—as parties to a contract—have as much of a right to add binding terms to a contract as corporations do.282 If a consumer

278   See, e.g., Bragg v. Linden Research, Inc., 487 F. Supp. 2d 593, 611 (E.D. Pa. 2007) (finding that a unilateral modification provision was unconscionable). 279   See Morrison v. Amway Corp., 517 F.3d 248 (5th Cir. 2008). 280   See CHEN, supra note 6; Bilton, supra note 89; Glusac, supra note 17; Husson, supra note 89; King, supra note 22. 281   See News Release, Fed. Trade Comm’n, FTC Announces Settlement with Bankrupt Website, Toysmart.com, Regarding Alleged Privacy Policy Violations (July 21, 2000), http:// www.ftc.gov/ opa/2000/07/toysmart2.shtm (discussing the FTC’s suit against Toysmart and the company’s attempt to take action directly in violation of its privacy policy). Cases that have found otherwise only serve to emphasize the problem that lies in interpreting privacy as non-contractual. See, e.g., In re Jet Blue Airways Corp. Privacy Litigation, 379 F. Supp. 2d 299 (E.D.N.Y. 2005); Dyer v. Nw. Airlines Corp., 334 F. Supp. 2d 1196 (D.N.D. 2004). This does not, however, undermine the notion that the FTC still enforces privacy policies as promises to consumers. 282   See Lessig, supra note 253; Feigin, supra note 259.

BARFIELD PRINT.indd 148

13/11/2018 11:52

Mixed reality  149 sets a “do not track” flag in her browser, courts should enforce it as a matter of contract law. Courts have long held that consumers “agree” to online contracts by continuing to use a website or online service.283 It would be nonsensical not to apply the same logic to a corporation: by continuing to provide service to a customer who has set a “do not track” flag, the corporation should be legally bound by that term.284 This proposal is of course not uncontroversial, but it exposes the poor reasoning underlying the law of online contracting.285 Functionally speaking, under current law only corporations are allowed to draft online contracts. Citizens are only granted the pro-forma right to agree to pre-set corporate terms. Citizens are denied a voice in setting the terms under which their information is gathered.286 It is this perverse state of affairs that we must correct before the law of online contracting can govern everyday life. Giving consumers control does not mean that control should be complicated. To impose all of the transaction costs of privacy protection on consumers—the current state of affairs—is planned failure. In order for consumers to be able to express their preferences effectively, privacy controls must be simple, unitary (in one place, applicable to all counterparties), default, and legally enforceable.287 Only under these circumstances will the transaction costs of private agreement over privacy be manageable for consumers. The alternatives are significantly less attractive: outright paternalistic government regulation on the one hand, or the current wild west of data protection on the other. The advent of Mixed Reality technology makes this return to consumer control over personal data of significant importance. Consider the average smartphone, which comes pre-loaded with numerous apps, each of which has different permissions to track the consumer’s real-world location, social network interactions, and even tap into the basic reality that the consumer is experiencing—what she is hearing and seeing.288 Consumers must have control over their own reality, and this includes the ability to control their own information as it is propagated through these networks.

283   See Yen-Shyang Tseng, Governing Virtual Worlds: Interration 2.0, 35 Wash. U. J.L. & Pol’y 547, 556 (2011) (“Generally speaking, courts have tended to enforce all of these forms of licenses, even though the licenses may unilaterally impose one-sided terms with little to no room for negotiation.”). 284   See Lessig, supra note 253; Feigin, supra note 259. 285   See, e.g., Woodrow Hartzog, Website Design as Contract, 60 Am. U. L. Rev. 1635, 1662 (2011) (noting various problems with trying to achieve privacy by consumer contracting). 286   See Newman, supra note 232; Popkin, supra note 233. 287   See Lessig, supra note 253; Feigin, supra note 259; see also Séverine Dusollier, The Master’s Tools v. The Master’s House: Creative Commons v. Copyright, 29 Colum. J.L. & Arts 271, 272 (2006) (discussing the purpose of Creative Commons to address the “recent expansion of copyright” and how it is “overreaching and detrimental both for future creators and for the users of copyrighted works”); About the Licenses, Creative Commons, http://creative commons.org/ licenses/ (last visited Feb. 19, 2012) (providing for a standard option set of contractual licenses that have revolutionized online exchanges). 288   See Yukari Iwatani Kane & Scott Thurm, Your Apps Are Watching You, Wall St. J. (Dec. 17, 2010), http://on.wsj.com/wq7Wiw (describing how iPhone and Android transmit various data about the phone without the user’s knowledge).

BARFIELD PRINT.indd 149

13/11/2018 11:52

150  Research handbook on the law of virtual and augmented reality

5. CONCLUSION The proposals here are not exhaustive. Rather, they are examples of potential benefits to be gained through application of a method. That method involves applying the common law approach—reasoned, careful, limited, and iterative decision-making based on the closest legal analogy—in order to find potential solutions to emerging technological problems.289 Mixed reality tools are pushing intellectual property regimes into realspace. In some senses, this is nothing new. Books are real, and intellectual property governs our ability to copy them. But the law of intellectual property licensing online has drifted from its moorings. Offline, the law of copyright has generally been limited to restricting the ability of a buyer to make copies, perform unlicensed performances or screenings, or create nonparody derivative works. Not so online, where copyright law has the ability to control the social rules of multiple-million member online communities. Mixed Reality technology brings this online over-extension of copyright licensing back into everyday life. As Mixed Reality merges virtual experiences with everyday life, there is a very real risk that courts will continue to draw on the law of online intellectual property licensing. The confluence of the judicial acceptance of pro forma corporate contracts coupled with the strength of contracts backed by out-of-proportion copyright infringement damages means that the law of online contracting and intellectual property licensing is a terrible fit for offline, everyday life. In everyday life, when a consumer buys a car, the consumer expects to be able to paint it any color. Yet when a consumer buys a garage door, there is a non-trivial question of law as to whether the manufacturer of the garage door may force the consumer to buy new remote controls from only the garage door manufacturer. And when a consumer buys a Playstation 3, there are very real legal threats from Sony when the consumer modifies her own property and teaches others how to do the same. This overextension of copyright licensing and online contracting law is undermining property rights in the real world, providing perverse incentives for online reputation purveyors to whitewash the reputations of network users, and burying consumers under thousands of differing privacy policies, many of which are not enforceable or which may be changed at any time by the software provider. But there is hope. This chapter proposes applying simple rules evolved in the full context of real-world situations to merged virtual and real experiences. A consumer’s property interest in her goods should permit her to make aftermarket modifications of her own property. A copyright holder’s rights in a given copy should be exhausted when the copyright holder sells a given copy away, never mind that the transaction is spuriously characterized as a license. And history has shown that consumers not only want privacy but can enforce their preferences quite effectively. However, consumers must be granted simple, unitary, and default tools that permit them to have an active say in the information gathering regimes to which they are subject, rather than the option to pick which one of a set of corporate-drafted terms they may agree to. As Mixed Reality merges virtual and real space, it provides serious challenges to law, 289

  See Perzanowski & Schultz, supra note 158; see also Moses, supra note 256, at 241.

BARFIELD PRINT.indd 150

13/11/2018 11:52

Mixed reality  151 but also offers serious hope. The law of intellectual property as applied to the real world is subject to traditional constraints that render it much less problematic than the unconstrained law of intellectual property as applied to Mixed Reality. The law of contracting in the real world grants both parties, not just corporations, the ability to contribute terms to the contract. Consumers actually can police their privacy quite effectively, if given simple, opt-in, and default options to do so. Mixed Reality opens the door to the application of common sense rules that have very effectively mediated the tensions between corporation and consumer, citizen and state. Mixed Reality need not be a dystopian vision. It may be the method by which we can restore balance to the law.

SOURCES Cases Barnes v. Yahoo!, Inc., 570 F.3d 1096, 9th Cir.(Or.), 2009 Bowers v. Baystate Technologies, Inc., 320 F.3d 1317, Fed.Cir. (Mass.), 2003 Bragg v. Linden Research, Inc., 487 F. Supp. 2d 593 (E.D. Pa. 2007) Cartoon Network LP, LLLP v. CSC Holdings, Inc., 536 F.3d 121, 2nd Cir. (N.Y.), 2008 Chamberlain Group, Inc. v. Skylink Technologies, Inc., 381 F.3d 1178, Fed.Cir. (Ill.), 2004 Chicago Lawyers’ Committee for Civil Rights Under Law, Inc. v. Craigslist, Inc., 519 F.3d 666, 7th Cir.(Ill.), 2008 CompuServe Inc. v. Cyber Promotions, Inc., 962 F.Supp. 1015, S.D.Ohio, 1997 Davidson & Associates v. Jung, 422 F.3d 630, 8th Cir.(Mo.), 2005 Davidson & Associates, Inc. v. Internet Gateway, Inc., 334 F.Supp.2d 1164, E.D.Mo., 2004 Doe v. MySpace, Inc., 528 F.3d 413, 5th Cir. (Tex.), 2008 Doe v. SexSearch.com, 502 F.Supp.2d 719, N.D.Ohio, 2007 Dyer v. Northwest Airlines Corporations, 334 F.Supp.2d 1196, D.N.D., 2004 eBay, Inc. v. Bidder’s Edge, Inc., 100 F.Supp.2d 1058, N.D.Cal., 2000 Fair Housing Council of San Fernando Valley v. Roommates.Com, LLC, 521 F.3d 1157, 9th Cir. (Cal.), 2008 Harris v. Blockbuster Inc., 622 F.Supp.2d 396, N.D.Tex., 2009 In re Jetblue Airways Corp. Privacy Litigation, 379 F.Supp.2d 299, E.D.N.Y., 2005 Lexmark Intern., Inc. v. Static Control Components, Inc., 387 F.3d 522, 6th Cir. (Ky.), 2004 MAI Systems Corp. v. Peak Computer, Inc., 991 F.2d 511, 9th Cir. (Cal.), 1993 Mazur v. eBay Inc., 2008 WL 618988, N.D.Cal., 2008 MDY Industries, LLC v. Blizzard Entertainment, Inc., 629 F.3d 928, 9th Cir. (Ariz.), 2010 Morrison v. Amway Corp., 517 F.3d 248, 5th Cir.(Tex.), 2008 ProCD, Inc. v. Zeidenberg, 86 F.3d 1447, 7th Cir.(Wis.), 1996 RealNetworks, Inc. v. Streambox, Inc., 1999 WL 1448173, W.D.Wash., 1999 Recording Industry Ass’n of America, Inc. v. Verizon Internet Services, Inc., 351 F.3d 1229, D.C.Cir., 2003 Recording Industry Ass’n of America v. Diamond Multimedia Systems, Inc., 180 F.3d 1072, 9th Cir. (Cal.), 1999 Reit v. Yelp!, Inc., 907 N.Y.S.2d 411, N.Y.Sup., 2010 Rosenberg v. Harwood, No. 100916536, 2011 WL 3153314 (D. Utah May 27, 2011) Triad Systems Corp. v. Southeastern Exp. Co., 64 F.3d 1330, 9th Cir. (Cal.), 1995 UMG Recordings, Inc. v. Augusto, 628 F.3d 1175, 9th Cir. (Cal.), 2011 Vernor v. Autodesk, Inc., 621 F.3d 1102, 9th Cir. (Wash.), 2010 Wall Data Inc. v. Los Angeles County Sheriff’s Dept., 447 F.3d 769, 9th Cir. (Cal.), 2006

Statutes and Regulations Communications Decency Act, 47 U.S.C. § 230(c)(1), (e)(2) (2010) Digital Millennium Copyright Act, 17 U.S.C. § 1201(a)(1)–(2), (b)(1) (2006) In re Implementation of Section 6002(b) of the Omnibus Budget Reconciliation Act of 1993 (DataTraffic Growth), 25 FCC Rcd. 11407, 11412-25 (2010) Pub. L. No. 106-113, S 1000(a)(9), 113 Stat. 1501, 1536 (1999) (enacting into law § 3002 of the Intellectual

BARFIELD PRINT.indd 151

13/11/2018 11:52

152  Research handbook on the law of virtual and augmented reality Property and Communications Omnibus Reform Act of 1999, S.1948, 106th Cong., reprinted at 113 Stat. 1501A-545 to -548 (1999) and codified as amended at 15 U.S.C. S 1125(d)) Restatement (Second) of Torts S 569 (1977) Trademark Act of 1946, Pub. L. No 79-489, 60 Stat. 427 (codified as amended in scattered sections of 15 U.S.C.) 17 U.S.C. § 504(c) (2010)

BARFIELD PRINT.indd 152

13/11/2018 11:52

PART II INTELLECTUAL PROPERTY LAW

BARFIELD PRINT.indd 153

13/11/2018 11:52

5.  Virtual copyright

Michael Risch and Jack Russo*

1. INTRODUCTION In the iconic book 1984,1 first published in 1949, George Orwell presented a dystopian view of the future where “Big Brother” observed all, controlled all, and persecuted all aspects of individualism and independent thinking as “thoughtcrimes” captured by “Thought Police” and punishable by death. In the iconic Macintosh introduction Super Bowl commercial “1984,”2 Apple Computer, Inc. (now just “Apple”) presented the antidystopian image of a young woman successfully challenging the status quo (of older men) to bring democracy to computing controlled by those who would similarly reject individualism and independent thinking. In 1984 (the year), one of us was legal counsel to the inventors and the company they established for the innovation described as “virtual reality,”3 a term that eventually became the generic and descriptive way to refer to what was then an entirely new class of computer software-based applications. These virtual reality applications were designed with the antidystopian goals envisioned by the personal computer revolution. Though that business, VPL, ultimately went under, the inventions, innovations, and intellectual properties created more than 30 years ago became the foundation for a multitude of new inventions and new innovations that have given rise to virtual reality (“VR”) applications on computers and mobile devices.4 We believe the goals of VR continue to include individualism, independent thinking, and individual creativity in a world that continues to democratize information (as demonstrated by how easily we can provide online references to the history of these developments in the footnotes to this chapter). Thus, nearly 25 years ago,5 we wrote that new developments in computer technology, dubbed “virtual reality” applications, would seek to put the computer user into the computing environment as a way to traverse and interact more intimately with informa-

*  Jack Russo is the founder and managing attorney of Computerlaw Group LLP; more about him and his firm can be found at http://www.computerlaw.com. Michael Risch is Associate Dean for Faculty Research & Development & Professor of Law at Villanova University Charles Widger School of Law; he is formerly a partner with Jack Russo, and remains Of Counsel to the firm. Portions of this work appeared in other forms in past work by the authors. 1   George Orwell, Nineteen Eighty-Four: A Novel (1949). 2   1984 Apple’s First Macintosh Commercial as republished at http://www.youtube.com/wat​ ch?v=OYecfV3ubP8. 3   See Virtual Reality Society, at http://www.vrs.org.uk/virtual-reality-profiles/vpl-research. html, for details about Jaron Lanier and VPL. Jack Russo was the startup counsel for Jaron Lanier and the original incorporating attorney for VPL Research, Inc. 4   See, e.g., The Virtual Engagement Platform, at http://www.immersiveentertainment.com. 5   Jack Russo & Michael Risch, Copyright Protection for Virtual Realities: Brave New World Emerges, National Law Journal (October 17, 1994).

154

BARFIELD PRINT.indd 154

13/11/2018 11:52

Virtual copyright  155 tion. We predicted then that significant legal issues would arise from this new software technology and we imagined correctly that many challenges would exist in the way intellectual property laws in the United States (and throughout the world) would come to address these new applications. This chapter discusses a subset of these issues: the scope of copyright protection for virtual reality works.6 Since their introduction in the 1980s the number of virtual reality applications has grown rapidly, and significant efforts have been made to make the necessary technology available and affordable to the general public.7 As one commentator has put it: “A virtual world is a computer that you operate with natural gestures, not by composing computer programs, but by walking around, looking around, and using your hands to manipulate objects.”8 Another has said, “In a virtual world, we are inside an environment of pure information that we can see, hear, and touch.”9 Multiple virtual “objects” can exist within a virtual “world,” multiple virtual worlds can exist within a virtual “reality,” and multiple virtual realities can be explored with the same computer equipment and, indeed, during the same session in which the user is interacting with that computer equipment. From a copyright point of view, these objects and worlds are all separate works. Indeed, the simplest of stories can be part of a world and also a separate work as well.10 Thus, virtual reality applications pose factual, legal, and even metaphysical questions. Metaphysically, virtual objects can have all of the same properties as the real objects they represent; consequently, as software and hardware technologies improve, it will become almost impossible to distinguish the virtual from the real. Factually, virtual reality applications will present novel ways of looking at traditional information and new ways of interacting with computing environments. Legally, at the point at which virtual objects can barely be distinguished from real objects,11 it will also be nearly impossible to tell whether someone has “copied” or “prepared a derivative work” from part or even all of a virtual world when such copying or use is not admitted.

 6   This chapter is based in part on our prior work. We focus here on VR applications and not also on the AR (augmented reality), MR (mixed reality), and other “x/R” variants that have proliferated and continue to grow as further smartphone and mobile devices expand personal computing through Internet connectivity throughout the world, but analogous principles should apply to all of these creative developments based on original human individual creativity and independent thinking and expression thereof in the new media and communication technologies.  7   See William Bricken, Virtual Reality: Directions of Growth 1 (Meckler, Ltd., 1991) (attributing the first computer-driven reality applications to VPL Research, Inc. of Redwood City, California and Autodesk of Sausalito, California. Broader definitions of virtual reality demonstrate that the concepts existed at least since the introduction of computers, if not before then).  8   Howard Rheingold, Virtual Reality: The Revolutionary Technology of ComputerGenerated Artificial Worlds—And How It Promises and Threatens to Transform Business and Society 70 (Simon & Schuster 1991).  9   Michael Benedikt, Cyberspace: First Steps 363 (MIT Press 1992), citing Meredith Bricken, Virtual Worlds: No Interface to Design. 10   MDY Industries, LLC v. Blizzard Entertainment, Inc., 629 F.3d 928 (9th Cir. 2010); Micro Star v. FormGen, Inc., 154 F.3d 1107 (9th Cir. 1998). 11   This is sometimes referred to as the point of “singularity” and many other predictions have been made on this topic. See, e.g., Raymond Kurzweil, The Singularity is Near (2005).

BARFIELD PRINT.indd 155

13/11/2018 11:52

156  Research handbook on the law of virtual and augmented reality We do not expect to see traditional characters and animation treated any differently in virtual reality. Mickey Mouse is still Mickey Mouse, and Pikachu lives in trading cards, cartoons, augmented reality, and virtual reality. It is whether and how realistic depiction, gesture control, modularization, and sharing fit within copyright’s limiting doctrines that will create important and difficult questions for future developers, judges, juries, and appellate courts. This chapter discusses the history and applications of virtual reality, analyzes the current case law as it will likely apply to virtual reality, and presents new issues in the scope of protection of virtual reality applications.

2.  BIRTH OF A NEW TECHNOLOGY A virtual reality is almost reality; as the name implies, it is “virtually” real. Virtual reality environments, such as rooms, cities, or entire worlds, are created by computer software executing on high-performance computer hardware. While the terms “virtual reality” and “artificial reality” are generally treated as the same in, for example, stereo reproductions of music, use of the term “artificial reality” historically occurred first and existed before the rise of high-performance personal computers.12 It was not until computing speed, performance, and ability improved and became more common in the 1990s that simulated visual, auditory, and even tactile environments were realistic enough to be properly termed “virtual realities.” Nearing 2020, computers are now so powerful that the average smartphone can provide a virtual reality experience directly and to multiple users interacting in the same virtual world.13 2.1  Various Forms of Virtual Reality At the very broadest interpretation, virtual reality “includes all computer-generated worlds that a viewer can enter and change through the imagination.”14 Thus, a virtual world includes, for example, the environment presented in the Disney movie Tron, an early 1980s fantasy that represents the workings of a computer as “little people” and their vehicles moving within a three-dimensional world. Indeed, the draw and paint programs which graphic artists manipulate and the CAD/CAM programs that architects and engineers use to generate twodimensional objects can also be viewed as creating relatively simple virtual worlds. According to a more moderate definition, virtual realities include environments in which some combination of sensory perception—such as sight, sound, touch, and eventually taste and smell—is simulated to recreate reality. An example is “surround sound” audio reproduction that is recorded on 100 channels from 100 positions; such an audio 12   Myron Krueger, Artificial Reality (1983); Myron Krueger, Artificial Reality II (1991). 13   The “AR” (aka Augmented Reality) form of this exists in the Pokémon Go game that became hugely popular in 2016, and that itself derived from a videogame (called “Pocket Monsters Red and Green”) from the 1990s. See http://www.wikipedia.org/wiki/Pok%C3%A9mon. 14   Charles Hagen, Virtual Reality: Is It Art Yet? New York Times, http://www.nytimes. com/1992/07/05/arts/art-view-virtual-reality-is-it-art-yet.html?pagewanted=all (Jul. 5, 1992).

BARFIELD PRINT.indd 156

13/11/2018 11:52

Virtual copyright  157 reproduction is perceived by most listeners almost exactly as though the live musical creator is present and performing.15 Finally, the narrowest and most modern definition is of virtual reality as a human–computer interface in which the user only partially controls the computer and in which the user seems to have a “natural” involvement in the environment.16 With a natural interface, the user ceases to think of the computer as an object to be manipulated, but more as an environment to explore; the user’s attention and psychology of experience melds into the environment itself.17 Perhaps the most advanced—if fictional—example of this is the “holodeck” in Star Trek: The Next Generation, in which the user is immersed in an environment which, for all intents and purposes, is real. Only if the user “overrides” the computer does he or she gain the ability to control the actions of the computer-simulated beings; otherwise, the computer controls all of the sights, sounds, and actions, while the user merely acts and reacts within the environment. Using uninvented forcefield technology to simulate tactile sensations, Star Trek characters on the holodeck can fight enemies, ride horses, and even swim while seeing, feeling, hearing, and even tasting and smelling everything that is simulated.18 The fictional holodeck extends virtual reality further than is currently possible; it represents a future ideal of virtual reality’s potential.19 The broadest definition of virtual reality encompasses, to some extent, other new technologies emerging in the computer industry, such as multimedia, cyberspace, and the internet. Multimedia refers to the combination of visual display and audio recordings in the computer environment. Cyberspace is generally any nondimensional area where information is exchanged and processed. The internet is the worldwide set of network connections which connect computers.20 This section focuses more on the narrower definitions of virtual reality. 2.2  Invention History The earliest beginnings of virtual reality can be seen in early patents for simulations. One such simulation is of the human hand and was used to communicate with the deaf and blind.21 Figure 5.1 shows a diagram of the device. The patent issued for a keyboard 15   Most current stereo reproductions use just two channels. Even “surround sound” is produced with two channels, but with encoding to help the processor decide which speakers to send signals to. More advanced sound systems, of course, exist in many movie theaters. See, e.g., https:// en.wikipedia.org/wiki/IMAX. 16   See Michael B. Spring, Informating with Virtual Reality, at 7–9, in Sandra Helsel & Judith Roth, Virtual Reality: Theory, Practice, and Promise (1991). 17   See generally Mihaly Csikszentmihaly, Flow: The Psychology of Optimal Experience (1991). 18   See Rick Sternbach & Michael Okuda, Star Trek: The Next Generation Technical Manual 156 (1991). 19   The broadest definition of virtual reality encompasses, to some extent, other technologies, such as multimedia, cyberspace, and the Internet. This chapter focuses more on the narrower definitions of virtual reality. 20   Compare Stop Capitalizing the Word Internet, New Republic (Jul. 28, 2015), https://­newrepub​ lic.com/article/122384/stop-capitalizing-word-internet; with David Post, It’s “the Internet.” Please, The Volokh Conspiracy (Aug. 11, 2011), http://volokh.com/2011/08/11/its-the-internet-please/. 21   See Charles Laenger, et al. “Method and Apparatus For Communicating with People,” US Patent No. 4,074,444 (issued February 21, 1978).

BARFIELD PRINT.indd 157

13/11/2018 11:52

158  Research handbook on the law of virtual and augmented reality

Figure 5.1  Human hand simulation c­onnected to a movable hand. Letters typed into the keyboard were converted into electrical signals which moved that hand into the shape which corresponded to the letter. This invention was a forebear of more recent inventions, such as the robot that wins rock-paper-scissors 100 percent of the time.22 Another, more “virtual reality-like” simulator was invented by Morton Heilig, who patented the “Sensorama Simulator” in 1962. Howard Rheingold describes one “motorcycle” version of the Sensorama Simulator as a pinball machine-like apparatus which presents a three-dimensional film from the viewpoint of a motorcycle driver and plays sounds, including the ignition of the motorcycle, consistent with a motorcycle ride.23 Figure 5.2 is a diagram of the Sensorama Simulator. Simulators have evolved into both hardware-type simulators, such as airplane simulators used in flying schools, and software-type ­simulators, such as Microsoft’s aircraft flight program for Windows. While simulators such as the Sensorama and its modern-day equivalents were necessary for the development of VR, they were not sufficient. The reputed “grandfather” of virtual reality in its present form is Myron Krueger, who first coined the term “artificial reality” in 1974; nine years later, he published a book by the same name.24 Krueger has long since been a visionary in the area of virtual reality. Krueger’s work evolved from an almost artistic viewpoint; his artificial reality “laboratories” were rooms which were alive. Krueger’s first project was a room that created different sound and light patterns depending on where people walked in the room. His projects recognized the movements of people in the room and, using artificial intelligence, automatically simulated them on a 22   Ishikawa Watanabe Laboratory, Janken (rock-paper-scissors) Robot with 100% Winning Rate, YouTube (June 25, 2012), http://www.youtube.com/watch?v=3nxjjztQKtY. 23   See Rheingold, supra note 8, at 49–58. See generally Morton Heilig, “Sensorama Simulator” U.S. Patent No. 3,050,870 (issued Aug. 28, 1962); see also Morton Heilig, “Stereoscopic Television Apparatus for Individual Use,” US Patent No. 2,955,156 (issued Oct. 4, 1960). 24   Myron Krueger, Artificial Reality (1983); see also Rheingold, supra note 8, at 114 (stating when Myron Krueger coined the term “Artificial Reality”).

BARFIELD PRINT.indd 158

13/11/2018 11:52

Virtual copyright  159

Figure 5.2  The Sensorama Simulator

video screen.25 This work represented the beginnings of the virtual reality camera, which could take photographs in three dimensions. Today, such image capture is commonplace; smartphone companies give away 3D cameras with purchases, and the phone cameras themselves can be used to stitch several photos together into a single image. Furthermore, these three-dimensional images have been used to create immersive visual settings. Today there are many CAVE providers, using technology to provide an immersive virtual environment.26 In addition to the development of more realistic simulators and visual environments, the move to more natural data input technology has progressed throughout the years. For example, in 1966, Francis Rossire received a patent on a stress sensor measurement system in aircraft.27 Also, Rockwell developed an optical method to examine different characteristics in materials.28 While these sensors are quite simple by today’s standards, 25   See Rheingold, id. at 113–28 (discussing Krueger’s most recent—at that time—virtual reality projects). 26   Wikipedia, Cave Automatic Virtual Environment (2017), https://en.wikipedia.org/w/index. php?title=Cave_automatic_virtual_environment&oldid=793383517. 27   See generally Francis Rossire, “Stress Sensor and Control System for Limiting Stresses Applied to Aircraft,” US Patent No. 3,229,511 (issued Jan. 18, 1966). 28   See generally Kenneth James, et al. “Optical Deformation Sensor,” U.S. Patent No. 4,420,251 (issued Dec. 13, 1983).

BARFIELD PRINT.indd 159

13/11/2018 11:52

160  Research handbook on the law of virtual and augmented reality

Figure 5.3  Gesture-based control they are examples of early technology that processes physical inputs and were the beginnings of today’s modern physical data input devices, such as gloves and other interactive clothing. In a more applied example, in the 1960s Nicholas Alferieff developed an apparatus which generated different signals by detecting the different button combinations on a handheld device.29 Figure 5.3 shows the drawing that Alferieff delivered to the Patent and Trademark Office. This device sent different signals out depending on the combination of keys pressed; rather than automatic sensing, however, the device used buttons that were pressed. This type of technology is ubiquitous today, from court reporter transcription machines (which are a direct application) to video game controllers (which use combinations of buttons and joystick positions to control avatar movement). In 1983, Gary Grimes of Bell Labs was awarded a patent for a data input system (in the shape of a glove) which detects the position of the hand and outputs an appropriate code. Figure 5.4 is a depiction of Grimes’s glove.30 In addition, four Japanese inventors received early patents for glove-based interfaces for generating musical tones.31 These devices worked in a similar manner, but extended the idea by generating music and other sensory outputs. One of the first businesses to develop virtual reality applications was VPL Research.32 VPL was founded by Jaron Lanier in the early 1980s.33 It developed the DataGlove, which used fiber optic bundles to discern movements of the hand and relate those movements to 29   See generally Nicholas Alferieff, “Man to Machine Communication Keyboard Device,” US Patent No. 3,428,747 (issued Feb. 18, 1969). 30   See generally Gary Grimes, “Digital Data Entry Glove Interface Device,” US Patent No. 4,414,537 (issued Nov. 8, 1983). 31   See generally Hideo Suzuki, et al. “Musical Tone Apparatus Mounted on a Performer’s Body,” US Patent No. 4,905,560 (issued Mar. 6, 1990). 32   Several other companies were original virtual reality developers, but are not discussed in detail here. They include Autodesk and Sense8, among others. 33   See supra note 3.

BARFIELD PRINT.indd 160

13/11/2018 11:52

Virtual copyright  161

Figure 5.4  Data input glove a computer. VPL also developed software to process the signals sent by the DataGlove.34 Today, users can select from an array of gloves to control their VR experience.35 VPL further went on to develop commercial head-mounted visual displays, or “­goggles,” which are worn over the head.36 VPL obtained a number of patents on its work, which was used in a number of consumer, commercial, and military applications.37 Today,

  See Rheingold, supra note 8, at 153–62 (noting VPL development of software to process signals). 35   Pierre Pita, Full List of Glove Controllers for VR, Virtual Reality Times (Mar. 14, 2017), https://virtualrealitytimes.com/2017/03/14/full-list-of-glove-controllers-for-vr/. 36   Goggles are the evolutionary product of the three-dimensional stereoscopic films developed by Morton Heilig. 37   See generally French Firm Gets VPL Patents, San Jose Mercury News, at G1 (Dec. 8, 1992) (discussing how, due to nonpayment of a debt, VPL’s patents reverted to a French investment company); see also Virtual Reality Company Barely Exists, San Francisco Chronicle, at C1 (Dec. 34

BARFIELD PRINT.indd 161

13/11/2018 11:52

162  Research handbook on the law of virtual and augmented reality head-mounted goggles are also ubiquitous, from extremely complex units that coordinate with the computer to smooth image transitions, to stereoscopic cardboard boxes that rely on cellphones to do all the 3D processing work.38 2.3  Current Uses of Virtual Reality Today’s computer hardware and software technology places virtual reality at a level somewhere between the moderate and narrowest definitions. At current technological levels, virtual reality is an environment simulated by a computer that creates a three-dimensional, multisensory environment that mimics reality when perceived by the user. The illusion can be viewed on a computer screen, but the effect is more realistic if headphones, eyewear, gloves, and interactive clothing are used to communicate directly between the user and the computer.39 The effect is enhanced even further by larger virtual environments. The additional components are used to make the computer interface more natural and seamless and the computer output more realistic. Artificial intelligence can make virtual reality applications more realistic by improving both computer responses to input and the overall human–computer interaction. Virtual reality has taken many forms commercially, and it continues to expand. The categories that commercial applications fall into are: military, education, entertainment, games, and medical and professional applications. 2.3.1 Military For example, NASA used a virtual workstation that combines voice and gesture recognition with motion tracking. The user is able to see and manipulate a world covering a full 360-degree range.40 When the user’s head turns, he or she sees different parts of the virtual world under display. When the user reaches out to grab an object, the computer senses if his or her hand is in the correct spot; if the hand is not in the correct position, the hand that is seen in the virtual environment closes on “air.” The user is essentially in the NASA-constructed virtual world, which could be a simulation of the real world or of another planet, or which could be an entirely unique imaginary environment, or anything in between—all while sitting safely in a computer-filled room on Earth. 2.3.2 Education Educationally, virtual reality is used to recreate scenes from the past and immerse the student in the scene. Programmers can include both physical, historical elements and 8, 1992) (stating that in addition it laid off virtually all of its staff, and its founder, Jaron Lanier, left the company). 38   Kurt Wagner, VR Still Isn’t Mainstream, but Oculus Thinks Its New $200 Headset Could Fix That, Recode (Oct. 12, 2017), http://www.recode.net/2017/10/12/16463464/oculus-facebook-vr-virt​ ual-reality-headsets-rift-santa-cruz-go-cost-samsung. 39   See Sandra K. Helsel, Beyond the Vision: The Technology Research and Business of Virtual Reality (Meckler Pub. 1992) (citing T.B. Coull, Texture-Based Virtual Reality on a Desktop Computer using WorldToolKit). 40   See Scott Fisher, Virtual Environments: Personal Simulations and Telepresence, at 105–8, in Sandra Helsel & Judith Roth, Virtual Reality: Theory, Practice, and Promise (Meckler Publishing, 1991).

BARFIELD PRINT.indd 162

13/11/2018 11:52

Virtual copyright  163 people or can create an interactive “expert system.” The simulation would respond to questions asked by students. By interacting with the scene, the student gains and retains more knowledge than if the information were read in a book. Also, virtual reality applications can be tailored to model different physical settings. Thus, students can learn geography by walking through the area being studied. 2.3.3 Entertainment Virtuality is used in a variety of entertainment media. In addition to the previously mentioned Star Trek, the movies Lawnmower Man, Virtuosity, Strange Days, Johnny Mnemonic, The Matrix, The Congress, Jurassic Park, TekWar, and Wild Palms featured virtual reality. In William Shatner’s TekWar books, Tek is a virtual reality program that acts as a drug. Images shown are two-dimensional, but the audience can imagine what it would be like to be immersed in such a situation. Chairs with special eye goggles and speakers are available to enhance each virtual reality experience. In addition, amusement parks, called Cinetropolises, are being designed to offer a myriad of “rides” in multimedia and three-dimensional virtual reality applications.41 Disneyland’s Star Tours, for example, is always a big hit. But the emergence of affordable goggles has brought entertainment to the third dimension. Movies are now routinely shown in 3D in theaters, and numerous movies and animations have designed for the VR experience. What separates goggles from the movies is the ability to look up, down, and even around and see the movie from all angles. Indeed, regular television programming, including sporting events, is now displayed through goggles. 2.3.4  Computer games Computer games are currently the most popular form of virtual reality. In the late 1980s Mattel introduced the PowerGlove—based on VPL’s DataGlove—which allowed the user to use hand movements to control Nintendo games.42 The PowerGlove sensed motion in the fingers as well as overall hand motion to perform its tasks as a game controller. As noted previously, modern gaming offers a variety of virtual inputs and outputs using gloves, wireless, goggles, and immersive environments. 2.3.5  Medical and professional Medically, virtual reality can be used to train doctors. Using sensitive gloves and goggles, doctors can simulate difficult operations before undertaking them. Doctors can also explore circulatory systems via an exploratory virtual reality “vehicle.” Also, they can practice surgery techniques on virtual patients before stepping into an operating room. Professionally, virtual reality is used to improve people’s ability to do their jobs. One application is the “walkthrough” design of homes. The customer can literally simulate a walkthrough and examine a home before it is built. During the walkthrough, the customer can grab and move items to new locations. Another use of virtual reality is the recreation

41   See generally Ronald Grover, Where Buying a Ticket Puts You Right in the Action, Time Magazine, at 73–75 (Mar. 7, 1994). 42   The PowerGlove represented a low-cost version of VPL’s DataGlove and was developed in part by VPL, among others.

BARFIELD PRINT.indd 163

13/11/2018 11:52

164  Research handbook on the law of virtual and augmented reality of crime scenes by lawyers in court. These applications are based on new types of authorship that will compete for broad legal protection under copyright and other law. 2.4  The Reality Trend Thus, since the early days of the virtual, the trend has been in one direction: toward reality. While there are many animated VR experiences, even the animation has become more realistic. Pixar is careful, for example, to ensure that its computer-designed images are not quite real, because they could be. Similarly, the Star Wars franchise received great criticism for portraying a dead actor using CGI animation in a nonanimated movie. Such is the trend toward reality. TED talks now tout VR’s mimicry as a tool for creating empathy,43 participating in real events half a world away,44 or realistically reporting the news.45 The creation of these works will vary—some will be real images and some will be animation, both realistic and fanciful.

3.  COPYRIGHT PROTECTION Copyright subsists in original works of authorship; virtual reality applications will almost always include original authorship, either in animated images or in the selection of videography of real items. The important issues in the protection of virtual reality applications, however, will be the scope of copyright protection, the degree of copying required before infringement will be found, and the impact of the fair use doctrine on the creation and reproduction of virtual worlds and virtual objects. Here, we focus on competing worlds rather than infringement by users. User infringement will have its own special issues, such as immunity as third party providers storing content at the behest of consumers.46 Generally, existing statutory and case law will readily extend protection to virtual worlds and even to particular original virtual objects, but the very nature of virtual reality requires that subsequent participants be allowed greater freedom to adapt, modify, and extend existing virtual worlds and existing virtual objects without liability for infringement except in cases where: (1) strikingly similar or nearly identical copying occurs for virtual worlds and virtual objects that simulate the real world and real objects; or (2) substantial similarity exists for unique virtual worlds and unique virtual objects. To understand how copyright protection applies to virtual reality, it is important to 43   Chris Milk, How Virtual Reality Can Create the Ultimate Empathy Machine (TED Talks, Mar. 2015), http://www.ted.com/talks/chris_milk_how_virtual_reality_can_create_the_ultimate_­empathy​ _machine. 44   Chris Milk, The Birth of Virtual Reality as an Art Form (TED Talks, Feb. 2016), http://www. ted.com/talks/chris_milk_the_birth_of_virtual_reality_as_an_art_form. 45   Nonny de la Peña, The Future of News? Virtual Reality (TED Talks, May 2015), http://www. ted.com/talks/nonny_de_la_pena_the_future_of_news_virtual_reality. 46   Mark A. Lemley & Eugene Volokh, Law, Virtual Reality, and Augmented Reality, 166 U. Pa. L. Rev. 1051, 1111 (2018). Under the Digital Millennium Copyright Act, providers are generally not liable for infringement of their users, so long as they remove infringing content once they find out about it.

BARFIELD PRINT.indd 164

13/11/2018 11:52

Virtual copyright  165 understand generally the terminology associated with virtual reality and the copyright principles applied to computer software. 3.1  Key Definitions Virtual worlds are composed of virtual objects. A virtual object is any reproducible part of the virtual environment, whether the user has interacted with it or not. Thus, the light cycles in Tron and the aliens on Star Trek: The Next Generation’s holodeck are virtual objects, as are cars, plant flowers, clouds, and even the background sky in any modern simulation. Virtual works are composed of virtual objects and the ways they interact. More technically, a virtual work is fixed in a tangible medium of expression, namely some form of recording of a sequence within a virtual world. Included in this definition of virtual works could be the following: computer code that creates virtual environments, the audiovisual presentation to the user, the interactive media including tactile components of the environment that are experienced by the user, and video recordings of audiovisual components to be played in goggles.47 Analysis of copyright protection of virtual reality requires understanding of the copyright principles involved in the computer industry generally. Virtual objects, worlds, and works consist of both an audiovisual portion and an underlying source code portion. Generally, the audiovisual portion is conceived and the source code portion is written so that when interpreted by a computer it presents the audiovisual portion. Of course, the audiovisual portion may be captured in computer code with the aid of machines: digital cameras, scanners, or animation and CAD software. The audiovisual portion, when combined with the way it is manipulated by the user, is called the user interface. The user interface itself will generally contain its own “look and feel” in which all the portions work together; this combination may also be original authorship. 3.2  General Principles The following principles govern the copyrightability in computer programs, including their user interfaces. First, original expression is protected by copyright.48 What is “original” is determined by the answer to two questions: First, has the work been independently developed, that is, is the work not simply copied from another source? Second, even if independent, does the work have more than a de minimis level of creativity, that is, is the work more than “garden variety,” not “so mechanical or routine as to require no creativity whatsoever,” 47   The movie Lawnmower Man is an example of video recording of virtual reality. As a side note, Lawnmower Man included virtual environments in which two people had “virtual contact” within a virtual environment. Such technology, when used in separate places and with appropriate tactile sensations, would revolutionize telecommunications. The idea has since been used in other stories, such as the television show Heroes. 48   See Feist Publications, Inc. v. Rural Telephone Service, 499 U.S. 340 (1991) (describing the requirements for copyrightability).

BARFIELD PRINT.indd 165

13/11/2018 11:52

166  Research handbook on the law of virtual and augmented reality not “devoid of even the slightest trace of creativity,” and not one “in which the creative spark is utterly lacking or so trivial as to be virtually nonexistent”?49 Second, ideas alone, including functional aspects of code, when not patented under federal law and when not maintained as a trade secret under state law, may not be protected and cannot be protected by copyright or trademark law.50 What is “idea” and what is “expression” in particular computer programs is generally determined on a case by case basis. Sometimes the idea and the expression cannot be separated, thus merging into one; the merger doctrine disallows protection in such cases. Furthermore, industry standards and other common elements will sometimes be considered unprotected under the copyright scènes a faire doctrine.51 The difference between protected original expression and unprotected idea will depend on which of three judicial approaches, as further discussed below, is adopted by the court undertaking the analysis in each case. Under the above principles, “look and feel” protection extends to all originality in a computer program’s user interface, including the original expression of the selection, organization, and presentation of user-visible functions, whether in menus, graphical displays, or other visual or sensory metaphors.52 The general principles just recited, as well as the underlying federal statutes and the cases interpreting them, all support this treatment. Though the holdings and rationales have fluctuated, both the Copyright Office and the courts have extended the copyright in computer programs to the look and feel of computer programs generally. The development of the Copyright Office position and the leading cases is instructive and will likely act as important precedent to extending copyright to virtual works and virtual objects. 3.3  Copyright Office Policies on Software-Generated Visual Displays In its history, the US Copyright Office has taken several differing positions in applying copyrights to visual displays of computer software. The Office originally allowed registration of visual displays as separate audiovisual works (on Form PA) or as separate literary works (on Form TX). Forms PA and TX (Copyright Office designations) are similar, but cover different types of works: audiovisual and literal. The Office then changed its position and held that the registration of the underlying program also protected all elements of its audiovisual display. The Copyright Office held public hearings starting on September 9, 1987 to obtain comments and recommendations on how it should proceed in this area. At the time, of

  See generally Feist Publications, 499 U.S. at 340.   Baker v. Selden, 101 U.S. 99 (1880). 51   Apple Computer, Inc. v. Microsoft Corporation, 799 F.Supp. 1006, 1022–1026 (N.D. Cal. 1992) (specific implementation of desktop metaphor is protected, though other implementations of the metaphor are not infringing). 52   A single user-visible command should not be protected; however, the view that an entire collection of user-visible functions is beyond legal protection has mixed support in the case law. As with the original selection, organization, and presentation of data, the original selection, organization, and presentation of such commands may be protected by copyright post-Feist. See Kregos v. Associated Press, 937 F.2d 700 (2nd Cir. 1991). 49 50

BARFIELD PRINT.indd 166

13/11/2018 11:52

Virtual copyright  167 course, virtual reality technology was brand new and not at the forefront of discussion. However, as with much of the Copyright Act, considerations of what future technology might bring were key. At these hearings, three separate views emerged as to the proper procedure to be followed in computer software registration cases. To some degree these views also represented the conflict within the industry and academy on the broader question of how the Copyright Act should be interpreted in computer program cases. 3.3.1  Single registration argument The majority view, supported by many industry associations, recommended that the Copyright Office continue to allow a single registration to cover the entire work, including visual displays. Under this view, the registrant would also have the option of submitting additional deposit material setting forth the visual displays. A modified version of the majority view suggested that the Copyright Office develop a new registration form specifically designed for computer software applications, one which would more specifically cover registration of all aspects of the originality embodied in a computer program. The majority position was supported by the need for a uniform Copyright Office practice, by the need to not prejudice existing registrations, and by the desire to allow for flexibility in the future. 3.3.2  Separate registration argument A second view, supported mainly by Apple, argued for a requirement of separate registration applications for visual displays and underlying program code. The position was supported by the desire to have independent registrations as evidence of independent copyright protection for visual displays that may exist across differing computer program code and differing computer hardware. In this manner, it was argued, it would be clearer that an infringement of visual displays can occur independent of any infringement of the underlying program code and independent of the computer hardware on which the substantially similar visual displays were performed. 3.3.3  No registration argument A third view, supported by some user-oriented associations, contended that the Copyright Office should not allow any registration of visual displays of computer software. It was argued that such displays are generally functional and therefore not copyrightable because they are either ideas or methods of operation. In response, the Copyright Office noted that each case must be examined separately and that Congress and the courts (and not the Copyright Office) ultimately determine the eligibility and ineligibility of subject matter for copyright and the scope of copyright in particular works. The Copyright Office concluded that original works of authorship embodying expression are properly the subject matter of copyright regardless of whether such works are presented in the form of visual displays or computer programs. In June of 1988, the Copyright Office issued a report and internal guidelines that upheld a single registration application practice for protection of all aspects of a computer program’s originality, including visual displays.53 Whether the registration application is

53   See 36 Patent, Trademark & Copyright Journal (BNA) 151, 155 (1988); Copyright Office Circular 61, at 4–5 (Sept. 2017), http://www.copyright.gov/circs/circ61.pdf.

BARFIELD PRINT.indd 167

13/11/2018 11:52

168  Research handbook on the law of virtual and augmented reality made on Form TX (for literary works) or Form PA (for audiovisual works) turns on what aspect of the work predominates. If text predominates then Form TX is to be used and if graphical displays predominate then Form PA is to be used. In either case, however, the applicant can submit deposit material to support the areas where protection is sought. By this mechanism, the Copyright Office in effect confirmed, at least at the examination and registration levels but for practical purposes even further, that copyright protection can subsist beyond the literal program code for the computer software to the look and feel of the computer software, and perhaps even beyond that. 3.3.4  Copyrightability of virtual reality applications Given the above principles and Copyright Office policies, there is little question that copyright protection will extend in some form to virtual works and even to virtual objects. Virtual reality applications are computer programs, and the portion that interacts with the user is indeed a user interface whose “look and feel” is protectable if original, and if beyond some constitutionally mandated de minimis level of creativity under Feist.54 If a group of rectangles can be protected in a breakout game,55 then a just-about level of creativity will suffice. The copyright protection subsists “in original works of authorship fixed in any tangible medium of expression, now known or later developed, from which they can be perceived, reproduced or otherwise communicated, either directly or with the aid of a machine or device.”56 There is no question but that original virtual objects, virtual works, and virtual worlds—all original authorship in a new medium of expression—easily fit within this statutory definition. Federal courts recognize that the statutory definition is flexible enough to permit extension of copyright principles to originality embodied in or otherwise expressed as part of virtual reality applications. From a practical point of view, the Copyright Office policies mean that virtual reality developers can complete copyright registration by registering the virtual work with the Copyright Office on a single form. Form PA should generally be used, because the audiovisual portion of virtual reality will typically predominate over the rest of the work. In any case, developers should submit both source code and interface deposits to fully avail themselves of the benefits associated with registering a copyright.57 3.3.5  Arguments in favor of broad copyright protection Virtual reality applications are most closely analogous to video games (as to interactive aspects) and computer programs (as to underlying implementing technology). Case law in

54   See generally Feist Publications Inc. v. Rural Telephone Service Co. 499 U.S. 340 (1991) (noting the requirements for copyrightability). 55   Atari Games Corp. v. Oman, 979 F.2d 242, 247 (D.C. Cir. 1992); see also Tetris Holding, LLC v. Xio Interactive, Inc., 863 F. Supp. 2d 394 (D.N.J. 2012) (extending protection to blocks in Tetris game). 56   17 U.S.C. § 102(a). 57   Even after the Berne Convention (which eliminated the registration requirement), such benefits include avoiding a claim of innocent infringement, obtaining a presumption of validity, and shifting the burden of producing evidence to the defendant with regard to the originality and the existence of de minimis creativity in the copyrighted work.

BARFIELD PRINT.indd 168

13/11/2018 11:52

Virtual copyright  169 these areas is well settled and can be readily applied to determine the scope of copyright protection in virtual reality. In Stern Electronics v. Kaufman,58 the Second Circuit determined that despite the fact that each game was different (depending on what the user does), a video game was original because of the “repeated appearance of the same sequence of numerous sights and sounds in each” different experience.59 The same court held that both the audiovisual display and the underlying computer code were protectable by copyright and the audiovisual display is itself original even though it is created by underlying computer code. As the Stern court put it, originality exists when “someone first conceived what the audiovisual display would look and sound like.”60 In addition, virtual realities (and the virtual objects, works, and worlds within them) are each generally “sufficiently permanent or stable to permit it to be perceived, reproduced, or otherwise communicated for a period of more than a transitory duration.”61 Because each experience in each virtual world—primarily due to interaction with particular virtual objects and the user’s creation of new virtual objects and new virtual works in each session—may be different, it is more difficult to claim permanence or stability of virtual reality in the traditional sense than is the case for video games. However, the virtual environment is generally fixed in some state before and during every use, even if a particular user does not see or hear certain aspects of it. As discussed by the court in Stern, this level of permanence and stability is sufficient to allow protection. In addition, both the audiovisual components and the computer program components are each protected by copyright. Thus it is possible to infringe an audiovisual work with program code that is different than the original, and it is possible to copy a computer program while not infringing an audiovisual copyright. 3.3.6  Arguments against broad copyright protection Virtual reality applications, however, are different in many ways from both their computer software and video game predecessors. Such differences cause reassessment of currently well-settled issues. For example, virtual objects generally simulate the properties of real objects. Virtual worlds generally simulate real worlds. Virtual works generally simulate sequences and interactions experienced in real life. At the same time, however, virtual objects, virtual worlds, and virtual works may be subject to user adaptation, modification, and recreation. In Feist Publications Inc. v. Rural Telephone Service Co.,62 the US Supreme Court affirmed that “pure” fact cannot be copyrighted and collections of facts, if not beyond some de minimis level of creativity (even if original), also cannot be copyrighted. Thus, it could be argued that the more realistic a virtual object, a virtual work, or a virtual world becomes, the less protection will exist under copyright laws. When a virtual object, work, or world becomes as real as the real-life item it is designed to simulate, then the environment would merely be an exact copy of reality. If reality is considered a “fact”     60   61   62   58 59

669 F.2d 852 (2d Cir. 1982). Id at 856. Id at 856. 17 U.S.C. §101; MAI Systems Corp. v. Peak Computer, Inc., 991 F.2d 511 (9th Cir. 1993). See Feist Publications Inc. v. Rural Telephone Service Co., 499 U.S. 340 (1991).

BARFIELD PRINT.indd 169

13/11/2018 11:52

170  Research handbook on the law of virtual and augmented reality within the meaning of Feist, then no copyright protection would be allowed, at least for the audiovisual portions that mimic reality or “fact” identically. This is not a new concept; courts foresaw this concern in the 1990s.63 Copyright would still protect the underlying original computer software that was implementing the particular virtual object, virtual work, or virtual world, but it would not protect that which is indistinguishably derivative of the real world. Consider, for example, Gracen v. Bradford Exchange,64 in which Judge Posner ruled that the winner of a contest to best capture the essence of Dorothy from the Wizard of Oz could not claim a copyright in the painting. It was too difficult, the opinion argued, to tell whether someone copied from the painting or from the original. In a later case, Posner seems to step back from the rule in Gracen, but the new application might still preclude copyright of realistic virtual objects: “When . . . a work . . . is based on work in the public domain, the only ‘originality’ required for the new work to be copyrightable (the very term is a misnomer) is enough expressive variation from public-domain

Figure 5.5  MeshWerks wire model with CGI additions

63   Engineering Dynamics, Inc. v. Structural Software, Inc., 26 F.3d 1335, 1342 (5th Cir. 1994) (“In some computer programs, the user interface may merge almost wholly with the expression, processes, or ideas embodied in the program—voice-activated or virtual reality programs or those attuned to the human heartbeat furnish some examples that may trouble courts in the future”). 64   698 F.2d 300 (7th Cir. 1983).

BARFIELD PRINT.indd 170

13/11/2018 11:52

Virtual copyright  171 or other e­ xisting works to enable the new work to be readily distinguished from its predecessors.”65 This general rule has even been extended to three-dimensional models. In Meshwerks, Inc. v. Toyota Motor Sales, U.S.A.,66 the court considered the copyrightability of wire models made to look like Toyota cars. These models were then used as the basis for computer graphics adding details, such as the full shape of the car body, and then the outer features of the car. These graphics were then used in commercials. The court ruled that the wire models were not copyrightable; their original expression, if any, emanated from the actual Toyota car on which they were based. The court did not, however, discuss whether the computer animation based on the real object would be copyrightable. One court that did address such a question considered scaled train models. In Osment Models, Inc. v. Mike’s Train House, Inc.,67 the court considered whether a model train station designed to look like a real train station, but not scaled to exact measurements, constituted sufficient originality to be copyrightable. The court decided that it was a question of fact for the jury to decide, but that such small changes in a realistic model might well be protected. 3.3.7  Balancing the competing arguments The competing positions can be balanced with reference to the particular virtual work at issue. A virtual work is not merely a direct copy of reality; it is an expression of reality conveyed through an author’s vision of what reality includes and how it behaves. The ability to render a realistic image into an electronic expression is at least as creative as rendering a realistic image into a photograph. Even photographs that capture minimal expression have always been subject to copyright protection regardless of the extent of automation of photographic equipment.68 The same standard will likely apply to virtual reality applications.69 In addition, the computerization of virtual objects, virtual works, and virtual realities provides opportunities for creative extensions to real-life objects, works, and realities. Virtual worlds can and often do allow activities which seem real, but are not 65   Bucklew v. Hawkins, Ash, Baptie & Co., LLP., 329 F.3d 923, 929 (7th Cir. 2003); accord L. Batlin & Son, Inc. v. Snyder, 536 F.2d 486 (2d Cir. 1976). See also Lewis Galoob Toys, Inc. v. Nintendo of America, Inc., 964 F. 2d 965 (1992) (minor modifications by user gameplay not a creation of derivative works). 66   Meshwerks, Inc. v. Toyota Motor Sales USA, Inc., 528 F.3d 1258 (10th Cir. 2008). 67   Case No. 2:09-CV-04189-NKL (W.D. Mo. Dec. 27, 2010). 68   Burrow Giles Lithographic Co. v. Sarony, 111 U.S. 53 (1884) (holding that copyright protection can be extended to photographs); but see Bridgeman Art Library, Ltd. v. Corel Corp., 36 F. Supp. 2d 191 (S.D.N.Y. 1999) (“slavish” reproduction of museum art not protected by copyright). 69   See Atari Games Corp. v. Oman, 888 F. 2d 878, 883 (D.C. Cir. 1989) (“The level of creativity necessary and sufficient for copyrightability has been described as ‘very slight,’ ‘minimal,’ ‘modest’”); Berkla v. Corel Corp., 66 F. Supp. 2d 1129 (1999) (granting protection to photorealistic images); see also Thomas Wilson & Co. v. Irving J. Dorfman, 433 F.2d 409, 411 (2nd Cir. 1970); cf. Star Athletica, LLC v. Varsity Brands, 137 S. Ct. 1002, 1010 (2017) (“Nor could someone claim a copyright in a useful article merely by creating a replica of that article in some other medium—for example, a cardboard model of a car. Although the replica could itself be copyrightable, it would not give rise to any rights in the useful article that inspired it”).

BARFIELD PRINT.indd 171

13/11/2018 11:52

172  Research handbook on the law of virtual and augmented reality possible in the real world. For example, virtual worlds can simulate the inner workings of the human body; in this environment, the user could tour the circulatory system as if he or she were a blood cell. Also, the user may “fly” through a very realistic copy of a city. It is the extension of reality that truly gives virtual works their creative expression—even if a true rendering of reality were not considered creative. For these reasons, notwithstanding Feist and arguments that can be made from it, virtual realities, including virtual worlds, works, and objects, will all receive some protection by copyright. The scope of protection will vary depending on the approach taken by the court in balancing the competing interests of creative incentives and competitive value.70 These same issues have been addressed in the copyright “look and feel” cases which have been decided over the last thirty years.

4.  COPYRIGHT “LOOK AND FEEL” IN THE COURTS While virtual reality applications can be copyrighted as computer software and are copyrightable even though they are distinct from standard computer software, the type of protection afforded to virtual reality applications can be interpolated from related decisions in precedential “look and feel” cases. The three views expressed in the Copyright Office roughly correspond to the views also found in the conflicting judicial opinions addressing these issues: Like Apple’s position in the Copyright Office, one body of case law provides broad protection; as suggested by some of the user associations in the Copyright Office hearings, another would provide only limited or narrow protection; and, somewhat similar to the majority view upheld by the Copyright Office, the most recent cases appear to take a middle-ground approach of protecting certain user interface elements on a case by case basis. But there remains tension in the law, as some courts vary on whether they should consider a work on its own merits or only in relation to how others might want to reuse them.71 Because no US Supreme Court case has yet directly resolved the conflicting decisions (the court was evenly divided the last time it came up in the late 1990s), they must each be applied separately to virtual reality to examine the range of possibilities in the courts. 4.1  Cases Favoring Broad Protection Several cases have favored broad protection of computer software. In general, their rationale is that the “idea” of a computer program can be stated narrowly and everything else may be properly considered “expression” subject to copyright protection. These cases have established several grounds for copyright protection.

70   See, e.g., Berkla v. Corel Corp., 66 F. Supp. 2d 1129 (E.D. Cal. 1999) (awarding protection to photorealistic plant images, but requiring virtual identity to show copyright infringement). 71   Oracle America, Inc. v. Google Inc., 750 F.3d 1339, 1358 (Fed. Cir. 2014) (discussing circuit divide over abstraction-filtration-comparison test).

BARFIELD PRINT.indd 172

13/11/2018 11:52

Virtual copyright  173 4.1.1  Structure, sequence, and organization Though not strictly a “look and feel” case, the decision in Whelan Associates, Inc. v. Jaslow Dental Laboratory Inc.72 extended the scope of copyright protection for computer programs and provided a foundation for many later software cases. In Whelan, the Third Circuit upheld the trial court’s finding of copyright infringement based upon the defendant’s copying of the “structure, sequence and organization” of plaintiff’s software. While recognizing that copyright protection does not extend to the idea or functionality of the program, the court found that similarities in the file structures, screen outputs, and certain subroutines, while not comprising a majority of the total number of lines of code in defendant’s software, were similarities in expression, and therefore constituted copyright infringement.73 In so finding, the court set forth a test for determining the line between unprotectable ideas and the protectable expression of those ideas; this test favors finding protectable expression: “[T]he purpose or function of a utilitarian work would be the work’s idea, and everything that is not necessary to that purpose or function would be part of the expression of the idea.”74 The Whelan decision not only extends copyright protection beyond literal copying of the structure, sequence, and organization of that program code itself, but also formulates a broad test that limits unprotectable ideas to only those “necessary to the program’s purpose or function”; everything else is or could be protectable expression. Compatibility, familiarity, ease of use, user training, and “industry standards” play no obvious role in this analysis. Nearly 30 years later, in Oracle v. Google,75 the Federal Circuit considered whether a collection of Java command names organized into “packages” was subject to copyright protection. Google had used these command names to create a compatible software interpreter/compiler. In doing so, Google did not copy the underlying implementation code. Instead, it implemented the same functionality using its own code, with the only remaining overlap being the name and package organization of the commands. Thus, if a programmer wanted to see if a file exists, it could call the io.file.exists() method in either Java or in Google’s compatible program. Both would check whether the file exists, but in their own ways. The district court had ruled that the names of the programs merged with their function. Specifically, one could not have a compatible API without using these function names.76 Further, the court ruled that the names were an uncopyrightable method of operation.77 On appeal, the Federal Circuit reversed, holding that the structure, sequence, and organization showed the minimal creativity required, and that the names themselves did not merge with their function. Any names could have been chosen at the time they were

  797 F.2d 1222 (3d Cir. 1986).   See id at 1238–9 (discussing aspects of software as a form of “expression” and not just functional elements). 74  See id at 1236 (defining test to apply to determine what aspects of a work are “expression” and what aspects of a work are functional). 75   Oracle America, Inc. v. Google Inc., 750 F.3d 1339. 76   Oracle America, Inc. v. Google Inc., 872 F. Supp. 2d 974, 1000 (N.D. Cal. 2012). 77   17 U.S.C. § 102(b). 72 73

BARFIELD PRINT.indd 173

13/11/2018 11:52

174  Research handbook on the law of virtual and augmented reality selected, it reasoned, and the fact that the same name was required to maintain compatibility did not mean that the original names lacked creative expression. Furthermore, the court found that protection was not barred as a method of operation, because some creative expression existed in the names and variable definitions that were not essential to its operation. The case was remanded for trial on whether Google’s use of the packages was a fair use. The jury agreed that it was; despite the protection of the commands, Google was not liable. Finally, some cases have considered dynamic nonliteral elements. For example, in MDY Industries, LLC v. Blizzard Entertainment, Inc.,78 the Ninth Circuit affirmed the ­copyrightability of server-based elements of World of Warcraft, including how all of the screens, players, stories, and data interacted to make a single virtual world. Similarly, in Micro Star v. Formgen, Inc.,79 the court found that map layout files were derivative works owned by the software developer, or “sequels” integral to telling the Duke Nukem story. The court so held even though the files were created by users, merely described where game elements should be placed, and did not actually include game graphics. As applied to virtual reality, Whelan v. Jaslow, Oracle v. Google, and MDY v. Blizzard stand for the proposition that all virtual works are protectable, including the combination of gestures and other commands used to manipulate such works. Under these cases, absent a fair use, no portion of a virtual work, including the structure, sequence, or organization of events, can be used in any other work without infringing. For example, in a virtual representation of a city, the “idea” of the work would be representation of a city, and that would not be protectable. However, the first developer to create a representation of a certain city would literally “lock out” any other similar representations of the same city through the broad copyright protection. Indeed, under Whelan, once access is shown, it is extremely difficult to rebut a finding of infringement based on any similarities between the two works at any level beyond the underlying idea for the first work. 4.1.2  Multiple visual display sequences In Broderbund Software. Inc., v. Unison World, Inc., the federal district court in San Francisco applied the Whelan principle to the protection of the program’s visual displays.80 While the plaintiff originally made claims pertaining to its greeting card application program called “The Print Shop,” the only question at trial considered infringement of the audiovisual copyright in the screen displays.81 The Broderbund court applied the Whelan case to protect audiovisual copyrights in computer programs beyond their literal displays. The court held that the structure, sequence, organization, and naming of visual displays and associated commands, as well as the layout on the screen, were protectable. Under Broderbund, all portions of a virtual work can be protected by a single copyright   629 F.3d 928 (9th Cir. 2010).   154 F.3d 1107 (9th Cir. 1998). 80   See generally Broderbund Software, Inc., v. Unison World, Inc., 648 F.Supp. 1127 (N.D. Cal. 1986) (applying the Whelan principle to a program’s visual displays). 81   See id at 1129 (noting the issues that were brought to the court for it to decide). 78 79

BARFIELD PRINT.indd 174

13/11/2018 11:52

Virtual copyright  175 registration.82 In addition, the placement of different virtual objects in a virtual world can be protected. For example, one world may place an apple tree on the left portion of the viewer display, a raspberry bush in the middle, and a lake on the right side of the display. Broderbund could be applied to hold that a virtual world with an orange tree on the left, a strawberry bush in the middle, and a pond on the right is infringement where access is proven and there is no proof of independent development of the competing work. 4.1.3  Single visual display and command names In Digital Communications Associates, Inc. v. Softklone Distributing Corp.,83 the court found that plaintiff’s Crosstalk telecommunication product’s single status screen, even though just one display of many setting and command names, was protectable expression and was copyrightable as an arrangement and compilation of such information independent of its underlying program code. The Digital Communications case, applied to virtual reality, could be extended to hold that a simple, static virtual work is protectable. For example, a virtual airport which features a text-based television monitor, similar to those found in actual airports, from which the user may select a destination to “fly” to might be protected against a similar listing of the same cities, and even a similar display listing different cities. Indeed, under Digital Communications, any static display, if original, is protected—even if its content is otherwise factual.84 In sum, the overall thrust of the above decisions is that, in theory, voice, audio, touch, and other forms of sensory perception, if original and if fixed in some form of expression, can be protected by copyright if, in one court’s words, “the overall structure, the order of commands . . . the choice of letters, words, or ‘symbolic tokens’ to represent each command,” and other features may be expressed in a number of other ways.85 Under the above line of cases, copyright is extended broadly without a strong concern for competitive principles, compatibility, familiarity, or industry standards. 4.2  Cases Favoring Competitive Principles Many other cases take a more strictly competitive approach to determine what is copyrightable “expression” in computer software cases. If applied to virtual reality applications, these cases would limit the scope of protection and would give weight to compatibility, familiarity, and industry standards across virtual works and between virtual worlds.

82   See id at 1135 (finding that a one-copyright registration can protect all copyrightable portions of a virtual work). 83   See generally Digital Communications Associates, Inc. v. Softklone Distributing Corp., 659 F.Supp. 449 (N.D. Ga. 1987). 84   Digital Communications, as with Broderbund and Whelan, are all pre-Feist cases; a different result may well have been determined in each of these cases in light of Feist. 85   Lotus Dev. Corp. v. Paperback Software Intern., 740 F.Supp. 37, 67 (D Mass. 1990). While Lotus v. Paperback is likely now overruled in the First Circuit (as discussed later), its analysis is consistent with opinions in other circuits.

BARFIELD PRINT.indd 175

13/11/2018 11:52

176  Research handbook on the law of virtual and augmented reality 4.2.1  Compatible data formats In Synercom Technology Inc. v. University Computing Co.,86 the court issued an early potential blow to copyright protection of visual displays—even though the case itself did not, strictly speaking, deal with the subject of visual displays. Rather, the court addressed whether plaintiff’s sequencing of data input formats for a statistical analysis program could be protected by copyright. The court rejected the plaintiff’s position and held that the sequence and ordering of data input was an unprotected idea.87 The Synercom court analogized the sequence and ordering of the plaintiff’s forms to the familiar “H” clutch pattern used in stick-shift automobiles. The court reasoned that another auto maker would be free to come along and create another car using the same “pattern,” and that to allow such copying would be “socially desirable.”88 For the Synercom court, it was enough that the input formats represented a barrier to competitors who wished to be able to have their programs also use data which had already been formatted for the plaintiff’s program; accordingly, the court gave priority to compatibility and competition. Taken to its limits, the rationale in Synercom would prevent the use of copyright whenever competitors need third party information (such as compatible screen displays or, in virtual reality applications, compatible virtual objects) to take advantage of an installed base of users who have familiarity through user training or general use. For example, a virtual car’s “H” pattern stick shift could not be copyright protected. At the very least, Synercom implies that virtual reality developers can reverse engineer a data storage format to allow users to import virtual objects and even virtual worlds into a competitive program. This is a central defense issue in litigation in recent cases and will be a critical defense to makers of compatible virtual worlds who wish to address the needs of the installed base of existing users. 4.2.2  Industry standard presentations In Plains Cotton Co-op v. Goodpasture Computer Service, Inc.,89 the Court of Appeals for the Fifth Circuit affirmed the district court’s denial of a preliminary injunction where plaintiff sought to enjoin, prior to trial, “organizational copying” that defendants submitted was due to the methods of the cotton industry for which the program was designed.90 In Plains Cotton, the plaintiff, an agricultural cooperative, had developed computer software called the “Telcot system.” The Telcot system operated on a mainframe 86   See generally Synercom Technology Inc. v. University Computing Co., 462 F.Supp. 1003 (N.D. Tex. 1978). 87   See id at 1013–14 (holding that sequence and input of data was not an expression that should be afforded copyright protection, but an unprotected idea that may be used by predecessors). 88   See id at 1013–14 (“The pattern (analogous to the computer ‘format’) may be expressed in several different ways: by a prose description in a driver’s manual, through a diagram, photograph, or driver training film, or otherwise. Each of these expressions may presumably be protected through copyright. But the copyright protects copying of the particular expressions of the pattern, and does not prohibit another manufacturer from marketing a car using the same pattern. Use of the same pattern might be socially desirable, as it would reduce the retraining of drivers”). 89   See Plains Cotton Co-op v. Goodpasture Computer Service, Inc., 807 F. 2d 1256, 1262 (5th Cir. 1987). 90   See id at 1258.

BARFIELD PRINT.indd 176

13/11/2018 11:52

Virtual copyright  177 computer and permitted cooperative members to obtain regularly updated displays of information pertaining to cotton prices and availability. The individual defendants had worked on the development of the Telcot system for plaintiff but did not sign ­confidentiality agreements. The corporate defendant, Goodpasture Computer Service, Inc., hired the individual defendants and put them to work on the personal computer version of the Telcot system. Just 20 days after arriving at Goodpasture, the former Plains employees had completed a design of the personal computer version of the Telcot system, and just several months later, Goodpasture began marketing it under the name “GEMS.” The district court found that the four former Plains employees had access to the Telcot system, that one of them had brought a diskette containing Telcot system programming design information to Goodpasture, and that at least one Telcot subroutine had been directly copied into the source code for GEMS. Nevertheless, the district court refused to enter a preliminary injunction because of its finding that the underlying program code for the two programs had not been copied. The appellate court affirmed, and in doing so made clear that it was not prepared to embrace a broad scope of protection for software copyrights.91 While finding that defendant’s program was “very similar to Telcot on the functional specification, programming and documentation levels,” the Court refused to follow Whelan and instead held that the structure, sequence, and organization of the plaintiff’s program was not necessarily expression but could, as a factual matter, be found after trial to constitute an unprotectable “idea,” in particular where the structure, sequence, or organization is a standard recognized in a particular industry.92 Similar to Synercom, the Plains Cotton holding lends support for the argument that compatibility, ease of use, user training, and “standards” play an important role in determining the existence and scope of copyright in a particular work. As applied to virtual reality, the Plains Cotton case can be extended to limit protection in certain types of applications. For example, architect and interior design applications may look very similar due to industry needs. Each application would necessarily consist of a virtual house in which the user could walk around; the user could grab and move or resize doors, windows, or furnishings. Under Plains Cotton, a court would likely hold that the s­ imilarities in such programs would be predetermined by industry standards.93 4.2.3  Collection of functions and ideas In Data East USA. Inc. v. Epyx Inc.,94 the Ninth Circuit reversed the finding of copyright infringement by the district court after trial and held that the similarities between the plaintiff’s and defendant’s respective “karate” video games were dictated by the functions

91   See id (stating that the court declined to embrace Whelan and that Plains’ case was more aligned with Synercom’s analogy). 92   See id (“In [Synercom], Judge Higginbotham held that ‘input formats’ of a computer program—the organization and configuration of the information fed to the computer—were ideas, not expressions, and thus were not protected by copyright”). 93   See also Bateman v. Mnemonics, Inc., 79 F.3d 1532, 1547 (11th Cir. 1996); Engineering Dynamics, Inc. v. Structural Software, Inc., 26 F.3d 1335, 1341–42 (5th Cir. 1994). 94   See generally Data East USA, Inc. v. Epyx, Inc., 862 F.2d 204 (9th Cir. 1988).

BARFIELD PRINT.indd 177

13/11/2018 11:52

178  Research handbook on the law of virtual and augmented reality of the game and not by their expression. The Ninth Circuit held that the similarities in the programs stemmed from the “idea” of a martial arts game. The Ninth Circuit’s decision is important for a number of reasons. First, it demonstrates that what is “idea” and what is “expression” is determined factually based on the context of a given work as against the genre of which it is a part with no assurance or certainty of outcome. In a direct extension of the Data East facts to virtual reality, a virtual reality karate application would likely receive much “thinner” copyright protection as against other virtual reality karate applications. A court would likely hold that the nature of a karate game would give rise to similar dress on the game participants, similar “karate moves” in the external user interface (for example, by using the PowerGlove), and similar “karate moves” executed by the karate participants in the visual display. In summary, as applied to virtual reality applications, the above cases favor a limited protection in order to foster competition. This means that as virtual reality applications emerge, additional “compatible” applications can be expected to arise competing for the same user attention with similar commands, functions, gestures, and sequences;95 in addition, there is uncertainty as to whether infringement has occurred, and if so, the extent of such infringement. 4.2.4  Menu commands A longrunning case involving only the interactive “feel” of a software product provided some guidance. Borland created the Quattro Pro spreadsheet program to compete with Lotus 1-2-3. The program had several user interfaces. The “native” interfaces looked ­different than Lotus 1-2-3. However, the Quattro program had an interface available to users called its “1-2-3 interface” or “emulation interface.” The emulation interface mimicked the menu structure of 1-2-3, which Borland learned from third party books.96 The district court found that Borland’s use of the Lotus interface infringed Lotus’s copyright.97 On appeal, however, the First Circuit held that the menu commands were merely a method of operation, like the play button on a VCR.98 Thus, the menu commands were unprotectable in themselves and in combination. Denying copyright to all commands stands in contrast to both the Mitel and Oracle cases discussed in this chapter, both of which considered and rejected the broad holding of Lotus. Many virtual reality applications use a collection of gestures such as pointing to indicate the direction that a person wishes to move. The Lotus v. Borland decisions 95   See Computer Associates v. Altai, 982 F.2d 693 (2nd Cir. 1992) (“The interest of copyright law is not simply in conferring a monopoly on industrious persons, but in advancing the public welfare through rewarding artistic creativity, in a manner that permits the free use and development of non-protectable ideas and processes”); see also Jon O. Newman, New Lyrics for Old Melody: The Idea/Expression Dichotomy in the Computer Age, 17 Cardozo Arts & Ent. L.J. 691 (1999) (Judge Newman authored Altai). 96   The idea that copying the structure from a book rather than 1-2-3 itself might protect against infringement seems a bit silly in retrospect. 97   See Lotus Development Corp. v. Borland International, Inc., 799 F.Supp. 203, 223 (D. Mass. 1992). 98   Lotus Development Corp. v. Borland International, Inc., 49 F.3d 807 (1st Cir. 1995), aff’d by an equally divided court 516 U.S. 233 (1996).

BARFIELD PRINT.indd 178

13/11/2018 11:52

Virtual copyright  179 could be extended to hold that such a collection of gestures cannot be protected, even if they form part of an interactive, expressive feel of the virtual reality application’s user interface. 4.3  Analytic Approach: Screen by Screen Analysis Some cases have taken a more analytic approach in which audiovisual expression is protected on an element by element or screen by screen basis. This approach correctly discerns those screens and those elements of those screens which are disqualified from copyright protection by doctrines of functionality, merger, scènes a faire, lack of creativity, or lack of originality. At the same time, courts struggle to recognize the overall originality of the combination of the elements on each screen and the sequence of screens, whether each element is original or in the public domain. For virtual reality applications, this means that different virtual worlds will be broken down into separate “screens.”99 On these screens, the court would determine which objects and user input commands or gestures are functional, limited in choice given the application, or unoriginal;100 the court would then extract them from the analysis. The court would at the same time attempt to determine the overall program and the way in which each object and screen, whether protectable or not, is original. This methodology is most concisely encapsulated in the abstraction-filtration-comparison test of Computer Assocs. Int’l v. Altai,101 which expressly split from Whelan’s more protective approach, but it has been adopted in other circuits as well.102 4.3.1  External flow and sequence In Manufacturer’s Technologies, Inc. v. CAMS, Inc.,103 the court analyzed how copyright protection extended to a cost estimating computer program called the “COSTIMATOR.” The defendants, three of which had acted as plaintiff’s sales representatives for the COSTIMATOR program, began development of a functionally similar program to sell for $1000–$2500 (as compared to the $20,000 price for plaintiff’s program). Plaintiff sued for, among other things, copyright infringement based on the similarity of the screen displays in defendants’ software. The district court found copyrightability in the screen displays alone without regard to the underlying program code, and also found that defendants had infringed the visual display copyrights. It relied in part upon the Copyright Office’s decision to register the software. The opinion went on to hold that the display was distinct from the underlying  99   Determining how to break virtual worlds into analyzable groups may be difficult for virtual worlds. Dynamic displays with flowing images do not lend themselves to a singular set of “screens.” However, courts will nevertheless likely attempt this analysis. 100   See, e.g., Harbor Software, Inc. v. Applied Systems, Inc., 925 F. Supp. 1042 (S.D.N.Y. 1996) (comparing a list of proffered elements for copyrightability to determine which will be compared for infringement); Mitek Holdings, Inc. v. Arce Engineering Co., Inc., 864 F. Supp. 1568 (S.D. Fla. 1994) (same); Ilog, Inc. v. Bell Logic, LLC, 181 F. Supp. 2d 3 (D. Mass. 2002) (same). 101   982 F.2d 693, 706 (2nd Cir. 1992). 102   Gates Rubber Co. v. Bando Chemical Industries, Ltd., 9 F.3d 823 (10th Cir. 1993). 103   See generally Manufacturer’s Technologies, Inc. v. CAMS, Inc., 706 F.Supp. 984 (D. Conn. 1989).

BARFIELD PRINT.indd 179

13/11/2018 11:52

180  Research handbook on the law of virtual and augmented reality code, providing protection for the visual display, even though the underlying computer program may not have been copied.104 Significantly, the court rejected the argument that the external flow and sequencing of the screens for a cost estimating program is dictated by functional considerations. The court pointed to evidence of four different cost estimating programs which all accomplished the same cost estimating functions but without using similar screen displays or sequences of displays.105 The district court concluded that the similarities between the visual displays and the sequence of visual displays constituted an infringement even though there was “no evidence that the defendants had access to the plaintiff’s source code nor did plaintiff put forth any evidence of source or object code similarity.”106 As applied to the previous example about architect and interior design virtual reality applications, the Manufacturer’s Technologies case would examine the plaintiff’s and defendant’s programs, as well as other similar programs, to determine whether such programs must be similar due to functional constraints of the features of the application. In contrast, a court following Plains Cotton would assume that the products in each industry may be similar to foster competition,107 regardless of whether the facts indicate that such programs need not be similar. 4.3.2  Icons and other symbolic representations In 1988, Apple Computer, Inc. filed suit in the district court in San Jose, California against Microsoft Corporation and Hewlett-Packard Company. Apple claimed that both version 2.03 of Microsoft’s Windows operating environment computer software and HP’s New Wave desktop computer software, which operated with Windows, infringed the audiovisual copyrights in Apple’s Macintosh visual displays and images. At a time when such registrations were in doubt, Apple had been quite successful in having the Copyright Office register the Macintosh desktop and application software visual displays containing icons and other “desktop” symbolic representations as audiovisual works and the Macintosh program code as literary works. The registration certificates attached to the complaint showed that Apple attempted to register all aspects of authorship embodied in the Macintosh computer programs, including their audiovisual displays. Apple’s case against Microsoft and HP claimed infringement of its graphical user interface. In an early decision, the district court ruled in Microsoft’s favor, and found that certain screen displays were in fact licensed by Apple to Microsoft under the terms of a written nonexclusive “license to use these derivative works in present and future software programs.”108 Notwithstanding the adverse determination on the license issue, Apple submitted a

  See id at 993.   See id at 994. 106   Id at 1000. 107   Note especially that the Plains Cotton court specifically held that industry standards dictated similar programs, even though direct copying of a portion of the source code was found. 108   See Apple Computer, Inc. v. Microsoft Corp., 709 F.Supp. 925, 927 (N.D. Cal. 1989). This license agreement was the subject of some intense license negotiations between Apple and Microsoft at a time when Apple needed additional software applications from Microsoft. The negotiations are described in John Sculley, Odyssey (1987) (autobiography of former Apple CEO). 104 105

BARFIELD PRINT.indd 180

13/11/2018 11:52

Virtual copyright  181 list of 189 alleged similarities in numerous icon and display categories. Some of these categories included: a trash can; the ability to overlap different windows of information; representation of different objects with different icons; ability to automatically arrange objects on the screen; and the ability to move an object on the screen with a pointing device such as a mouse. In discussing the problem of determining idea and expression, the court noted the “metaphysical line drawing between idea and expression by which courts rationalize their decisions.”109 Further, though the court initially accepted Apple’s argument in favor of the overall “look and feel” protection for the Macintosh user interface—that is, that the arrangement of unprotectable items could be considered protectable110—upon ­reconsideration, in its final decision on the issue, the district court ultimately rejected Apple’s approach and instead held that a collection of unprotectable ideas were not necessarily protectable.111 Specifically, the court held that the idea of a desktop and a windowing system gives rise to certain objects, such as trash cans and individual windows, which may be expected in the interface due to the idea. The Ninth Circuit affirmed this decision, making clear that a collection of features that are ideas, dictated by functionality, or scènes a faire, are limited only to thin copyright protection, requiring virtual identity of the works to show copyright infringement.112 This opinion is a balance between the protective and competitive principles discussed in prior sections. It recognizes that graphical computer programs might obtain copyright protection even if functional, but also limits the infringement analysis to only those expressive elements. Thus, the Apple decision may have the greatest impact on virtual reality applications. It implies that, although it may be desirable to protect any entire user interface, at some point the objects in that interface must be examined for their functional aspects on a case by case and item by item basis. In a literal extension of Apple to virtual reality, one developer may create a virtual work environment, which would naturally include a virtual desk to work at. Such a desk would include a trash can, a calendar, an inbox, an outbox, a to do list, different applications, a clock, and a file drawer. Under Apple, a court may be persuaded, as this court was, to protect the entire interface. But then, however, a court would examine each of the virtual objects on the desktop and determine which objects, such as a trash can, are dictated based on the functional need for disposal. While there are many different types of trash cans in reality, when comparing the designs courts will often hold that the idea of a trash can, without more, is simply rendered unprotectable. 4.3.3  Command presentations In Mitel, Inc. v. IQTel, Inc.,113 the court extended this to apply to four-digit numbers used to control a modem microchip. While the command names were not used, the numbers     111   1992). 112   113   109 110

See Apple Computer, Inc. v. Microsoft Corp., 779 F.Supp.133, 135 (N.D. Cal. 1991). See id at 136. See generally Apple Computer, Inc. v. Microsoft Corp., 799 F.Supp.1006, 1022–6 (N.D. Cal. Apple Computer, Inc. v. Microsoft Corp., 35 F.3d 1435, 1446 (9th Cir. 1994). 124 F.3d 1366 (10th Cir. 1997).

BARFIELD PRINT.indd 181

13/11/2018 11:52

182  Research handbook on the law of virtual and augmented reality represented different descriptions of the status of the controlled microchip. It ruled that, despite the fact that such commands are a method of operation, there may be expressive elements at a lower level of abstraction. Thus, while the method of operation may not be protected, the numbers used to execute that operation might be protected.114 The court did not stop there, however. It went on to consider what was copied and performed a filtration of unprotected items that might have been copied. In doing so, the court went on to hold that a sequential list of numbers lacked originality, but the arbitrary assignment of numbers to a description required at least some creative thought. Finally, the court held that the commands were unprotectable under the scènes a faire doctrine because they were dictated by external demands more than individual creativity. The implications for virtual reality are relatively straightforward. Under this rationale, a sequence of hand gestures will not be an unprotectable idea simply because the gestures are used to move someone in virtual reality. The collection may well be expressive. However, the extent of protection vis a vis accused infringers will depend on how creative—indeed, arbitrary—the gestures are. Use of standard, noncreative gestures (pointing to move in a direction) may be unprotected, but complex movements may garner protection if they are arbitrary rather than driven by outside pressures for standardization. In Brown Bag Software v. Symantec Corp.,115 the Ninth Circuit upheld a summary judgment granted by the district court against a plaintiff seeking protection of the user interface of an outlining software product. Living Videotext developed products called ThinkTank (for the IBM PC) and MORE (for the Apple Macintosh). These products extended the metaphor of outlining and automation through use of a personal computer with dropdown menus at the top and user messages at the bottom of the visual display. After Living Videotext released ThinkTank and MORE, John Friend developed a clone-like competitive product called PC-OUTLINE for Brown Bag. Living Videotext did not initially pursue an infringement claim against John Friend or against Brown Bag. Living Videotext merged into Symantec and Symantec marketed both ThinkTank and MORE. After the merger, Symantec hired John Friend, the former competitor, to elevate ThinkTank and MORE to the next level of outlining software innovation. Friend and Symantec ultimately developed an advanced (for the time)116 outlining product called GRANDVIEW. After GRANDVIEW was released, Brown Bag filed suit in the district court claiming copyright and trade dress infringement of PC-OUTLINE by Symantec’s GRANDVIEW software product. The district court granted summary judgment in favor of Symantec on the copyright claims.117 The district court’s opinion did not address the Lanham Act trade dress issues raised by Brown Bag. The Ninth Circuit affirmed the district court’s ruling on the copyright claims but remanded the action for the district court’s consideration of the Lanham Act claims.   Id at 1372.   See generally Brown Bag Software v. Symantec Corp., 960 F. 2d 1465, 1476 (9th Cir. 1992). 116   Both authors of this chapter were long-time users of GrandView, and consider its simple but powerful DOS interface the gold standard even today. Mr. Russo was the startup counsel for David and Peter Winer and the original incorporating attorney for Living Videotext, Inc., the publisher of their ThinkTank and MORE software products. 117   See Brown Bag Software v. Symantec Corp., 12 U.S.P.Q.2d (BNA) 1191 (N.D. Cal. 1989). 114 115

BARFIELD PRINT.indd 182

13/11/2018 11:52

Virtual copyright  183 The Ninth Circuit’s approach expressly affirms detailed analytical dissection of the copyrightable expression from uncopyrightable features embodied in the user interface.118 In evaluating Brown Bag’s position that the overall look and feel of the programs at issue had to be compared and evaluated, the Ninth Circuit rejected Brown Bag’s position because Brown Bag did not request such an analysis.119 More importantly, however, the Ninth Circuit noted that copyright protection may extend to a new arrangement of otherwise unprotectable elements.120 The Ninth Circuit’s decision restates this two-step approach. First, the court must consider the overall look and feel of the user interface and evaluate whether the selection, arrangement, and presentation of the user interface is original and at least minimally ­creative—even if such selection, arrangement, and presentation is of commands and other functional features which, on their own, are not individually or separately protectable. Second, when comparing the original work with an accused infringer, the court should not consider those similar elements that are unprotectable, and should require nearly identical copying when most of the elements are unprotectable. As applied to virtual reality applications, Brown Bag indicates that the overall virtual experience can be protected. When combined with the Apple ruling, this means that the overall arrangement of a virtual desktop can indeed be protected, even if some parts of it are unprotectable. The balance between incentives for creation and incentives for competition will be made on a case by case and screen by screen basis, and this will also be true for virtual reality applications. While not involving virtual reality, this balance played out in a video game case. In Incredible Technologies, Inc. v. Virtual Technologies, Inc.,121 the court considered competing arcade golf games. It first ruled that the use of a trackball and its accompanying gestures and instructions for use could not be protected against appropriation. It then ruled that any similarities in the video imagery were caused by scènes a faire—they were the types of images one would expect to see in a golf game: golfers, fairways, greens, trees, sky, and so forth. Thus, a similar look and feel to the realistic game was held to be noninfringing.122

5.  STANDARDS FOR DETERMINING INFRINGEMENT It is likely that more exacting standards, such as those used in the most recent look and feel cases, will be used for determining infringement—particularly when virtual objects,

  See Brown Bag, 960 F.2d at 1477.   See id at 1476 (stating that the record fails to include Brown Bag asking the court to make an “overall look and feel analysis”). 120   See id at 1476, n. 4. 121   400 F.3d 1007 (7th Cir. 2005). 122   Compare Kim Seng Co. v. J & A Importers, Inc., 810 F. Supp. 2d 1046 (C.D. Cal. 2011) (similarities in arrangement of two realistic pictures of food explained by the subject matter rather than copying), with Diamond Foods, Inc. v. Hottrix, Llc, No. Case No. 14-cv-03162-BLF (N.D. Cal. Jul. 18, 2016) (providing protection and denying motion to dismiss based on compilation and similarities of unrealistic elements). 118 119

BARFIELD PRINT.indd 183

13/11/2018 11:52

184  Research handbook on the law of virtual and augmented reality virtual works, and virtual realities are closely simulating the real world. In such accurate simulation cases, upon proof of access, more than the traditional substantial similarity will likely be required for proof that there was copying in the first place. Striking similarity or near-identical copying will likely be needed to justify a finding of infringement in cases where virtual realities simulate true real-world realities. In short, similarities that, in the normal course of events, would arise independently in the two works due to a likeness to some real object are not necessarily probative of defendant’s copying from plaintiff’s work. 5.1  Effect of User Participation in Virtual Worlds Unlike ordinary audiovisual works, a key purpose of the most immersive and realistic virtual worlds is to allow user participation, interaction, modification, adaptation, and recreation of virtual objects. Without the user, virtual reality is useless; the same is not true for many other computer applications. This is true even for “viewer only” applications, where the user interacts with the world by a simple head turn. Within that context, authors of virtual objects, virtual works, and virtual worlds expect to provoke and inspire a creative process that will generate additional objects, works, and worlds out of the interaction. Given this goal, a greater showing of similarity should be required to support a finding of infringement, even if a subsequent virtual reality is based upon a preexisting work. Essentially, the new virtual work is protectable for the originality added over a preexisting work,123 but infringement of the preexisting work will be more difficult to prove because in most cases the application is designed to be adapted, modified, and used for the further development of the virtual work. This issue has come to a head in the lucrative video game streaming market. Players can make money by showing the world their good (and sometimes bad) gameplay.124 The question of who owns which portions of the game is still being determined, though providers appear to have agreed that performing simple user gameplay will constitute infringement of an underlying game,125 even if the users might have some additional rights. But where the game is more realistic, and the user modifies it for personal use rather than performance to others, the questions may become more difficult to answer. 5.2  Effect of Idea/Expression Dichotomy Except for wholly original creations, federal courts can be expected to recognize that virtual objects, virtual worlds, and virtual realities that simulate real objects, real works,

123   17 U.S.C. §103(a) (“[P]rotection for a work employing preexisting material in which copyright subsists does not extend to any part of the work in which such material has been used unlawfully”); but see Lewis Galoob Toys, Inc. v. Nintendo of America, Inc., 964 F. 2d 965 (1992) (user participation in work must be sufficiently concrete and permanent to qualify as new derivative work and changes during game play do not constitute new works). 124   See, e.g., https://go.twitch.tv/. 125   Brianna Howard et al., To Avoid Copyright Disaster, the Future of Game Streaming Is Licensing, VentureBeat (Aug. 23, 2017), https://venturebeat.com/2017/08/23/to-avoid-copyright-disaster-the-​ future-of-game-streaming-is-licensing/.

BARFIELD PRINT.indd 184

13/11/2018 11:52

Virtual copyright  185 and real-world realities are themselves based upon preexisting materials, thereby limiting the scope of copyright protection. There is, therefore, a requirement of closer copying before infringement will be found. Though copyright law contains no statutory formula for what constitutes infringement, absent copying of underlying code, the idea/expression dichotomy contained in Section 102(b) of the Copyright Act will likely permit a greater degree of freedom for virtual authors and virtual artists to create new virtual objects, new virtual works, and new virtual worlds based on preexisting virtual reality materials. Because of the nature of virtual reality authorship, federal courts will strive even more rigorously to filter from the infringement analysis any virtual elements that derive directly from unprotected “idea, procedure, process, system, method of operation, concept, principle, or discovery, regardless of form” excluded from copyright protection under Section 102(b) of the Copyright Act.126 The more realistic the virtual reality, the more likely that protection will be limited so that First Amendment freedom is not impaired. “Free speech and right of access to information under the First Amendment protects . . . reproduction efforts . . . When factual information contained in a film is inextricably related to the film’s expression of that information, the use of the film to secure broad distribution of the facts may be permitted.”127 In virtual reality applications, it will be more common to have expression of a virtual object indistinguishable from the factual object it simulates, much as the wire model of the car had no expression distinguishable from the car. In such cases, the courts will have a more difficult task of determining whether copying occurred. Absent striking similarity or admitted copying, courts may not be able to determine whether expression was copied when expression of an object is completely and accurately intertwined with the idea of that object. In that case, the merger doctrine may limit the scope of protection.128 Analogies for this defense exist within the current body of computer law. In NEC v. Intel,129 for example, the court held that many of the similarities observed in two hardware microcode programs were caused by functional constraints of the physical hardware, and were likely to arise even if the programs were created independently. Similarly, idea and expression should be dissected in virtual reality cases. As discussed previously, the entire arrangement of a virtual world, or its look and feel, is protectable, even if the idea of the objects in the world is closely intertwined with the expression of

126   17 U.S.C. § 102(b). See Gentieu v. John Muller & Co., Inc., 712 F.Supp. 740, 744 (W.D. Mo. 1989) (“In the case at bar, the plaintiff, in using her copyrighted expression (lighting, choice of background, camera angle, etc.) has not expanded on the idea of a photograph of a naked baby any more than the sculptor who adds nothing to the idea of a plaster nude statute. This is not meant to minimize the plaintiff’s work in any way; its simplicity is its creativity. However, by utilizing her expression in such a way as to create a naked baby and nothing else the plaintiff limits her copyright protection to the identical copying of the copyrightable elements of the work”). See also Mazer v. Stein, 347 U.S. 201 (1954). 127   Encyclopedia Britannica Educational Corporation v. Crooks 542 F.Supp 1156, 1180 (W.D.N.Y. 1982). 128   Cf. Deven R. Desai & Gerard N. Magliocca, Patents, Meet Napster: 3D Printing and the Digitization of Things, 102 Geo. L.J. 1691, 1707–8 (2014) (arguing that 3D scans of real objects are not protected because they merge with the object itself and add no creative expression). 129   See generally NEC v. Intel, 10 U.S.P.Q.2d 1177 (N.D. Cal. 1989).

BARFIELD PRINT.indd 185

13/11/2018 11:52

186  Research handbook on the law of virtual and augmented reality those objects. At the same time, substantial similarity between specific objects in two different worlds will not necessarily justify a finding of infringement because a single object may not be protectable when expression and idea merge for that object.130 5.3  Application of First Amendment and Fair Use Doctrines Will First Amendment and fair use principles play a larger role in VR-based litigation? Because expression and fact are closely related in virtual worlds and in virtual objects, federal courts will extend Copyright Act fair use principles to permit greater freedom to virtual artists. Fair use is a judicially created rule of reason, originating in the mid-1800s.131 The fair use defense is now codified in Section 107 of the Copyright Act.132 It provides in relevant part that: the fair use of a copyrighted work . . . is not an infringement of copyright. In determining whether the use made of a work in any particular case is a fair use the factors to be considered shall include—(1) the purpose and character of the use, including whether such use is of a commercial nature or is for nonprofit educational purposes; (2) the nature of the copyrighted work; (3) the amount and substantiality of the portion used in relation to the copyrighted work as a whole; (4) the effect of the use upon the potential market for or value of the copyrighted work.133

Thus, the copyright laws provide a four-part test to be used to determine fair use. Notwithstanding the statutory provisions, Congress intended that the federal courts judicially determine fair use on a case by case basis: “Beyond a very broad statutory explanation of what fair use is and some of the criteria applicable to it, the courts must be free to adapt the doctrine to particular situations on a case by case basis.”134 Notable cases have allowed for fair use in parody, artistic reproduction with minor changes, copying of entire works for search purposes, and even reuse of the entire work for different expressive purposes. In summary, the fair use doctrine requires a judicial weighing of the public benefit of additional creativity and additional dissemination of information—the first two factors in Section 107—as against the effect of a substantial taking on the marketability of a preexisting copyrighted work—the last two factors in Section 107. The Supreme Court has recognized that the main purpose of the Copyright Act is for the public benefit, that is, to encourage creation and dissemination of information.135 Thus, when a use is transformative, federal courts give greater weight to the first two “public interest” factors over the last two “private interest” factors affecting authors and their works.136 130   For a detailed discussion of several nonprotectable elements in two graphical operation systems, see Apple Computer, Inc. v. Microsoft Corp., 799 F.Supp. 1006 (N.D. Cal. 1992). 131   See, e.g., Folsom v. Marsh, 9 Fed.Cas. 342, 343 (No. 4901) (C.C.D. Mass. 1841). 132   17 U.S.C. § 107. 133   Id. 134   Key Maps v. Pruitt, 470 F.Supp. 33, 37 (S.D. Tex. 1978) (citing Notes on the Committee of the Judiciary, House Report No 94-1476). 135   See Sony Corp. v. Universal Studios, Inc., 464 U.S. 417 (1984) (noting that the purpose of the Copyright Act is to promote and furnish the cultivation of science and the arts). 136   Campbell v. Acuff-Rose Music, Inc., 510 U.S. 569 (1994).

BARFIELD PRINT.indd 186

13/11/2018 11:52

Virtual copyright  187 5.4  Effect on Virtual Worlds and Virtual Works Federal courts will determine that the public interest in use and extension of virtual reality is important enough to allow copying of substantial parts of virtual worlds and perhaps complete copying of at least some virtual objects (for placement in new virtual worlds) without liability for infringement. Though exact copying of entire virtual worlds likely will not be permitted, courts should allow a greater degree of freedom for virtual artists to incorporate preexisting virtual objects into newly created or substantially enhanced virtual worlds for both internal and commercial distribution. Historically, the fair use defense was not available for infringements in commercial cases.137 In commercial cases, the profitmaking motive and the competitive effect on the plaintiff’s potential market are usually viewed as deciding factors prohibiting application of the fair use doctrine.138 That has changed more recently, however. The Supreme Court ruled that commercial use is not presumptively unfair, especially when the use is transformative. In Campbell v. Acuff-Rose Music,139 the Court held that copying a large portion of the “heart”140 of the song “Pretty Woman” into a rap song was parody and a fair use. The Court held that in order to achieve parody, enough of the original work must be copied so that the new work can be related to the original work. Additionally, the Court ruled that parody did not usurp the market of the original product and was thus eligible as a fair use.141 Since Campbell, courts have applied fair use to protect uses in a variety of nearly direct copies put to different expressive purposes. In Cariou v. Prince,142 the court found that an artist’s modification of another artist’s photographs was fair use, notwithstanding the commercial nature of the art. Similarly, in Blanch v. Koons,143 an artist’s reimagining of a portion of a commercial photographer’s advertisement was protected by fair use. In White v. West Pub. Corp.,144 the court determined that republication of legal briefs filed with the courts, unchanged and in their entirety, was transformative because the use changed from persuasion to legal research.145 Finally, in Bill Graham Archives v. Dorling Kindersley Ltd.,146 the court allowed a direct, though smaller, copy of several Grateful Dead concert posters in an illustrated coffee table book. The court found the copies transformative, notwithstanding the complete use of each work, and thus disregarded the commercial nature, finding instead that there would be no market substitution.147

137   See, e.g., Sony Corp. of America v. Universal City Studios, Inc., 464 U.S. 417, 451 (1984); Walt Disney Products v. Air Pirates, 581 F.2d 751 (9th Cir. 1978). 138   See 3 Nimmer on Copyright § 13.05[A](4). 139   510 U.S. 569 (1994). 140   In this case, the “heart” was much of the music and the first two lines of lyrics. 141   See Campbell, 510 U.S. at 591–2 (1994). 142   714 F.3d 694 (2d Cir. 2013). 143   467 F.3d 244 (2d Cir. 2006). 144   29 F. Supp. 3d 396 (S.D.N.Y. 2014). 145   See also AV Ex Rel. Vanderhye v. iParadigms, LLC, 562 F. 3d 630 (4th Cir. 2009) (finding fair use in copying of entire term papers used for detection of plagiarism). 146   448 F.3d 605 (2d Cir. 2006). 147   But cf. Kienitz v. Sconnie Nation LLC, 766 F.3d 756 (7th Cir. 2014) (criticizing transformative test and considering only whether work is a complement or a substitute).

BARFIELD PRINT.indd 187

13/11/2018 11:52

188  Research handbook on the law of virtual and augmented reality These cases all stand for the proposition that virtual objects could be fairly copied into other virtual works if there were no demonstrable adverse impact on the market for the original work.148 In the short term, copying virtual objects and distributing derivative works could theoretically affect the market of the author who created the virtual object. However, as more and more virtual artists exist and as technology advances to the point that virtual objects are easily created with “camera-like” technology, little impact on markets will be created by some copying of realistic virtual objects. The trend over time would predict that the fair use doctrine will extend to permit greater freedom for the adaptation, use, display, and distribution of virtual works; however, the fair use doctrine will not extend as readily when there is copying of entirely unique or less realistic virtual objects because such copying would more likely affect a potential market and therefore be less likely to fall in the fair use category. Other fair use decisions reinforce this view, though less directly. In Sega v. Accolade,149 for example, Sega accused Accolade of copyright infringement when Accolade reverse engineered Sega’s computer object code to determine how it functioned; the Ninth Circuit held that such reverse engineering is a fair use. In reaching its decision, the Sega Court held that there is public benefit in the growth of creative expression. The Sega Court additionally held that while the author of the original work may suffer some loss of profits, the benefit of the growth of expression and entry into the market of others outweighs a small decrease in profits.150 As applied to virtual reality applications, Sega holds that reverse engineering of a virtual reality application’s object code is generally allowable to create compatible modules of virtual objects for that application. More generally, Sega can be interpreted to allow copying of portions of virtual works to extend overall expression in a new virtual world when the benefits of this outweigh the market impact on the original author’s work. Finally, as libraries of virtual objects are developed and shared online, search facilities will allow other users to find them for viewing and reuse. Like other searching, this will likely be considered fair use.151 This may be made more complicated if developers use open source software to create such objects; reuse of such objects may also violate license agreements.152 The more the copyrighted virtual object or virtual world is based on physical reality, the more easily the fair use defense could be argued; the purpose of the use would be to more closely simulate reality in a creative environment. An analogy is the fair use copying 148   But see Monge v. Maya Magazines, Inc., 688 F.3d 1164, 1176 (9th Cir. 2012) (finding that use of unpublished wedding photos was not transformative expression). 149   See generally Sega Enterprises Ltd. v. Accolade, Inc., 977 F. 2d 1510 (9th Cir. 1992). 150   See id at 1524 (“In any event, an attempt to monopolize the market by making it impossible for others to compete runs counter to the statutory purpose of promoting creative expression and cannot constitute a strong equitable basis for resisting the invocation of the fair use doctrine”); see also Sony Computer Entertainment, Inc. v. Connectix Corp., 203 F.3d 596 (9th Cir. 2000). 151   See, e.g., Authors Guild, Inc. v. HathiTrust, 755 F.3d 87 (2d Cir. 2014); Authors Guild v. Google, Inc., 804 F.3d 202 (2d Cir. 2015); Perfect 10, Inc. v. Amazon.com, Inc., 508 F.3d 1146 (9th Cir. 2007). 152   Aaron Williamson, Software Litigation Opens Pandora’s Box of Key Open Source Issues, 7 Landslide 53 (2015).

BARFIELD PRINT.indd 188

13/11/2018 11:52

Virtual copyright  189 of maps. Maps have limited copyright protection because of the public domain nature of the underlying data.153 As idea and expression become more intertwined in realistic virtual objects, courts will find that the nature of the copyrighted work allows a fair use defense. The more realistic the virtual object is, the more limited the protection will be; when protection is found and even when infringement may be found, the fair use doctrine will provide greater latitude given the nature of virtual reality applications.

6. CONCLUSION There are no “Thought Police” and the Orwellian vision of a dystopian future has not come to fruition. The continued revolution in computing, from mainframes to minicomputers to personal computers and now to smartphones and other mobile devices, has assured a proliferation of new media and instant forms of communications worldwide. VR applications are enhanced new media with amplified communications potential. The now existing and growing field of virtual reality applications and their virtual objects, virtual works, and virtual worlds will generate major intellectual property litigation seeking to decide what may properly be viewed as part of the nature of the world and reality for all to share without limitation, as against the creative effort of virtual authors and virtual artists to set forth their vision of reality within a virtual world. In this respect, courts will struggle to maintain the proper balance between the rights of virtual authors and virtual artists to exclusive control of their virtual works and the rights of the public to share and build upon those virtual works in an environment that is designed to allow users to participate, interact, and create new forms of creative ­expression and new forms of user experiences in a less limited creative environment. Eventually copyright and other intellectual property law will recognize yet another dimension to protection and infringement for virtual reality applications, and that is the nature of the experience imparted to the user who participates in that world. Ultimately, the originality contained in virtual creativity may need to be measured by the overall experience imparted to the user rather than the actual audiovisual displays or other expression presented or available from the underlying authorship. Though courts apply copyright protection to virtual worlds and even to virtual objects, more rigorous infringement standards and more expansive interpretations of the idea/ expression and fair use doctrines will likely apply—particularly for virtual worlds and virtual objects that purposefully simulate the real world and real information. The coming challenge will be to distinguish those cases in which protection justifies licensing and those where the policies favor unrestricted use for further development of this new technology.

153   See Christianson v. West Pub. Co., 149 F. 2d 202, 203 (9th Cir. 1945); Key Maps v. Pruitt, 470 F.Supp. 33, 37 (SD Tex. 1978).

BARFIELD PRINT.indd 189

13/11/2018 11:52

190  Research handbook on the law of virtual and augmented reality

SOURCES Cases Apple Computer, Inc. v. Microsoft Corp., 709 F.Supp. 925 (N.D. Cal. 1989). Apple Computer, Inc. v. Microsoft Corp., 779 F.Supp. 133 (N.D. Cal. 1991). Apple Computer, Inc. v. Microsoft Corp., 799 F.Supp. 1006 (N.D. Cal. 1992). Apple Computer, Inc. v. Microsoft Corp., 35 F.3d 1435 (9th Cir. 1994). Atari Games Corp. v. Oman, 888 F. 2d 878 (D.C. Cir. 1989). Atari Games Corp. v. Oman, 979 F.2d 242 (D.C. Cir. 1992). Authors Guild, Inc. v. HathiTrust, 755 F.3d 87 (2nd Cir. 2014). Authors Guild, Inc. v. Google, Inc., 804 F.3d 202 (2nd Cir. 2015). AV Ex Rel. Vanderhye v. iParadigms, LLC, 562 F. 3d 630 (4th Cir. 2009). Baker v. Selden, 101 U.S. 99 (1880). Bateman v. Mnemonic, Inc. 79 F 3d. 1532 (11th Cir. 1996). L. Batlin & Son, Inc. v. Snyder, 536 F. 2d 486 (2nd Cir. 1976). Berkla v. Corel Corp., 66 F. Supp. 2d 1129 (1999). Bill Graham Archives v. Dorling Kindersley Ltd., 448 F.3d 605 (2nd Cir. 2006). Blanch v. Koons, 467 F.3d 244 (2nd Cir. 2006). Bridgeman Art Library, Ltd. v. Corel Corp., 36 F. Supp. 2d 191 (S.D.N.Y. 1999). Broderbund Software. Inc., v. Unison World, Inc., 648 F.Supp. 1127 (N.D. Cal. 1986). Brown Bag Software v. Symantec Corp., 12 U.S.P.Q.2d (BNA) 1191 (N.D. Cal. 1989). Brown Bag Software v. Symantec Corp., 960 F. 2d 1465 (9th Cir. 1992). Bucklew v. Hawkins, Ash, Baptie & Co., LLP., 329 F.3d 923 (7th Cir. 2003). Burrow Giles Lithographic Co. v. Sarony, 111 U.S. 53 (1884). Campbell v. Acuff-Rose Music, Inc., 510 U.S. 569 (1994). Cariou v. Prince, 714 F.3d 694 (2nd Cir. 2013). Christianson v. West Pub. Co., 149 F. 2d 202 (9th Cir. 1945). Computer Associates v. Altai, 982 F.2d 693 (2nd Cir. 1992). Data East USA. Inc. v. Epyx Inc., 862 F.2d 204 (9th Cir. 1988). Diamond Foods, Inc. v. Hottrix, Llc, No. Case No. 14-cv-03162-BLF (N.D. Cal. Jul. 18, 2016). Digital Communications Associates, Inc. v. Softklone Distributing Corp., 659 F.Supp. 449 (N.D. Ga. 1987). Encyclopedia Britannica Educational Corporation v. Crooks 542 F.Supp 1156, 1180 (W.D.N.Y. 1982). Engineering Dynamics, Inc. v. Structural Software, Inc., 26 F.3d 1335 (5th Cir. 1994). Feist Publications, Inc. v. Rural Telephone Service, 499 U.S. 340 (1991). Folsom v. Marsh, 9 Fed.Cas. 342, 343 (No. 4901) (C.C.D. Mass. 1841). Gates Rubber Co. v. Bando Chemical Industries, Ltd., 9 F.3d 823 (10th Cir. 1993). Gentieu v. John Muller & Co., Inc., 712 F.Supp. 740 (W.D. Mo. 1989). Gracen v. Bradford Exchange, 698 F.2d 300 (7th Cir. 1983). Harbor Software, Inc. v. Applied Systems, Inc., 925 F. Supp. 1042 (S.D.N.Y. 1996). Ilog, Inc. v. Bell Logic, LLC, 181 F. Supp. 2d 3 (D. Mass. 2002). Incredible Technologies, Inc. v. Virtual Technologies, Inc., 400 F.3d 1007 (7th Cir. 2005). Key Maps v. Pruitt, 470 F.Supp. 33 (SD Tex. 1978). Kienitz v. Sconnie Nation LLC, 766 F.3d 756 (7th Cir. 2014). Kim Seng Co. v. J & A Importers, Inc., 810 F. Supp. 2d 1046 (C.D. Cal. 2011). Kregos v. Associated Press, 937 F.2d 700 (2nd Cir. 1991). Lotus Development Corp. v. Paperback Software Intern., 740 F.Supp. 37 (D Mass. 1990). Lotus Development Corp. v. Borland International, Inc., 799 F.Supp. 203 (D. Mass. 1992). Lotus Development Corp. v. Borland International, Inc., 49 F.3d 807 (1st Cir. 1995), aff’d by an equally divided court, 516 U.S. 233 (1996). MAI Systems Corp. v. Peak Computer, Inc., 991 F.2d 511 (9th Cir. 1993). Manufacturer’s Technologies, Inc. v. CAMS, Inc., 706 F.Supp. 984 (D. Conn. 1989). Mazer v. Stein, 347 U.S. 201 (1954). MDY Industries, LLC v. Blizzard Entertainment, Inc., 629 F.3d 928 (9th Cir. 2010). Meshwerks, Inc. v. Toyota Motor Sales USA, Inc., 528 F.3d 1258 (10th Cir. 2008). Micro Star v. FormGen, Inc., 154 F.3d 1107 (9th Cir. 1998). Mitek Holdings, Inc. v. Arce Engineering Co., Inc., 864 F. Supp. 1568 (S.D. Fla. 1994). Mitel, Inc. v. IQTel, Inc., 124 F.3d 1366 (10th Cir. 1997). Monge v. Maya Magazines, Inc., 688 F.3d 1164 (9th Cir. 2012). NEC v. Intel, 10 U.S.P.Q.2d 1177 (N.D. Cal. 1989).

BARFIELD PRINT.indd 190

13/11/2018 11:52

Virtual copyright  191 Oracle America, Inc. v. Google Inc., 872 F. Supp. 2d 974 (N.D. Cal. 2012). Oracle America, Inc. v. Google Inc., 750 F.3d 1339 (Fed. Cir. 2014). Perfect 10, Inc. v. Amazon.com, Inc., 508 F.3d 1146 (9th Cir. 2007). Plains Cotton Co-op v. Goodpasture Computer Service, Inc., 807 F. 2d 1256 (5th Cir. 1987). Sega Enterprises Ltd. v. Accolade, Inc., 977 F. 2d 1510 (9th Cir. 1992). Sony Computer Entertainment, Inc. v. Connectix Corp., 203 F.3d 596 (9th Cir. 2000). Sony Corp. v. Universal Studios, Inc., 464 U.S. 417 (1984). Star Athletica, LLC v. Varsity Brands, 137 S. Ct. 1002 (2017). Stern Electronics v. Kaufman, 669 F.2d 852 (2nd Cir. 1982). Synercom Technology Inc. v. University Computing Co., 462 F.Supp. 1003 (N.D. Tex. 1978). Tetris Holding, LLC v. Xio Interactive, Inc., 863 F. Supp. 2d 394 (D.N.J. 2012). Thomas Wilson & Co. v. Irving J. Dorfman, 433 F.2d 409 (2nd Cir. 1970). Walt Disney Products v. Air Pirates, 581 F.2d 751 (9th Cir. 1978). Whelan Associates, Inc. v. Jaslow Dental Laboratory Inc., 797 F.2d 1222 (3d Cir. 1986). White v. West Pub. Corp., 29 F. Supp. 3d 396 (S.D.N.Y. 2014).

Statutes 17 U.S.C. §101. 17 U.S.C. § 102(a). 17 U.S.C. §102(b). 17 U.S.C. §103(a). 17 U.S.C. § 107.

Patents Nicholas Alferieff, “Man to Machine Communication Keyboard Device,” U.S. Patent No 3,428,747 (issued Feb. 18, 1969). Gary Grimes, “Digital Data Entry Glove Interface Device,” U.S. Patent No 4,414,537 (issued Nov. 8, 1983). Morton Heilig, “Sensorama Simulator” U.S. Patent No 3,050,870 (issued Aug. 28, 1962). Morton Heilig, “Stereoscopic Television Apparatus for Individual Use,” U.S. Patent No 2,955,156 (issued Oct. 4, 1960). Kenneth James, et al. “Optical Deformation Sensor,” U.S. Patent No 4,420,251 (issued Dec. 13, 1983). Charles Laenger, et al. “Method and Apparatus For Communicating with People,” U.S. Patent No 4,074,444 (issued February 21, 1978). Francis Rossire, “Stress Sensor and Control System for Limiting Stresses Applied to Aircraft,” U.S. Patent No 3,229,511 (issued Jan. 18, 1966). Hideo Suzuki, et al. “Musical Tone Apparatus Mounted on a Performer’s Body,” U.S. Patent No 4,905,560 (issued Mar. 6, 1990).

BARFIELD PRINT.indd 191

13/11/2018 11:52

6.  Trademark law and the right of publicity in augmented reality Brian D. Wassom, Amber M. Underhill, and Andrew L. Rossow

The emerging medium known as augmented reality (“AR”) may still be in its infancy, but it heralds an evolution in human creativity and communication that most of society has yet to fully grasp. The essential characteristic of AR is to liberate digital data and imagery from two-dimensional screens and allow that content to be perceived and interacted with as if it were just as volumetric and accessible as any physical object. Applying Marshall McLuhan’s axiom that “the medium is the message”—in other words, that the form of a medium embeds itself in the content transmitted through it, creating a symbiotic relationship by which the medium influences how the message is perceived—this means that AR will endow digital content with new, visceral familiarity, contributing to an erosion of the barriers between what is “real” and “not real,” and between what is “there” and “not there” in the world around us. Within such an environment, mixed messages and a certain degree of confusion are to be expected. The increasing adoption of AR for commercial messaging means that trademarkrelated disputes related to the medium cannot be far behind. AR has already begun to offer brand owners new and interesting ways to create, use, and expand upon their commercial identities. Those who would use those brand names and trademarks without permission, however, now also have the same abilities. As is the case any time a new mass medium is introduced, attorneys and jurists will be called upon to apply and adapt longstanding legal principles to uncharted factual territory. We begin this discussion with the caution that the speculation that follows is exactly that. By postulating certain hypotheticals and brainstorming the legal claims that certain parties may allege based on them, we do not mean to imply that any one of the legal theories discussed is more or less meritorious than others. As attorneys in active private practice, we are keenly aware that each case rises and falls on its own facts. Circumstances will certainly arise in the near future that bear similarities to those discussed here, but they will also contain distinguishing factors of their own.

1.  UNDERSTANDING TRADEMARKS The following concise background on trademark law will frame the discussion that follows on the types of trademark dispute that are most likely to arise in the augmented medium. It is not intended to be a comprehensive analysis, nor does it extend beyond the boundaries of United States law (although many principles of trademark protection have been more or less harmonized throughout most of the world). A trademark is anything—such as a word, phrase, symbol, device and/or design—that 192

BARFIELD PRINT.indd 192

13/11/2018 11:52

Trademark law and publicity in augmented reality  193 distinguishes the source of one product from the goods sold by others.1 Marks that identify the source of services, as opposed to goods, are most accurately called “service marks,” but this chapter will include them in its discussion of “trademarks” for simplicity’s sake. Most trademarks in contemporary society have taken the form of words (either as such or in a particular stylized form), logos, and/or slogans. The advancement and proliferation of communications media, however, has brought with it new ways to convey source and commercial identity. Distinct sounds, haptic sensations, and even smells can function as trademarks, so long as they can be shown to distinguish a source of goods or services in the relevant marketplace. Containers, décor, and even product shapes can be protected as “trade dress”—­essentially, the exterior trappings of a product that perform the same source identification function as a mark appended to the product. Functional features of a product are not eligible for trademark protection as such, however, since functionality is the realm of patent law. In order to be distinctive of a particular source of goods, it is not always necessary that consumers be able to precisely name the originator of the goods or the owner of the mark. It is enough that they recognize goods bearing that mark as originating from a particular source, even if the full identity of that source is unknown to them. Generally speaking, trademark rights arise in the first instance as a result of using the mark in commerce. These are called “common law” rights. Registering a mark with the United States Patent and Trademark Office under the Lanham Act provides important additional benefits,2 such as nationwide priority in the mark, constructive notice to others of the claimed rights, and additional remedies for infringement. Under the right circumstances, registration can be obtained on an “intent to use” basis for a certain period of time before actual use in commerce begins. Actual use in commerce, however, remains the foundation of rights in a mark. As long as the mark remains in sufficient use, these rights can continue to exist indefinitely. Conversely, if a mark ceases to be used for too long, its owner can eventually lose claim to the mark. Not every word, symbol, or other would-be mark is capable of functioning as a trademark. The conceptual threshold for protectability is not especially high, but the mark must be capable of actually distinguishing a specific product, as opposed to merely describing a type or genre of product. In making this determination, context is key. The word APPLE,3 for instance, is a generic term for the fruit of an apple tree, and could never serve as a trademark for a particular orchard’s produce. As applied to computers, however, it is an entirely arbitrary appellation, and thus can (and does) signal one ­company’s products as distinct from another’s. Courts and other regulatory bodies will apply a similar type of analysis in determining not only whether a mark can be protected at all, but also how “strong” of a mark it is. As applied to trademarks, the term “strength” can indicate both its conceptual distinctiveness and the actual degree of association consumers perceive between a mark and the product it signifies. A mark’s strength can be one factor that helps determine whether a symbol

1   US Patent and Trademark Office, Trademark Basics, www.uspto.gov/trademarks-getting-st​ arted/trademark-basics. 2   Lanham Act, 15 U.S.C. §1051 et seq. 3   Trademarks are customarily rendered in all caps.

BARFIELD PRINT.indd 193

13/11/2018 11:52

194  Research handbook on the law of virtual and augmented reality functions as a trademark, and the breadth of goods and services with which the mark owner may exclude the use of similar marks. When assessing a mark’s conceptual strength, it often helps to rank the mark in one of the five categories, each of which represents a descending degree of distinctiveness:4 Fanciful: A “fanciful” or “inherently distinctive” mark receives the highest degree of protection available under the Lanham Act. These marks have no logical or alternative meaning. Examples include DAQRI, an AR startup in Los Angeles, as well as KODAK, XEROX, KINKO’S, and KLEENEX. Arbitrary: An “arbitrary” mark is typically a word that is recognized in everyday life, but its meaning has no relation to the product or service it is attached to or identifying. These marks are slightly less strong than “fanciful” marks, but are still generally considered strong marks. Examples include DITTO, a company offering three-dimensional “virtual try-on” experiences for retail customers, BLACKBERRY (for mobile phones), and LOTUS (for software). Suggestive: A “suggestive” mark makes reference to, or calls to mind, the nature or characteristic of the goods or services to which it relates. These are generally weaker than “arbitrary” marks. Examples include TOTAL IMMERSION, the name of one of the earliest AR companies in France, BLU-RAY (for movie technology with a high capacity for data storage), CHICKEN OF THE SEA (for tuna), GREYHOUND (for buses), and COPPERTONE (for suntan lotion). Descriptive: These marks are considered weaker than suggestive marks, because they merely describe a characteristic of the product or service with no mental inference required. For this reason, they are often ineligible for trademark protection unless and until they can be shown to have acquired distinctiveness in the marketplace. The French company AUGMENT may well fall into this category; examples also include SUDSY SOAP, ALL BRAN, and VISION CENTER. Generic: These words can never be protected as trademarks and are free to be used by anyone, because they are basic, common descriptors for the category into which the product or service fits—such as “tape,” “shirts,” “computers,” or “smartphones and tablets.” Some marks that were once distinctive can become generic—and therefore ­unprotectable—by coming to be publicly used as a generic term. If an augmented reality company were to include “AR” as part of its proposed trademark (as a great many of them have), the Trademark Office may or may not agree to register the mark as a whole, but with the caveat that the applicant disclaim any exclusive rights in the “AR” portion, because that is generic for “augmented reality.” Examples of words that were once trademarks but became generic include lite beer, polo, aspirin, and cellophane. The strength of a mark is also an important factor when assessing whether the use of a mark infringes on one’s rights in another mark. 1.1  What it Means to “Infringe” a Trademark Any person who makes an unauthorized commercial use of a trademark in a way that is likely to cause confusion, mistake, or deception in the marketplace infringes the rights of that

4   See Abercrombie & Fitch Co. v. Hunting World, 537 F.2d 4 (2nd Cir. 1976); see also Two Pesos v. Taco Cabana, Inc., 505 U.S. 763, 768 (1992).

BARFIELD PRINT.indd 194

13/11/2018 11:52

Trademark law and publicity in augmented reality  195 trademark’s owner. The Lanham Act likewise imposes liability on anyone who uses a trademark or any false designation of origin, false or misleading description, or representation of fact that is likely to cause confusion, mistake, or deception as to the affiliation, connection, or association of such person with another person, or as to the origin, sponsorship, or approval of his or her goods, services, or commercial activities by another person.5 In all of these provisions, “likelihood of confusion” is the touchstone of liability. Infringement inhibits the trademark owner’s ability to associate their goods or services with their mark because the mark is so similar to another mark that consumers are confused as to which mark signifies which source. Courts assess whether trademark infringement has occurred by measuring the “likelihood of confusion” presented by the facts of a particular case. The tests used to determine likelihood of confusion differ among the various circuit courts, but they all emphasize varying portions of the following factors: ● ● ● ●

● ● ●

● ● ● ●

● ●

The similarity or dissimilarity of the marks in their entireties as to appearance, sound, connotation, and commercial impression. The similarity or dissimilarity and nature of the goods . . . described in an application or registration or in connection with which a prior mark is in use. The similarity or dissimilarity of established, likely to continue trade channels. The conditions under which and buyers to whom sales are made, that is, “impulse” vs. careful, sophisticated purchasing. The fame of the prior mark. The number and nature of similar marks in use on similar goods. The nature and extent of any actual confusion. The length of time during and the conditions under which there has been concurrent use without evidence of actual confusion. The variety of goods on which a mark is or is not used. The market interface between the applicant and the owner of a prior mark. The extent to which applicant has a right to exclude others from use of its mark on its goods. The extent of potential confusion. Any other established fact probative of the effect of use.6

Not all of these factors may be relevant or of equal weight, and any one of the factors may control a particular case. 1.2  Trademark Dilution Trademark dilution occurs when a third party uses a famous mark and weakens the value or quality of the mark. Dilution protects the distinctiveness, reputation, and connotation of a “famous mark.” A mark is “famous” if it is widely recognized by the general public of the United States as a designated source or origin of an owner’s goods or services.7

  See 15 U.S.C. §§1114, 1125.   See, e.g., In re E.I. du Pont de Nemours & Co., 476 F.2d 1357, (C.C.P.A. 1973). 7   15 U.S.C. §1125(c)(2)(A). 5 6

BARFIELD PRINT.indd 195

13/11/2018 11:52

196  Research handbook on the law of virtual and augmented reality The more distinctive or well known the mark is, the more value it holds and the more protection it is given. It is important to remember that only famous marks are entitled to protection against dilution. The Lanham Act provides four factors to determine whether a mark is considered famous:8 ●

the duration, extent, and geographic reach of advertising and publicity of the mark, whether advertised or publicized by the owner or third parties; ● the amount, volume, and geographic extent of sales of goods or services offered under the mark; ● the extent of actual recognition of the mark; and/or ● whether the mark was registered. There are two ways in which a trademark can be diluted—dilution by blurring and dilution by tarnishment. “Blurring” is an association arising from the similarity between a mark and a famous mark that negatively affects or impairs the commercial value of the famous mark.9 The Lanham Act also sets out a factor test for determining whether a mark is likely to cause dilution by blurring:10 ● ●

● ● ●

●

The degree of similarity between the mark or trade name and the famous mark; The degree of inherent or acquired distinctiveness of the famous mark; The extent to which the owner of the famous mark is engaging in substantially exclusive use of the mark; The degree of recognition of the famous mark; Whether the user of the mark or trade name intended to create an association with the famous mark; and/or Any actual association between the mark or trade name and the famous mark.

The Act also defines “dilution by tarnishment”11 as an association arising from the similarity between a mark and a famous mark that harms the reputation of the famous mark. In other words, this type of use threatens the very nature or meaning of the mark and could possibly alter the reputable nature of the mark’s associations.

2.  TRADEMARKS IN THE AUGMENTED MEDIUM 2.1  Depicting Trademarks through Visual AR Displays As more digital content finds its way into augmented media and virtual reality applications, trademark-laden commercial content will surely follow and find its way into these

    10   11    8  9

Id. 15 U.S.C. §1125(c)(2)(B). Id. Id.

BARFIELD PRINT.indd 196

13/11/2018 11:52

Trademark law and publicity in augmented reality  197 platforms. We have already seen this to some degree. There are scores of one-off AR applications available for download in the popular online app stores, many of which display images of content that either is or contains a trademark. Perhaps the most extreme (and disturbingly plausible) depiction of brand name saturation in augmented reality can be found in the work of filmmaker Keiichi Matsuda. His short videos depict life in a society where AR is “mature” and commonplace, entirely from the first-person perspective of an average citizen wearing digital eyewear. In his short video Augmented (hyper)Reality: Domestic Robocop,12 the AR user sees literally every flat surface in his modest kitchenette digitally plastered with branded advertisements. At one point he even manually raises the “advertising level” of his eyewear, suggesting that he is receiving micropayments or subsidized services for each ad in his field of view. Matsuda carried this concept further in a second video published in May 2016, simply titled Hyper-Reality.13 This short film, also shot entirely from a first-person view, shows us the perspective of a woman riding a public bus and then shopping in a grocery store. Literally everything she sees—every surface within the bus and the store, the sidewalks she travels, the storefronts she passes, the products she handles, even the people she passes—is awash in audiovisual augmented advertising. The effect is so complete that at one point in the film, when the woman’s device fails and her perspective reverts to only what her physical eyes and ears can perceive, the viewer is left feeling unnerved, even anxious. With consumer-facing communication come inevitable questions of how commercial goodwill is being used to attract consumer attention. That is the realm of trademark law. Because AR will enable forms of communication that have not previously been seen, many of the related trademark questions will also be novel. 2.2  Commercial Speech Communicated through Nonvisual Senses Consumers associate symbols with different forms of commercial products and services on a daily basis, but they do not necessarily think of them as “trademarks” per se. However, under the Lanham Act, “symbols” are registrable and can include exotic marks such as scents, sounds, and colors.14 A few famous examples of these symbols include the MGM lion roar at the beginning of MGM films, the sound a Harley-Davidson makes when it starts, and the Intel tones at the end of its commercial. Emerging AR technologies have already inspired a wide variety of conventional trademarks, including words, logos, and phrases. Soon, though, technologies that augment the sense of touch may lead to a rush of trademark applications seeking to protect a wide variety of artificial textures. A number of companies are experimenting with different means of tricking the mind into thinking one’s skin is perceiving whatever haptic sensation a content provider wishes to convey. For example, various competitors—from small startups such as Ultrahaptics to giants such as Disney—have introduced various methods 12   Keiichi Matsuda, Augmented (hyper)Reality: Domestic Robocop, Youtube (Jan. 6, 2010) www.youtube.com/watch?v=fSfKlCmYcLc. 13  www.youtube.com/watch?v=YJg02ivYzSs&feature=autoshare 14   Two Pesos v. Taco Cabana, 505 U.S. 763, U.S (1992), where the Court held that a color by itself is not protectable but a color combination can be afforded trademark protection, this only if there is proof of secondary meaning.

BARFIELD PRINT.indd 197

13/11/2018 11:52

198  Research handbook on the law of virtual and augmented reality of creating such sensations using precisely controlled puffs of air. The Finnish company Senseg has proposed synthetic surfaces made up of what it calls “tixels” that are able to mimic various textures by means of electrostatic fields. A number of other companies are creating gloves and other wearable devices that accomplish a similar effect through direct stimulation of the skin. What all of these companies recognize is that the potential of AR will never be fully realized until users can reach out and touch virtual objects through haptic interfaces. One way in which this technology seems likely to (literally) get into the hands of consumers is through retailers using haptic technology to further enhance the “feel” of their products. When that begins to happen, there will be a resurgence of interest in haptic trademarks. Of the less conventional trademarks, haptic marks are among the least common.15 Those commentators who have broached the subject in recent years have only identified a handful of such federally registered marks.16 They include a registration by American Wholesale Wine & Spirits for “a velvet textured covering on the surface of a bottle of wine”17—specifically, its Khvanchkara brand of wine. In the course of convincing the USPTO to register this mark, American Wholesale distinguished its “velvety covering” from that of the more iconic Crown Royal bag by noting that Khvanchkara is “tightly encased within the fabric,” and that the “FEEL of a LIMP bag is quite different from the FEEL of a TURGID velvety surface attached to a wine bottle.”18 Similarly, Touchdown Marketing has registered a trademark in the “pebble-grain texture” and “soft-touch feel” of its basketball-shaped cologne dispenser, and Fresh, Inc. has registered the “cotton-textured paper” that wraps its soap products. Conceptually, a distinctive touch ought to be just as protectable by trademark law as any other unique indicator of source. Indeed, in 2006, the International Trademark Association (INTA) adopted “a resolution supporting the recognition and registration of ‘touch’ marks.”19 In practice, however, it is very difficult to separate the way something feels from the function that texture performs—and to come up with a texture that is truly “distinctive” of one product as opposed to other brands within the same category of products. That is where haptic AR technologies come in. The ability to coat the surface of any product with a synthetic layer capable of mimicking any arbitrary texture the ­manufacturer chooses would finally break the connection between a product’s feel and the

15   Other trademark practitioners have called these “tactile,” “texture,” or even “touch” marks, but the term “haptic” is more definitionally sound and technologically consistent. 16   See Steve Baird, Touch Trademarks and Tactile Brands with Mojo: Feeling the Strength of a Velvet, Turgid, Touch Mark? Duets Blog (July 13, 2009) www.duetsblog.com/2009/07/articles/­ trademarks/touch-trademarks-and-tactile-brands-with-mojo-feeling-the-strength-of-a-velvet-turgidtou​ch-mark/. 17   U.S. Trademark Application Serial No 76,634,174 (Filed Mar. 23, 2005) available at http:// tsdr.uspto.gov/#caseNumber=76634174&caseType=SERIAL_NO&searchType​=­status​Sea​ rch (trademark canceled as of May 24, 2013). 18   Response to Office Action, U.S. Trademark Application Serial No 76,634,174 (Apr. 17, 2006) available at http://tsdr.uspto.gov/documentviewer?caseId=sn76634174&docId=ROA200604 18121514#docIndex=4&page=1. 19   Report of the World Intellectual Property Organization, Standing Committee On The Law Of Trademarks, Industrial Designs And Geographical Indications, Sixteenth Session, Geneva, November 13 to 17, 2006 at 10–11 (2006).

BARFIELD PRINT.indd 198

13/11/2018 11:52

Trademark law and publicity in augmented reality  199 function it performs. Consider, for example, a book cover that feels wet, or a plastic squirt gun that feels metallic. There is no necessary correlation between what these products are, or what they do, and the way they feel. Therefore, under trademark law as it is understood in the United States and most other jurisdictions, there should be no conceptual barrier to those manufacturers seeking trademark protection in those textures. Of course, not every artificial texture will automatically be eligible for trademark protection. Many haptic enhancements may still be chosen for functional reasons. The maker of an automotive steering wheel or a baseball, for example, might choose to make their products artificially sticky to enhance performance. A cell phone might be designed to get warmer in one’s pocket as it rings, in order to catch the user’s attention.20 And it could be that certain haptic enhancements still do not rise to the level of being sufficiently distinctive of a particular source to serve as a trademark. Still, by promising the ability to manipulate the sensation of touch independently from other aspects of a product, haptic AR technologies open up a new and exciting world of trademark possibilities. Consumers may soon reach out and touch . . . whatever retailers want them to.

3. POTENTIAL FLASHPOINTS FOR AUGMENTED TRADEMARK DISPUTES As with the challenges involved in trademark acquisition and development, many of the trademark disputes that AR companies encounter will be unremarkable compared to publishers in any other media. There will be questions over marks that are allegedly too descriptive, allegations that two marks are similar or dissimilar to each other, and surveys conducted to determine how actual consumers perceive certain marks. Nevertheless, the unique aspects of the medium do suggest certain unique ways in which trademark disputes may arise. 3.1  Location-Based Augmented Advertising The growth of the commercial internet over the past 25 years has been funded predominately by advertising revenue. Consumers get to browse free content on millions of web pages and on various search engines in large part because advertisers have paid good money to insert their ad next to whatever they are reading. This funding model is likely to continue well into the future. The primary purpose of all commercial advertising is to draw potential customers to the advertised business or product, and away from its competitors. Moreover, comparative advertisements—those that compare a product to its competition—have been around for decades. Courts have had opportunities to draw some basic lines between what is permissible to say in such advertisements and what is “deceptive” advertising. Generally speaking, it is permissible to describe your competitor’s goods and compare one product to

20   Technically, as noted elsewhere, the ability to discern heat is distinct from the sense of touch. For simplicity’s sake, however, this book will follow the popular approach of treating them as the same.

BARFIELD PRINT.indd 199

13/11/2018 11:52

200  Research handbook on the law of virtual and augmented reality another, but you cannot say things that are likely to confuse customers into believing that you are your competitor. You cannot say something materially false or misleading about your competitor or your own product. And you cannot do anything to confuse reasonable consumers into mistakenly believing there is some sort of connection, ­sponsorship, affiliation, or endorsement between your companies or products. These boundaries are not always easy to apply, however, and there are several contexts in which the courts have not been able to agree on how they apply. For example, the battle over “keyword advertising”—using an algorithm to display a “sponsored” ad whenever a user types a given term into a search engine—is still being fought, more than a decade after the practice began. Google explained its own keyword advertising system, called “AdWords,” this way: Google AdWords is Google’s advertising program. AdWords lets you create simple, effective ads and display them to people already searching online for information related to your business. So how is it possible to show your ads only to the most relevant audiences? The answer is keywordbased advertising.

When a searcher visits Google and enters a query—say, good beginner guitars—Google displays a variety of relevant search results, such as links to articles containing guitar purchasing advice or websites dedicated to novice musicians. Google also displays AdWords ads that link to online businesses selling guitars, music lessons, or other products and services related to the query. For example, imagine that you own a music store carrying a large selection of guitars. You could sign up for an AdWords account and create ads for entry-level guitars in your inventory. For each of your ads, you might select keywords (single words or phrases related to your ad’s message) such as beginner guitars or entry-level guitars.21 Company A potentially implicates trademark law when it purchases a search term that is also a trademark belonging to Company B. The fact that Company A’s advertising appears when a user searches for Company B’s trademark raises questions of whether Company A is “using” that trademark “in commerce” (most courts have said yes), and whether this use creates a likelihood that consumers will be confused regarding the potential association or sponsorship between the two companies or as to the source of Company B’s goods or services (which depends on how the two marks are displayed, and in what context).22 Answers to this latter question have been mixed. Some courts over the past decade have found that ads triggered by a trademarked keyword search cause a likelihood of ­confusion—especially when the resulting ad also incorporates the trademarked term,23 21   See Google AdWords, www.google.com/adwords/learningcenter/text/18911.html (last ­visited March 23, 2009). 22   See, e.g., Maral Kilejian and Sally Dahlstrom, Trademark Infringement Claims in Keyword Advertising, Franchise Law Journal Vol. 36, No. 1 (Summer 2016), available at www.americanbar. org/content/dam/aba/publications/franchise_lawyer/summer2016/flj-v36-1-kilejian.authcheckdam. pdf (collecting cases). 23   See, e.g., Storus Corp. v. Aroa Mktg., Civ. No. 06-2454-MMC; 2008 U.S. Dist. LEXIS 11698, at *12–13 (N.D. Cal. Feb. 15, 2008) (finding infringement where defendant’s sponsored ad was triggered by and incorporated plaintiff’s trademarked “smart money clip”).

BARFIELD PRINT.indd 200

13/11/2018 11:52

Trademark law and publicity in augmented reality  201 but even occasionally when it does not.24 On the other hand, several recent cases have rejected the proposition that merely purchasing a competitor’s trademark as a search term in and of itself creates confusion.25 This may suggest that the potential for confusion has decreased in many situations as online sponsored ads have become more commonplace. Where the potential for confusion exists, though, the question of who is responsible for it also remains open. Rosetta Stone is one of several companies to sue a search engine for allowing competitors to use its marks in keyword ads. As most other courts had done in similar cases, the trial court dismissed the suit as a matter of law, finding that Rosetta Stone could not prove that the search engine was liable. But in 2012, the US Court of Appeals for the Fourth Circuit overturned that holding, finding it possible that the search engine’s policy on the use of keywords in sponsored ads could amount to direct infringement, contributory infringement, or trademark dilution.26 Other cases have likewise gone either way on liability depending on how the particular trademark at issue appeared in the header or text of a sponsored ad. Augmented reality will take this jostling for position between advertisers to a new level. This is already happening in TV broadcasts of certain sports games, in which “digital billboard replacement” technology is used to superimpose digital ads on top of the ones that are physically present in the stadium. The Public Ad Project and the Heavy Projects have demonstrated similar concepts on mobile devices by sponsoring campaigns that replace physical billboards with artistic images when viewed through a mobile device. But what happens when AR eyewear becomes ubiquitous, and digital ad replacement becomes commonplace? Will advertisers pay AR service providers for the ability to superimpose their ads on top of what consumers see? If the past 25 years of ecommerce is any indication, then the answer is “yes”—and in a number of creative ways. So, for example, a business may pay to superimpose its logo on top of signs advertising a competitor’s products, completely blocking the physical ad from view. Or, the mere act of looking at Company A’s ad through your AR eyewear may trigger a virtual ad for Company B to pop up somewhere else in your field of vision. An example of this is looking at a McDonald’s sign through a digital device and instantly seeing a Burger King ­advertisement ­superimposed upon it. Similarly, your decision to look at something may prompt suggestions for goods and services relating to the thing you’re looking at. Self-described “pop culture hacker” Jonathan McIntosh captures all of these ideas in his parody video “ADmented Reality.”27 The video depicts a world in which every glance triggers another advertisement in one’s 24   See, e.g., Edina Realty, Inc. v. Themlsonline.com, Civ. 04-4371JRTFLN; 2006 U.S. Dist. LEXIS 13775 (D. Minn. March 20, 2006) (finding liability where “Defendant purchases search terms that include the Edina Realty mark to generate its sponsored link advertisement”); Fin. Express LLC v. Nowcom Corp., 564 F. Supp. 2d 1160, 1177 (C.D. Cal. 2008) (holding that defendant’s purchase of keywords that “are identical or strikingly similar to the trademarks held by plaintiff ” along with its offer of “services and products which are highly related to those offered by plaintiff ” and “simultaneous use of the Web as a marketing channel” may result in consumer confusion). 25   See, e.g., 1-800-Contacts, Inc. v. Lens.com, Inc., 722 F.3d 1229 (10th Cir. 2013). 26   See Rosetta Stone LTD. v. Google, Inc., 676 F.3d 144 (4th Cir. 2012). 27   Jonathan McIntosh, Admented Reality, Youtube (Apr. 5, 2012) www.youtube.com/watch?v​ =_mRF0rBXIeg&feature=kp.

BARFIELD PRINT.indd 201

13/11/2018 11:52

202  Research handbook on the law of virtual and augmented reality digital eyewear, to the point that reality itself becomes obscured in a sea of sponsored content. Other commentators have also foreseen augmented advertising and the legal issues it will raise. John C. Havens discussed them and some of the legal issues they raise in his insightful piece for Mashable called “Who Owns the Advertising Space in an Augmented Reality?”28 Noting that Google had already applied for a patent for digitally replacing physical ads within the Street View feature of Google Maps, Havens wrote that “the importance of virtual real estate may quickly supplant actual signage for advertisers. This is especially true when virtual signage could be switched dynamically for individual eye traffic depending on a viewer’s preferences.”29 He went on to quote Gabe Greenberg, director of social and emerging media at Microsoft, as saying that “if the experience presents the ads in a way that makes sense for the augmented reality experience and the user’s intention, this could be a powerful advertising tool for tomorrow’s marketplace.”30 These predictions are persuasive. Owners of all types of real estate—not only billboards or other physical signage—will face similar issues as AR becomes commonplace. Indeed, the lead author of this chapter has already had occasion to counsel multiple forward-thinking businesses on the question of what they can do to prevent unauthorized augmented content from being associated with their commercial real estate. In many cases, the answer will be “nothing.” Unless it is projected in a manner that makes it actually present in a given physical location, AR imagery will still reside on a two-dimensional (albeit transparent) screen. It may look like it exists in three dimensions or that it is plastered on the side of an actual building, but that is merely an illusion. The digital content is not really there. In that sense, it is much more akin to a digital map, painting, or movie than it is a billboard. For that reason, much AR content will not conflict with the rights afforded by property law, and the content itself will be protected by the First Amendment to the US Constitution and similar free speech guarantees (as described in more detail below). Legal efforts to prohibit AR content as such, therefore, will encounter stiff resistance from the courts. That said, if businesses are able to demonstrate that particular content creates a likelihood of confusion among potential consumers as to the source of goods and services, or as to the existence of a sponsorship, endorsement, or other relationship between two entities, the fact that the content is displayed in augmented form will not immunize its publisher from liability under the Lanham Act. Whenever AR is used to associate digital content owned by one party with physical content owned by another—especially if one or both sets of content is branded—the potential exists, in theory, for consumer confusion to result. That is, it is conceivable that a user will be misled into believing that the digital content is somehow sponsored by, or affiliated with, the physical content over which it is superimposed. In such circumstances, the augmented advertising may infringe trademark rights. Of course, it is also quite possible that the context of the experience (such as the user’s familiarity with how the

28   John Havens, Who Owns the Advertising Space in an Augmented Reality World? Mashable (June 6, 2011) http://mashable.com/2011/06/06/virtual-air-rights-augmented-reality/. 29   Id. 30   Id.

BARFIELD PRINT.indd 202

13/11/2018 11:52

Trademark law and publicity in augmented reality  203 system operates, or the content of the digital content itself) will clearly communicate otherwise. The degree of this risk cannot be assessed categorically ahead of time; rather, as in any other context, it is dependent on the factual circumstances involved. This basis for potential confusion, however, ought to be kept in mind when designing commercial AR experiences. The potential for confusion will come within digital services in which consumers expect to see advertising content from a variety of authentic sources within a viewpoint-neutral environment. One does not approach billboards, telephone directories, television commercial breaks, or internet banner ads as such with a predetermined expectation of the message those media will contain. Instead, one bases their determination about the source of a particular advertisement within those media on the content and context of the ad itself. For example, merely opening an internet browser tells consumers almost nothing about what sort of banner advertising they might encounter; they know by virtue of having surfed the internet that they will be served such ads by any random company that may have paid to place them there. But if they are discerning, consumers will notice that certain types of websites are more likely to serve up advertisements from a particular point of view, and that the behavioral advertising cookies in their browser will sometimes deliver ads based on their prior online activity. Similarly, to the extent that anyone still reads telephone directories, they ought to expect to see advertisements for local businesses rather than for those located elsewhere. When there are multiuser, viewpoint-neutral augmented reality browsers is when allegations of trademark infringement are expected to arise in earnest. The existing ability to digitally replace physical signs within mapping programs such as Bing and Google Maps offers a glimpse of what such a world will be like. Ubiquitous, always-on AR will feel very much like moving around within a three-dimensional version of those contemporary mapping programs. Once consumers find themselves there, what expectations will they have about the advertising they see? More than likely, they will realize that at least some of the augmented content they encounter is provided by the service provider itself (whichever company that turns out to be), while some is triggered by their own personal activities and preferences. Just as with behavioral advertisements on the internet today, no two users of the service are likely to encounter all of the same ads. Unlike the current web, however, augmented ads will necessarily correspond to physical places. It will be those relationships between digital and physical content that will raise new and unique questions of when a likelihood of confusion may exist. Sticking with the fast food example, then, will it be permissible in this context for Burger King to deliver users an ad every time they look in the direction of a McDonald’s restaurant or sign? If so, will the law of trademarks and unfair competition place limits on how obtrusive these ads can be? In other words, may they appear only in the periphery of a user’s vision? May they hover in space next to the Golden Arches, or even be superimposed over them? Moreover, the degree to which a service provider makes these decisions—or allows users to adjust such settings—could well determine whether the service provider may be held jointly liable for any resulting infringement. Courts deciding AR advertising cases in these contexts will apply the lessons learned in preexisting media, including the reasoning of the search engine keyword cases with which today’s courts are wrestling. Just as search engine algorithms use particular terms as

BARFIELD PRINT.indd 203

13/11/2018 11:52

204  Research handbook on the law of virtual and augmented reality keywords that prompt an ad to appear, so too can the physical objects that prompt similar virtual ads in AR devices be thought of as “keywords.” Whether it is a billboard, logo, or some other trigger, any object that prompts an algorithm to display an ad is performing the same function that keywords do today. A determination of whether that ad creates a likelihood of confusion will depend on how the likelihood of confusion factors applies to the particular case at hand. As with existing case law on sponsored advertising, moreover, courts are likely to be all over the map in how they decide such cases at first, until the model becomes more commonplace and a consensus forms about what boundaries it is fair to expect advertisers to observe in this space. As late as the middle of 2017, the possibility of genuine commercial confusion arising from AR experiences still seemed relatively remote. Prior to this point, having an AR experience required a user to download and open a particular, branded app on their device. These apps also usually offer only a very limited range of options in a predetermined number of situations. So, for example, the only way a user would see a Burger King ad atop the Golden Arches would be by using an app (or user-generated layer with an app that hosts such content) designed specifically for that purpose. In this situation, a trademark owner could object to the way that its trademark is being “used in commerce,” and the way in which the app is portrayed could conceivably be confusing. Assuming that the user understands where the app is coming from, however, one can hardly expect the user to be surprised or confused by what they see through it. In the summer of 2017, however, Apple and Android began to roll out AR-creation software designed specifically for their own proprietary operating systems—called ARKit and ARCore, respectively. These developments heralded an order of magnitude reduction in the barriers to creating customized AR content. Even as of this writing—which comes before either program has been made completely available to the general public—examples of these new tools being used for commercial advertising purposes are widespread. Their direct utility specifically for creating location-based content is still unclear, but they nevertheless herald an increased democratization in the creation of AR content. Other content creation platforms are either already on the market or continue to come online that are more specifically geared at location-based content, including Wikitude, Metaverse, and Motive.io. The Pokémon Go phenomenon that began in July 2016 planted the seeds of a similar issue, albeit in a slightly different form factor. By popularizing the concept of locationbased mobile gaming, Pokémon Go demonstrated to brick-and-mortar retailers a new method of driving foot traffic to their locations. Therefore, it was not long after the game took off that chains such as Starbucks and Sprint inked deals with Niantic, the game’s publisher, to locate Pokéstops and Gyms (the in-game terminology for geolocated, interactive digital content) at their physical stores. Inevitably, this will draw players to those locations. Players in the vicinity of these stores see the content displayed within the game itself, and the longer the sponsorship deal remains in place, the more players will seek out those specific stores with the expectation of finding AR content there. Those deals were well publicized and above board. With that precedent now established, however, will players of the next popular location-based game be acculturated to assume that geolocated content associated with the same or other retail locations is likewise sponsored by those retailers? If so, the potential could exist for confusion as to whether commercial digital content is sponsored by or affiliated with the retail establishment.

BARFIELD PRINT.indd 204

13/11/2018 11:52

Trademark law and publicity in augmented reality  205 3.2  Incorporating Third Party Trademarks into Augmented Content Trademarks frequently show up inside artistic works—especially in video games that attempt to create a realistic world in which players can immerse themselves. For the most part, courts uphold these uses as free speech, due in no small part to Brown v. Entertainment Merchants Association,31 the United States Supreme Court’s decision in 2011 that video games deserve First Amendment protection. Games and other immersive augmented reality environments will also attempt to create realistic digital worlds. In so doing, there will inevitably be some AR applications that recreate actual trademarks in the name of authenticity. The one fundamental difference between the AR medium and traditional digital expression, however, is that AR content is inherently tied to real physical locations. This distinction adds a layer of risk to replicating someone else’s trademark in AR, because associating that trademark with a real place or object could, in many foreseeable circumstances, heighten the likelihood that someone will draw a connection between the trademark and the physical place or object with which it is digitally associated. For example, players may see the mark digitally displayed on the wall of a business not associated with the trademark owner, or the mark may appear (wither physically or digitally) on a real object designed to serve as a target within the AR app. In either circumstance, the mark is no longer confined within a virtual, fictional world created by the artist, but instead is being associated with real objects or places that may be businesses or products with which the trademark owner does not wish to be associated. This could, in some cases, satisfy enough of the likelihood of confusion factors to create a real headache for both the trademark owner and the designer of the AR environment. Of course, it is equally possible—again, depending on the circumstances of the particular case—that the choice to make that particular association between trademark and physical place or thing could, in and of itself, be a creatively expressive decision that merits First Amendment protection. Regardless of the result, however, use of trademarks within AR content will inherently raise an additional dimension of legal complexity beyond that found in other digital works. 3.3  Fair Use, Free Speech, and the Unauthorized Augmentation of Trademarks Trademark ownership is not a complete monopoly on any and all uses of the word or symbol that forms the trademark. Although trademark rights are broad, they exist only to protect consumers from confusion and to safeguard business goodwill. As restrictions on the rights of others’ speech, moreover, trademark laws always exist in an uneasy tension with the First Amendment to the United States Constitution. “Because overextension of Lanham Act restrictions in the area of [artistic expression] might intrude on First Amendment values,” wrote the Second Circuit Court of Appeals in the frequently quoted opinion Rogers v. Grimaldi, “[courts] construe the Act narrowly to avoid such a conflict.”32 That case stands for the proposition that artists can freely refer

  131 S. Ct. 2729 (2011).   Rogers v. Grimaldi, 875 F. 2d 994, 998 (2d Cir. 1989).

31 32

BARFIELD PRINT.indd 205

13/11/2018 11:52

206  Research handbook on the law of virtual and augmented reality to trademarked goods and services by name in the titles of their songs, films, and other creative expressions. Such issues will inevitably arise in the context of augmented works just as they do elsewhere. First Amendment rights are fully available in augmented and virtual media. Judge J.P. Stadtmueller of the United States District Court for the Eastern District of Wisconsin held as much on July 20, 2017, in the case Candy Lab, Inc. v. County of Milwaukee (filed by this chapter’s lead author),33 when he issued a preliminary injunction striking down a municipal ordinance regulating the publication of location-based augmented reality content. The basis for the ruling—which relied heavily on Brown—was that developers’ AR content is free speech, and that the ordinance infringed on developers’ First Amendment rights. Quoting Brown, Judge Stadtmueller wrote: “whatever the challenges of applying the Constitution to ever-advancing technology, ‘the basic principles of freedom of speech and the press, like the First Amendment’s command, do not vary’ when a new and ­different medium for communication appears.”34 When the subject matter of user-generated AR content relates to a particular brand, no object will be more of a temptation in terms of serving as the trigger for that content than the very trademark that the brand owner uses to represent its goodwill to the public. Indeed, this has already happened at least once. In 2010, Mark Skwarek—a professor at NYU Polytechnic School of Engineering, President of SemblanceAR, and an expert witness in the Candy Lab case—released the iPhone app “The Leak in Your Home Town.” Through this app, one could view a physical sign bearing the BP logo at a local gas station, and see superimposed on that logo a digital broken pipe spewing oil, exactly like the one responsible for the then current spill in the Gulf of Mexico. Existing media also teach that a sizable portion of that commentary will be directed back toward the brands who advertise to consumers. For almost as long as companies have been setting up shop at , there have been detractors posting vitriol at . In today’s social media, popular sites such as Ripoff Report and Pissed Consumer base their entire business models on naming and shaming ­commercial brands. Although some early judicial decisions blocked these sites’ ability to reproduce the trademarks of the companies they criticize, most courts and other trademark dispute resolution organizations recognize such content as fair commentary that trademark holders cannot prevent.35 For example, in 2011, the United States District Court for the   No 17-CV-569-JPS.   Id (quoting Brown, 564 US at 789). 35   See, e.g., Taubman Co. v. Webfeats, 319 F.3d 770, 777–78 (6th Cir. 2003) (no Lanham Act violation where gripe site with domain name taubmansucks.com that provided editorial on conflict between website creator and plaintiff corporation did not create any possibility of confusion); Taylor Bldg. Corp. of Am. v. Benfield, 507 F.Supp.2d 832, 847 (S.D. Ohio 2007) (gripe site with domain name taylorhomesripoff.com that served as forum for criticizing home builder did not create any likelihood of confusion “because [n]o one seeking Taylor’s website would think—even momentarily—that Taylor in fact sponsored a website that included the word “ripoff ” in its website address”); Bally Total Fitness Holding Corp. v. Faber, 29 F.Supp.2d 1161, 1163–4 (C.D.Cal.1998) (gripe site with domain name www.compupix.com/ballysucks dedicated to complaints about Bally’s health club did not create likelihood of confusion because no reasonable visitor to gripe site would assume it to come from same source or think it to be affiliated with, connected with, or 33 34

BARFIELD PRINT.indd 206

13/11/2018 11:52

Trademark law and publicity in augmented reality  207 Eastern District of New York rejected a trademark infringement lawsuit that challenged the use of a reviewed company’s trademarks in the sub-URLs, metadata, and text of PissedConsumer.com.36 Despite copious use of the plaintiff’s marks throughout the website, the court found it implausible that any reasonable person would believe the site’s critical commentary to be sponsored by or associated with the trademark owner. For these reasons, it seems likely that courts will protect a healthy amount of breathing room for augmented commentary on brands that is superimposed upon their printed logos— although the exact parameters of what is fair and what is confusing remain to be drawn.

4.  THE BASICS OF PUBLICITY RIGHTS Just as trademarked objects can easily serve as triggers for digital content, so too can the physical characteristics of individual people. Therefore, the right of publicity—that weird, state-level transitional species between the common law of privacy and intellectual property—will play an increasingly prominent role in this debate going forward. Most of the 50 states recognize the statutory right of publicity to some degree. This right emerged from the common law of privacy to become more or less recognized as a form of intellectual property. Although the particulars vary slightly from state to state, it is essentially the right of an individual to control the commercial exploitation of his or her identity. Again, bear in mind that the following discussion is by necessity generic and not specific to any particular jurisdiction or fact pattern. Indeed, with so much variation in this area of law between states (not to mention ambiguity even within a particular state), any concise explanation of the law can be no more than a starting point. For present purposes, however, this high-level overview will be sufficient to frame the AR-specific discussion that follows. Perhaps the most widely cited summary of the right of publicity as generally understood across the United States (although not precisely the same way in every state) comes from the Restatement (Third) of Unfair Competition § 46: “[o]ne who appropriates the commercial value of a person’s identity by using without consent the person’s name, image, likeness, or other indicia of identity for purposes of trade is subject to liability.”

sponsored by Bally’s); MCW, Inc. v. Badbusinessbureau.com, L.L.C., No. 02 Civ. 2727, 2004 WL 833595, at *16 (N.D.Tex. Apr. 14, 2004) (Lanham Act unfair competition claims against consumer review websites called “ripoffreport.com” and “badbusinessbureau.com” that used plaintiff’s trademarks in connection with allegedly defamatory posts dismissed because no visitor to websites would believe that plaintiff markholder endorsed the comments on sites); Whitney Inf. Network, Inc. v. Xcentric Ventures, No 2:04-cv-47-FtM-34SPC, 2005 WL 1677256 (M.D.Fla. July 14, 2005) (unpublished memorandum and order) (dismissing trademark infringement and false designation of origin claims against “ripoffreport.com” because plaintiff mark holder, a seller of education courses, was involved in a different field than defendant, who sold advertising space on site and helped aggrieved consumers reclaim lost money, and because no consumer would “be confused by a consumer watch-dog type website that is not selling any real estate investment course”); Cintas Corp. v. Unite Here, 601 F.Supp.2d 571 (S.D.N.Y. 2009), aff’d 355 Fed.Appx. 508 (2d Cir. 2009) (per curiam) (rejecting assertion by Cintas that the website cintasexposed.com, run by a labor union and dedicated to criticizing the company’s labor practices, could cause customer confusion). 36   Ascentive, LLC v. Opinion Corp., 842 F.Supp. 2d 450 (E.D.N.Y. 2011).

BARFIELD PRINT.indd 207

13/11/2018 11:52

208  Research handbook on the law of virtual and augmented reality If the definition is broken down into elements, it can generally be simplified into (1) commercial value and (2) indicia of identity. 4.1  Commercial Value The “commercial” aspect of this right is intentional. It is what distinguishes the use of someone’s likeness in creative expression, protected under the First Amendment (such as a movie or song), from commercial speech, which is designed to target consumers for monetary gain. That makes the right somewhat akin to a trademark, and hence within the realm of governmental regulation and property rights. For a plaintiff to prevail on a right to publicity claim, she must typically prove that her identity has “commercial value”—that is, that there is reason to believe that her identity would be worth something to an advertiser, or that a customer might be more likely to pay attention to or purchase a product because the plaintiff’s identity was associated with it. For this reason, courts had long ruled that the right of publicity was only available to “celebrities,” and not the rest of society. Today, however, with the rise of digital (and especially social) media, it is difficult to hold that standard for celebrities alone. It is entirely realistic to argue that individuals can all attain commercial value in some context. One argument for establishing “commercial value” in social media is the value of personal relationships. On many social media sites, the identity of the people with whom one interacts in social media both incentivizes people to participate in the site and adds qualitatively significant value to the experience. And the more such interactions that occur on a particular social media site, the more benefit the owner of that site derives (in terms of advertising revenue, search engine tieins, or whatever the site’s business model may be). For example, social media platforms such as Facebook and Twitter allow users to request “blue-badge verification,”37 often rendered as a small blue check mark next to an individual’s name, proving that their account and identity is legitimate. This type of verification is difficult to obtain and is held strictly for those individuals or businesses that have value in the mainstream media or other public figures who have the potential of having their name or likeness used without their consent. LinkedIn also allows certain users to have the “Influencer” tag next to their account, but this is reserved for those individuals that LinkedIn determines to have high profile accounts.38 Therefore—in a very direct and measurable way, some would argue—digital (and especially social) media is a context in which literally every user’s identity has potential commercial value. Two judicial decisions stemming from lawsuits filed against Facebook in recent years have given some credence to this view,39 as did a lawsuit over a banking executive’s LinkedIn profile.40 37  See Facebook Badge Verification at www.facebook.com/help/100168986860974/?ref=u2u; and Twitter Badge Verification at https://support.twitter.com/articles/119135. 38  See LinkedIn Influencer at www.linkedin.com/help/linkedin/answer/49650/linkedin-influencer?​ lang​=en. 39   Cohen v. Facebook, Inc., 798 F. Supp. 2d 1090 (ND Cal. 2011) and Fraley v. Facebook, Inc., 830 F. Supp. 2d 785 (ND Cal. 2011). 40   Eagle v. Morgan, No. 11-4303. (E.D. Penn. 2013).

BARFIELD PRINT.indd 208

13/11/2018 11:52

Trademark law and publicity in augmented reality  209 4.2  Indicia of Identity The second element of the right of publicity is an individual’s “indicia of identity” being used without their consent. These indicia can take the form of one’s physical appearance, name, signature, or voice. An earlier example of this was argued in the Here’s Johnny Portable Toilets case.41 In that case, the founder of Here’s Johnny Portable Toilets, Inc. admitted that he selected the name for his company because the name was associated with the intro to The Tonight Show, then hosted by Johnny Carson, where the show’s announcer would introduce Mr. Carson with the well-known phrase “here’s Johnny.” In response, Mr Carson sued the defendant for invasion of privacy and his right to publicity, and won. Even after Mr Carson’s death in 2005, the Michigan and California courts recognized the postmortem rights of publicity held by Carson’s estate. To this day, therefore, the company is still bound by the court’s order and cannot use that name for its business. The right of publicity has distinct limits, however, especially when enforcing it would interfere with First Amendment-protected speech. In 2010 and 2014, respectively, actresses Karen Gravano (Mob Wives star and daughter of Salvatore “Sammy the Bull” Gravano) and Lindsay Lohan filed suit against the video game company Rockstar for allegedly using their “likenesses” in the video game Grand Theft Auto 5.42 Lohan argued that one of the characters, “Lacey Jonas”, who was an “unequivocal” reference to the Mean Girls actress, looked exactly like her and had the same background as her—a blond movie star with an eating disorder.43 Specifically, in one of the game’s scenes, the character asks the player to escort her home to avoid the paparazzi. Lastly, Lohan argued that a promotional image for the game depicting the character in a bikini making a peace sign was based on a picture wherein the actress made the same gesture in her swimsuit. Gravano argued that Rockstar took her background story of being a mobster’s daughter afraid of the repercussions of her father aligning himself with the government and publicly criticized by her father for being on a reality TV show. (Similarly, in the game, the character’s father would not let her be on a reality TV show.) The courts in both cases found for Rockstar, holding it did not appropriate either actress’ likeness and that, even if the characters were based on Lohan or Gravano, a video game is a work of fiction and falls under the protection of the First Amendment.

41   John W. Carson, d/b/a Johnny Carson v. Here’s Johnny Portable Toilets, Inc., 698 F.2d 831 (1983); see also 498 F.Supp. 71 (E.D. Mich. 1980). 42  See Lindsay Lohan v. Take-Two Interactive Software, Inc., Rockstar Games, Rockstar Games, Inc., and Rockstar North, 156443-2014 (Supreme Court of New York); see also www.­euro​gamer.net/ articles/2016-09-01-lindsay-lohans-grand-theft-auto-lawsuit-rules-in-rockstars-favour; also www.sil​ ive.com/news/2017/02/ex-mob_wife_karen_gravanos_law.html. 43   See the Lohan image reproduced in Chapter 15 of this book.

BARFIELD PRINT.indd 209

13/11/2018 11:52

210  Research handbook on the law of virtual and augmented reality

5. HOW MIGHT PUBLICITY RIGHTS BE INFRINGED IN AUGMENTED REALITY? 5.1 Facial Recognition as Infringing the Right of Publicity and Related Statutory Causes of Action Facial recognition has long been feasible, but fear of liability has kept its application limited. That self-restraint on the part of tech companies is slowly eroding, as seen by facial recognition’s increasing implementation on Facebook and incorporation into mobile devices such as the iPhone X. Augmented reality technologies will accelerate this trend. Social interaction will remain a major driving force behind the adoption of AR, just as it has been for web and mobile devices. And AR cannot function properly without the device on which it is displayed having the ability to recognize its physical surroundings and adapt its display accordingly. For social AR media, this means the device must be able to recognize individuals at a distance—and that is most easily accomplished via facial recognition. Enterprising plaintiffs’ attorneys have already speculated that an individual may be found to own a right of publicity in their facial features, such that facial recognition technologies will infringe that right unless they first obtain consent. Although we are unaware of this argument having yet been articulated in court (much less endorsed by one), the more ubiquitous facial recognition becomes, the more opportunities and ammunition will arise to fuel such arguments. Nor is the right of publicity the only theory to which facial recognition is vulnerable. A much more clear and present danger to those who would implement facial recognition is the laws already passed by a handful of states that create privacy rights in biometric data and restrict its unauthorized collection and processing. One particularly popular jurisdiction for bringing such claims has been Illinois, which continues (as of this writing) to see class actions filed under the Illinois Biometric Information Privacy Act of 2008 against companies collecting facial data. How far such restrictions against the collection of biometrics—and especially of facial data—will be permitted to expand, and what norms they will eventually set for the rest of the country, remains to be seen. Biometric data is already widely used for entirely utilitarian (and especially security) purposes, such as in mobile devices. Entire social networks and other user-generated content currently employ, and may come to entirely rely upon, the ability to use facial recognition to identify specific individuals. For example, Facebook uses technology to suggest whom users should tag in photos, with an individual’s name automatically appearing. As facial recognition capability becomes more democratized, such that not only corporations but also individuals are able to scan and store such data, courts may extend the First Amendment right of free speech (and its judicially recognized corollary, the right to gather data about which to speak) in a way that protects an individual’s right to collect and annotate their knowledge of others in this manner. Allowing people to own intellectual property or privacy rights in that data might arguably complicate matters too much for that technology to remain useful, to the detriment of society as a whole.

BARFIELD PRINT.indd 210

13/11/2018 11:52

Trademark law and publicity in augmented reality  211 5.2  Individuals as Keywords The previous discussion drawing analogies between search engine keyword advertising and three-dimensional AR can be extended to AR content triggered not only by locations or physical objects, but also by individuals. Keiichi Matsuda’s Hyper-Reality provides an apt example. Nearly every passerby in the film is shown with a digital symbol floating above their heads. Although this content is not explored in the brief video, one can imagine that at least some of this content is commercial advertising triggered by something specific about the individual. Perhaps the viewer is offered the opportunity to digitally purchase the outfit or makeup another person is wearing. If a specific individual has appeared in a program or film, perhaps passersby would be offered a link to that content. Or links to a person’s social media or dating app profiles may be depicted as floating over their heads, inviting others to connect with them. Each of these applications exists in two-dimensional form and has been speculated about in works of commentary; their implementation is only a matter of time. In these circumstances, however, how ought the profits be split? Will the individuals with whom this interactive content is associated be treated as “sponsors” or “brand ambassadors,” and thus independent contractors entitled to a share of the profits from any transaction they promote? If commercial transactions are initiated by a person’s presence without the person’s knowledge, would that be an unauthorized exploitation of the person’s identity? Time—and litigation—will tell. 5.3 Three-Dimensional Capture of Entire Bodies: Sex Appeal and the Right of Publicity Traditional biometric indicators may not be the only way in which augmented technologies catalog and exploit individuals’ physical attributes. Mass-market devices like Microsoft’s Kinect are already designed to recognize entire bodies. Today, there are companies simultaneously using more than 60 sensors more precise than the Kinect to digitally render individuals in real time with amazing accuracy. The year 2014 saw the introduction of the Kickstarter-funded Structure Sensor—an iPad accessory that allows the device’s camera to capture three-dimensional imagery from its surroundings in real time—and Google’s Project Tango, an experimental depth sensor that also renders ambient surroundings in 3D. Similar capabilities now come as standard in top of the line iPhones and other mobile devices. AR applications will take advantage of such capabilities in order to superimpose digital data on a person’s entire body. Many of these will be benign; entire markets will develop for virtual clothing and accessories, for instance. Security professionals already scan entire bodies for contraband, and have made progress in identifying individuals based on their gait as well.44 Other applications, however, will go beyond merely analyzing images of bodies to storing and repurposing those images. In an age where sexting is epidemic among teens and states such as California are forced to outlaw the salacious repurposing of such

44   “Gait Biometrics Shows Promise,” Homeland Security News Wire, Sept. 8, 2011, available at www.homelandsecuritynewswire.com/gait-biometrics-shows-promise.

BARFIELD PRINT.indd 211

13/11/2018 11:52

212  Research handbook on the law of virtual and augmented reality two-dimensional images (that is, as “revenge porn”),45 little imagination is required to conceive of the unsavory uses to which 3D personal imaging technologies could be put. If it can be done, it will be, as evidenced by the scandals involving TSA agents who stored 3D images of travelers taken through airport security scanners.46 Already, pornography is one of the leading use cases for the adoption of virtual reality goggles,47 and it was one of the few purposes to which a large segment of consumers were ready to put Google Glass.48 One independent filmmaker in Florida has published a short film called Ex Post Facto centered entirely around a fictional set of AR glasses capable of recording a woman within the wearer’s line of sight and then converting that data into a seductive digital avatar.49 It is difficult to imagine that similar imagery would not propagate rapidly—both consensually and otherwise—throughout any form of AR network. To date, in courts across the country, one of the most frequent reasons for invoking the right of publicity has been to enjoin the prurient use of girls’ and women’s images,50 which are often recorded unwittingly. It is logical to expect the same laws to be applied when those images are collected and manipulated by new digital media. 5.4  Virtual Assistants as Infringement Apple has “Siri.” Microsoft has “Cortana.” Amazon has “Alexa.” Google has “Google Assistant.” The Oscar-nominated film Her featured Samantha, while Ender Wiggins (in the sequels to Ender’s Game) had Jane. Vital to each iteration of the starship Enterprise was the Computer. Futurists have long anticipated the day when humans would be able to interact with computers using the same conversational speech individuals normally reserve for other people. In order to offer truly two-way interaction of this type, however, the programs need to be able to respond in kind. In other words, they need to seem more human. The virtual assistants and chatbots available as of this writing are beginning to approximate that experience. Siri has already reached its fourth generation, and in 2016 Samsung acquired “Viv,” a next-gen AI assistant built by the creators of Apple’s “Siri,” which is said to be “blindingly smart,” “infinitely flexible,” “omnipresent,” and “­embedded in a plethora of Internet-connected everyday objects.”51

  See California Penal Code 647(j)(4) PC.   See, e.g., “One Hundred Naked Citizens: One Hundred Leaked Body Scans,” Gizmodo http://gizmodo.com/5690749/these-are-the-first-100-leaked-body-scans (Nov. 16, 2010). 47   “Porn Industry Will Drive Mainstream Adoption of Virtual Reality,” www.pcr-online.biz/ news/read/porn-industry-will-drive-mainstream-adoption-of-virtual-reality/036278 (June 1, 2015). 48   “Google Glass May Be the Future of the Porn Industry,” www.huffingtonpost.com/2014/01/25/ google-glass-porn_n_4662944.html (Jan. 25, 2014). 49   Brian D. Wassom, Augmented . . . Rape? New Short Film Explores The Really Dark Side Of AR, www.wassom.com/ex-post-facto.html (Aug. 21, 2014). Fortunately, the video itself has since been removed from YouTube. 50   See, e.g., Arnold v. Treadwell, 642 F. Supp. 2d 723 (ED Mich 2009) (refusing to dismiss claim brought by aspiring model); Balsley v. LFP, Inc., 691 F. 3d 747 (6th Cir. 2012) (news anchor filmed during wet t-shirt contest). 51   Steven Levy, “Siri’s Inventors Are Building a Radical New AI that Does Anything You Ask,” Wired, Aug. 12, 2014, available at www.wired.com/2014/08/viv/. 45 46

BARFIELD PRINT.indd 212

13/11/2018 11:52

Trademark law and publicity in augmented reality  213 But users and UX designers alike crave more humanlike interaction. AR will offer a unique avenue for achieving this goal by adding visual (and perhaps other sensory) elements to digital companions. In the very near future, Siri-like assistants will be available that consumers can actually see and communicate with face-to-face, via their digital eyewear. Those designing virtual people will inevitably mimic real people, and this will potentially infringe the publicity rights of their muses. The simple case in this hypothetical is the company that offers the likenesses of recognizable celebrities as “skins” for a virtual assistant. That would amount to a commercial exploitation of an identity with defined monetary value, a straightforward infringement. Today, Waze provides users with the options to choose the voices of Owen Wilson, Armie Hammer, Keith Morrison, Morgan Freeman, and many others. But what if the real-life inspiration behind the program is less recognizable, the similarity is less than exact, or the imitation is unauthorized? Digital avatars will be fertile ground for right of publicity claims. These theories are likely to have greater resonance with respect to virtual assistants because of the commercial, utilitarian nature of the function that they perform. To be sure, similar issues will arise in more artistic contexts as well. In fact, consumers have already seen several analogous claims in recent years, involving “holograms” of such deceased celebrities as Tupac Shakur, Marilyn Monroe, Amy Winehouse, Freddie Mercury, and Michael Jackson. Publicity rights objections by Winehouse’s next of kin in 2014 put a stop to her hologram before it began. College athletes in both New Jersey and California have recently won publicity rights lawsuits against video game manufacturers that i­ ncorporated the players’ likenesses into football games. And several celebrities, including Bette Midler, Tom Waits, the Romantics, and Arnold Schwarzenegger have asserted (with varying degrees of success) publicity rights claims against those with soundalike voices. In conclusion, it bears reiterating that no one can predict with confidence ahead of time where the line between lawful expression and infringement of trademark or publicity right will fall in any given factual scenario; nor do we attempt to do so here. What is certain, though, is that widespread adoption of AR by potential consumers will bring with it commercially subsidized speech and advertising content, with a flurry of litigation activity soon to follow. As was the case with the internet before it, advertisers will not be able to resist inventing new business models for delivering commercial messages, and people whose interests and expectations are upset by these developments will certainly push back. Owners of physical real estate will take umbrage at the use of AR to depict commercial content as if it is “on” their property without their permission. Companies that paid handsomely for physical billboards will rebel against the idea of third parties overlaying those ads with their own digital content. Echoing the years of litigation spurred by keyword advertising on search engine websites, brand owners will certainly complain when their famous logos and icons are used to trigger the display of competitive marketing. In each case, the aggrieved parties will argue that this unauthorized association between the physical and the digital creates a likelihood of customer confusion in violation of the Lanham Act and other sources of trademark rights. Resolving these cases will require a keen understanding of evolving customer expectations and their understanding of this new medium, coupled with a healthy respect for the limitations on intellectual property protection imposed by free speech principles.

BARFIELD PRINT.indd 213

13/11/2018 11:52

214  Research handbook on the law of virtual and augmented reality Similar complaints will be made by those individuals—both famous and ordinary— who find their personal identities augmented in unauthorized and unwelcome ways. AR devices that use facial recognition or other identifying data to recognize an individual and associate commercial messages with that person—a personalized form of keyword advertising—will likely give rise to right of publicity litigation. Cooption of individuals’ physical features for more nefarious or prurient purposes will also become an unfortunate fact of life. If recent litigation is any guide, the right of publicity will remain one of a growing arsenal of legal tools available to combat such content. It is also safe to predict that some courts, especially in the early days of AR adoption, will overreact to the potential for confusion in the medium. Like the internet before it, AR will foster a wide variety of new approaches to commercial messaging. The good news, though, is that these bodies of law have already been stretched, shaped, and refined enough by previous digital media that the innovators and entrepreneurs of AR ought to be able to weather these challenges just fine—with, of course, guidance by the right legal counsel.

SOURCES Cases Abercrombie & Fitch Co. v. Hunting World, 537 F.2d 4 (2nd Cir. 1976). Arnold v. Treadwell, 642 F. Supp. 2d 723 (E.D. Mich 2009). Ascentive, LLC v. Opinion Corp., 842 F. Supp. 2d 450 (E.D.N.Y. 2011). Balsley v. LFP, Inc., 691 F. 3d 747 (6th Cir. 2012). Brown v. Entertainment Merchants Association, 131 S. Ct. 2729 (2011). Bally Total Fitness Holding Corp. v. Faber, 29 F.Supp.2d 1161, 1163–4 (C.D. Cal. 1998). Candy Lab, Inc. v. Milwaukee County, No 17-CV-569-JPS (E.D. Wis. July 20, 2017). Carson v. Here’s Johnny Portable Toilets, Inc., 698 F.2d 831 (1983). Cintas Corp. v. Unite Here, 601 F.Supp.2d 571 (S.D.N.Y. 2009), aff’d 355 Fed.Appx. 508 (2d Cir. 2009) (per curiam). Cohen v. Facebook, Inc., 798 F. Supp. 2d 1090 (N.D. Cal. 2011). Eagle v. Morgan, No. 11-4303. (E.D. Penn. 2013). Edina Realty, Inc. v. Themlsonline.com, Civ. 04-4371JRTFLN; 2006 U.S. Dist. LEXIS 13775 (D. Minn. March 20, 2006). In re E.I. du Pont de Nemours & Co., 476 F.2d 1357 (C.C.P.A. 1973). Fin. Express LLC v. Nowcom Corp., 564 F. Supp. 2d 1160, 1177 (C.D. Cal. 2008). Fraley v. Facebook, Inc., 830 F. Supp. 2d 785 (N.D. Cal. 2011).MCW, Inc. v. Badbusinessbureau.com, L.L.C., No. 02 Civ. 2727, 2004 WL 833595, at *16 (N.D. Tex. Apr. 14, 2004). Lohan v. Take-Two Interactive Software, Inc., Rockstar Games, Rockstar Games, Inc., and Rockstar North, 156443-2014 (Supreme Court of New York) (2014). Rogers v. Grimaldi, 875 F. 2d 994, 998 (2d Cir. 1989). Rosetta Stone Ltd. v. Google, Inc., 676 F.3d 144 (4th Cir. 2012). Storus Corp. v. Aroa Mktg., Civ. No. 06-2454-MMC; 2008 U.S. Dist. LEXIS 11698, at *12–13 (N.D. Cal. Feb. 15, 2008). Taubman Co. v. Webfeats, 319 F.3d 770, 777–8 (6th Cir. 2003). Taylor Bldg. Corp. of Am. v. Benfield, 507 F.Supp.2d 832, 847 (S.D. Ohio 2007). Two Pesos v. Taco Cabana, Inc., 505 U.S. 763, 768 (1992). Whitney Inf. Network, Inc. v. Xcentric Ventures, No. 2:04-cv-47-FtM-34SPC, 2005 WL 1677256 (M.D. Fla. July 14, 2005). 1-800-Contacts, Inc. v. Lens.com, Inc., 722 F.3d 1229 (10th Cir. 2013).

BARFIELD PRINT.indd 214

13/11/2018 11:52

Trademark law and publicity in augmented reality  215 Statutes California Penal Code 647(j)(4) PC. Lanham Act, 15 U.S.C. §1051 et seq. 15 U.S.C. §1114, 1125. 15 U.S.C. §1125(c)(2)(A). 15 U.S.C. §1125(c)(2)(B).

BARFIELD PRINT.indd 215

13/11/2018 11:52

7.  Trade dress in virtual worlds Jack Russo*

1. INTRODUCTION There are important intellectual property issues in newly emerging virtual worlds. Those issues relate to the design and use of creative elements in virtual, augmented, and mixed reality environments. This chapter focuses on the law of trade dress. Trade dress generally refers to characteristics of the visual appearance of a product and its packaging that signify the source of the product or services to consumers. The law of trade dress may be particularly relevant for virtual, augmented, and mixed reality worlds given the visual presentation of images in virtual worlds, and the numerous applications which produce the “look and feel” of real-world commercial businesses and real-world commercial products and services within virtual worlds. This chapter focuses on applicable federal statutes and case law applicable to trade dress within virtual, augmented, and mixed reality environments. The federal Lanham Act nowhere expressly defines the term “trade dress,”1 but it does broadly define “trademark” as “any word, name, symbol, or device, or any combination thereof used to identify goods.”2 The Supreme Court of the United States has made clear that “any combination thereof ” is more than sufficient to broadly protect trade dress for both products as well as services.3 Indeed, trade dress protection has extended beyond the United States into European Union and other jurisdictions.4 What role then will trade

*  The author gratefully acknowledges the assistance of Mr Alex Chan, a law student, who previously clerked at Computerlaw Group LLP where Mr Russo has his intellectual property law practice, as well as Ms Umeet Sajjan, who currently clerks with Mr. Russo at his Silicon Valley law offices. 1   Trademark Act of 1946, ch. 540, 60 Stat. 427 (codified as amended at 15 U.S.C. §§ 1051–1127 (1994)). 2   15 U.S.C. § 1127 (1994) (emphasis added); statutory authority is the Constitution’s Commerce Clause. 3   See Two Pesos, Inc. v. Taco Cabana, Inc., 505 U.S. 763, 773 (1992). 4   See Reckitt & Colman Ltd v Borden, Inc. [1990] 1 All E.R. 873 (the “Jif Lemon” case) (establishment of passing off requires showing of all of reputation and goodwill, misrepresentation, and damage); Moroccanoil Israel Ltd v Aldi Stores Ltd [2014] EWHC 1686 (29 May 2014) (no passing off claim was established because, despite the similarities between the two products in question, there had been no assumption of get-up such that nothing more than mere confusion was proved); Philips v Remington (C-299/99) (no “capricious additions” required to achieve distinctiveness); K. Purnima, Trade Dress Protection for “Look and Feel” of the Website, India Law Journal (retrieved October 13, 2015); and AB-InBev v. XYM, SPC No 1183 (September 26, 2014) (China’s highest court affirms trade dress infringement in spite of the defendant’s enforceable design patent).

216

BARFIELD PRINT.indd 216

13/11/2018 11:52

Trade dress in virtual worlds  217 dress play in coming mobile computer-based “realities” presented through augmented “real”5 as well as “artificial” virtual worlds? 6 The thesis here is simple and straightforward: In our new mobile computing-based virtual worlds of eXtended Reality (“xR”)7 in all its forms, including virtual reality (“VR”),8 augmented reality (“AR”),9 and mixed reality (“MR”),10 potentially broader “look and  5   One can think of high-powered binoculars as a form of “augmented” reality of distant images; when one adds computer-based technologies to binoculars far more is possible, as was demonstrated publicly more than 25 years ago. See Jon Markoff & Martin Fackler, “Cellphone as a Link from the Web to the World,” New York Times at www.nytimes.com/2006/06/28/ technology/28iht-locate.2071028.html (“The idea came . . . in 1991 on a sailboat off the coast of Mexico. To compensate for the elder Ellenby’s poor sense of direction, the two men decided that by tying together a compass, a Global Positioning System receiver and binoculars, it would be possible simply to point at an object or a navigational landmark to identify it”).  6   We use “virtual worlds” broadly to include all emerging forms of virtual, augmented, and mixed reality applications, including those anticipated by video arcade and then personal computer-based videogames decades ago. National Research Council, Virtual Reality: Scientific and Technological Challenges 13 (1995) (“In virtual reality (VR) or virtual environment (VE) systems, the human operator is connected to a computer that can simulate a wide variety of worlds, both real and imaginary”). See also Brenda Laurel, Virtual Reality, Scientific American (Sept. 1995), at 90 (“The important thing about VR is what it does rather than how its effects are achieved: it permits people to behave as if they were somewhere they are not”). Compare Howard Rheingold, Virtual Reality 17 (1991) (“In the future, less intrusive technologies will be used to create the same experience, and the computers will be both more powerful and less expensive, which means the virtualities will be more realistic and more people will be able to afford to visit them”).  7   See www.foundry.com/industries/virtual-reality/vr-mr-ar-confused (an overview of the differences between VR, AR, and MR demonstrates how “xR” is an umbrella for all forms of immersive and quasi-immersive experiences). See also https://en.wikipedia.org/wiki/Extended_reality (xR refers to “all real-and-virtual combined environments and human-machine interactions generated by computer technology and wearables”).  8   “While AR adds visible digital content to a person’s perception of the real world, VR replaces the real world altogether. Using goggles and speakers, VR places people inside a virtual environment, letting them move around in it and interact with it as if it were the real world.” Mark A. Lemley and Eugene Volokh, Law, Virtual Reality, and Augmented Reality, 166 University of Pennsylvania Law Review 1105 (2018); Stanford Public Law Working Paper No. 2933867; UCLA School of Law, Public Law Research Paper No 17-13. Available at SSRN https://ssrn.com/abstract=2933867 or http://dx.doi.org/10.2139/ssrn.2933867. See also https://en.wikipedia.org/wiki/Virtual_reality (VR is a “computer technology that uses virtual reality headsets or multi-projected environments, sometimes in combination with physical environments or props, to generate realistic images, sounds and other sensations that simulate a user’s physical presence in a virtual or imaginary environment”).  9   AR is “an enhanced version of reality created by the use of technology to overlay digital information on an image of something being viewed through a device (such as a smartphone camera).” Donald J. Kochan, Playing with Real Property Inside Augmented Reality: Pokémon Go, Trespass, and Law’s Limitations, 38 Whittier Law Review 70 (2018) (symposium) (quoting Merriam-Webster Online Dictionary); Chapman University, Fowler Law Research Paper No. 17-01, available at SSRN: https://ssrn.com/abstract=2919109. See also https://en.wikipedia.org/ wiki/Augmented_reality (AR is a “live direct or indirect view of a physical, real-world environment whose elements are ‘augmented’ by computer-generated or extracted real-world sensory input such as sound, video, graphics, haptics or GPS data”) and http://virtualrealitysummit.com/ news/airbnb-will-include-ar-vr-on-website-to-build-trust-between-hosts-guests/43014/ (will trust between Airbnb and its customers ultimately become part of the hospitality service provider’s goodwill thanks to VR, AR, and MR?). 10   MR (or “Mixed Reality”) is “the merging of the Internet with the physical world around

BARFIELD PRINT.indd 217

13/11/2018 11:52

218  Research handbook on the law of virtual and augmented reality feel” protection under federal (and state) trade dress principles will expand to provide protection to goodwill-generating computer-based works—even when those works are otherwise unprotected (and even when not protectable) by patent or copyright.11 How far trade dress law expansion will take us is open to debate but I predict that the federal courts will address multiple new questions about the scope of available trade dress protection and the role that the First Amendment will play in providing a broader (if not broadest) form of “fair use” as defense.12 The degree of “immersion” that each type of xR interactive technology enables for end users may well dictate the degree of protection or nonprotection that will be allowed the trade dress owner. For example, purely fictional forms of completely immersive VR may well be allowed broader “license” under broadening forms of First Amendment (that is, free speech) immunity than is the case when limited AR simply enhances a shopping experience. The federal courts—with the assistance of industry experts, informed legal counsel, and the community and its standards—will ultimately have to become far more expert in the scope of protection (and immunity) provided (or not) in virtual worlds; the case by case development now in its infancy must balance the continuing expansion of trade dress against First Amendment rights to express ideas, particularly in areas which neither patent nor copyright law necessarily cover. In trade dress protection, when asserting trade dress infringement or unfair competition, there must be a showing of the nonfunctionality of its trade dress. Therefore, a trade dress cannot be afforded protection if it is purely functional. As the Supreme Court has said, “the functionality doctrine prevents trademark law [including trade dress law], which seeks to promote competition by protecting a firm’s reputation, from instead inhibiting legitimate competition by allowing a producer to control a useful product feature.”13 This aspect can raise a question in the context of extended reality: Though published functionality alone will never be legally protected through trade dress (other than in us.” Joshua Fairfield, Mixed Reality: How the Laws of Virtual Worlds Govern Everyday Life, 27 Berkeley Technology Law Journal, p. 55, 2011; Washington & Lee Legal Studies Paper No 2012-23. Available at SSRN: https://ssrn.com/abstract=2064922. See also https://en.wikipedia. org/wiki/Mixed_reality (also known as hybrid reality, MR is “the merging of real and virtual worlds to produce new environments and visualizations where physical and digital objects co-exist and interact in real time . . . encompassing both virtual reality and augmented reality”). See also https://www.recode.net/2015/7/27/11615046/whats-the-difference-between-virtual-augmented-an​ d-mixed-reality (AR, VR, and MR are different, and they pursue, respectively, presence, utility, and flexibility.) 11   For a review of the history and scope of copyright protection, see Jack Russo and Michael Risch, Virtual Copyright (October 12, 2017). The Law of Virtual and Augmented Reality (Woody Barfield, Marc Blitz eds, 2018 forthcoming). Available at SSRN: https://ssrn.com/ abstract=3051871. 12   Unlike the US Copyright Act’s detailed statutory “fair use” as a complete defense to even willful copying (see 17 U.S.C. 108), there is no such detailed statutory defense in the Federal Lanham Act. Compare, e.g., Volkswagenwerk Aktiengesellschaft v. Church, 411 F.2d 350 (9th Cir.), supplemented, 413 F.2d 1126 (9th Cir. 1969) (unauthorized “VW” auto repair service provider did not infringe on automaker’s trademarks for use of its registered “VW” trademarks were for descriptive fair use in good faith), with Sony Corp. of Am. v. Universal City Studios, Inc., 464 U.S. 417 (1984) (the “Betamax case”) (home videotape recorder manufacturers did not commit contributory infringement of television program copyrights for it constituted statutory fair use to allow private viewers to shift broadcast time of television programs). 13   Qualitex Co. v. Jacobson Products, Co., Inc., 514 U.S. 159 (1995).

BARFIELD PRINT.indd 218

13/11/2018 11:52

Trade dress in virtual worlds  219 appropriate cases by patent law),14 the distinctive organization and presentation of functionality as part of the overall “look and feel” of virtual-world elements is potentially protected when commercial association and protectable goodwill have been generated; accordingly, unless entirely functional, a first mover’s distinctive trade dress in virtual worlds will garner protection under trade dress principles now well established under federal (Lanham Act) as well as state (Uniform Act15) trademark, service mark, and trade dress law. These protections can and will survive loss of trade secret protection (by publication), loss of patent protection (by ineligibility, invalidity, expiration or abandonment), as well as loss of copyright protection (by failure of expressive content, fair use, or other, defense). As with trademark protection generally, trade dress protection can last forever. Depending on the scope of protection, certain virtual worlds can achieve monopolylike status on a perpetual basis (just as now exists in some distinctive real-world trade dress).16 Accordingly, though federal law (and federal preemption) will continue to forbid nonpatent protection for published functionality alone, trade dress protection in virtual worlds provides a vehicle for goodwill protection of distinctive VR, AR, and MR works that are not purely functional. Closer questions will emerge about degrees of protection of VR works over MR works (and both over AR works) with the appropriate detailed dissection by the courts as assisted by appropriate experts and by industry standards, customs, and usage.17 I submit that entirely new fields of IP expertise will arise to help federal and state courts tackle these trade dress issues in the near future.18

14   See, e.g., TrafFix Devices, Inc. v. Mktg. Displays, Inc., 532 U.S. 23, 33 (2001) (“the trade dress in [Two Pesos, Inc. v. Taco Cabana, Inc., 505 U.S. 763 (1992)] did not bar competitors from copying functional product design features. In the instant case, beyond serving the purpose of informing consumers that the sign stands are made by MDI (assuming it does so), the dual-spring design provides a unique and useful mechanism to resist the force of the wind”). Of course, unpublished functionality is potentially protectable under state (as well as now federal) trade secret law if appropriate steps are taken to maintain secrecy. See Defend Trade Secrets Act  of 2016 (Pub.L. 114–53, 130 Stat. 376, enacted May 11, 2016, now 18 U.S.C. § 1836, et seq.) (federal protection of trade secrets in interstate commerce cases). 15   Though variations exist in the 50 states, most adopt a version of the  Model Trademark Bill (MTB), which provides for trademark and trade dress registration, or the Uniform Deceptive Trade Practices Act (UDTPA), which is comparable to Section 43(a) of the Lanham Act. 16   See, e.g., Isabelle Fraser, “Banksy Bans Disney Lawyers from his ‘Dismaland’ Theme Park,” The Telegraph (August 20, 2015), www.telegraph.co.uk/culture/banksy/11814061/Banksybans-Disney-lawyers-from-his-Dismaland-theme-park.html; compare “Disney Faces a Dismal Conundrum over New Banksy Art Project,” World Trademark Review (August 21, 2015), www. worldtrademarkreview.com/Blog/detail.aspx?g=d48eb8f3-e136-41ae-a1cf-454c861db734. 17   See, e.g., In re General Mills IP Holdings II, LLC, 124 USPQ2d 1016 (TTAB 2017) (trade dress registration of yellow-colored packaging denied because third-party usage of yellow packaging for breakfast cereals demonstrated that consumers “do not perceive the color yellow as having source significance for the goods”); compare Qualitex Co. v. Jacobson Products Co., 514 U.S. 159 (1995). 18   See Schuyler Moore, “The Legal Reality of Virtual Reality,” Forbes (Mar. 2017) reprinted at www.forbes.com/sites/schuylermoore/2017/03/10/the-legal-reality-of-virtual-reality/#f66124a20495.

BARFIELD PRINT.indd 219

13/11/2018 11:52

220  Research handbook on the law of virtual and augmented reality

2.  KEY PRINCIPLES SUMMARIZED Given the case law developments I describe below, spanning the past six decades, the following principles—among others—for trade dress law can be readily ascertained. They will guide the summary and more extended discussion that follows. ●

Competitive principles require that published ideas alone, when not patented under federal law,19 and when not maintained as trade secrets under state (and now federal) trade secret law, may not be protected and cannot be protected by federal copyright or by federal (or state) trademark or trade dress law; federal preemption bars that.20 ● What is “idea” and what is “expression” in particular virtual worlds will continue to be determined on a case by case basis using the “abstractions” test (the purpose of the abstraction test is to identify which aspects of the virtual world constitute expression and which are ideas), the merger doctrine (the idea–expression divide or the idea–expression dichotomy limits the scope of copyright protection by differentiating an idea from the expression or manifestation of that idea), and published industry standards under the copyright scènes a faire copyright fair use, or trade mark and trade dress descriptive use doctrines (under the doctrine of descriptive fair use, a party, even a direct competitor, may use another party’s descriptive trademark to describe their own product or services).21 ● Original expression in the “look and feel” of computer software generating virtual worlds is generally protected by copyright but subject to statutory limitations providing “fair use” and other copyright defenses; comparatively, however, the inherently distinctive nonfunctional trade dress in such presentations has received less judicial attention to date under applicable law.22 ● In the Copyright Act context, what is “original” raises two questions: First, has the work been independently developed, that is, not simply copied from another source? Second, even if independent, does the work have “some” greater than de minimis level of creativity, that is, in the words of Feist,23 is the work more than “garden variety,” not “so mechanical or routine as to require no creativity whatsoever,” not “devoid of even the slightest trace of creativity,” and not one “in which the creative spark is utterly lacking or so trivial as to be virtually nonexistent”?

19   Important issues have also now emerged regarding whether published functional ideas—even when organized in novel and nonobvious computer software applications—are eligible subject matter for patent protection. See, e.g., Alice Corp. Pty. Ltd. v. CLS Bank Intern., 134 S. Ct. 2347 (2014); see also A. Underhill, “Who Is Alice, And Why Is She Driving Patent Attorneys Mad As Hatters?” reprinted at https://abovethelaw.com/2016/02/who-is-alice-and-why-is-she-driving-patent-attorneys-mad-as-hatters/. 20  See Bonito Boats, Inc. v. Thunder Craft Boats, Inc., 489 U.S. 141 (1989). 21  Compare Lotus Development Corp. v. Borland International, Inc., 799 F. Supp. 203, 216–219 (D. Mass. 1992) (electronic spreadsheet with menus is idea but overall Lotus 1-2-3 menu-oriented user interface is copyrighted expression) with Apple Computer, Inc. v. Microsoft Corporation, 799 F. Supp. 1006, 1022-1026 (N.D. Cal. 1992) (overall desktop metaphor of Apple Macintosh graphical user interface is idea and dissection of particular elements found to be either licensed material or unprotected ideas). 22  See Feist Publications, Inc. v. Rural Telephone Service Company, 499 U.S. 340 (1991). 23   Feist Publications v. Rural Telephone Service Company, 499 U.S. 340 (1991).

BARFIELD PRINT.indd 220

13/11/2018 11:52

Trade dress in virtual worlds  221 ●

In the Lanham Act context, original trade dress in the “look and feel” of virtual works, when inherently distinctive, is additionally (or alternatively) protected as trade dress under the Lanham Act; indeed, Lanham Act protection may become the next major vehicle for the broadest available legal protection.

Under the above principles, subject to statutory fair use and First Amendment defenses, legal protection extends to all original expression as well as all distinctive trade dress in a virtual world work’s user interface, including potentially the original expression (and/or distinctiveness) in the selection, organization, and presentation of user visible functions. Though a single user visible command or menu button or other common element will not be protected, potentially an entire collection of user visible functions and features if distinctive will be protected once goodwill attaches—even color alone has been found sufficient for this.24 The distinctive selection, organization, and presentation of even just such elements is capable of being the “word, name, symbol, or device, or any combination thereof” which identifies the source of the goods or services and which acts as the vessel of its goodwill and quality protection. The general principles recited above, as well as the underlying broad federal statutory scope and the cases interpreting them, all support this treatment of trade dress law for virtual worlds.

3. NATURAL EXTENSION OF HISTORICAL TRADE DRESS LAW In the United States, the underlying trade dress jurisprudence may be traced back six decades, to an appeal by a Scotch whisky manufacturer before the Commissioner of Patents of the US Patent and Trademark Office. In Ex parte Haig & Haig Limited,25 the Commissioner reversed the decision of the Examiner of Trademarks, whom the Commissioner supervised, and approved Haig & Haig’s application to file on the Principal Register as a trademark for whisky the renowned bottle known as the “Pinch Bottle” (Figure 7.1). The term “pinch” for whisky was, of course, fully protected as a word-based trademark; the larger question in the appeal was whether the bottle itself—even stripped of the “pinch” text—was itself protectable, even though its design also functionally provided an easier way to hold the bottle itself. The Examiner’s decision was grounded on the opinion that Haig & Haig sought registration of a whisky container which was not inherently distinctive and was registrable at most only on the Supplemental Register.26 Note that in the US the Supplemental Register is the secondary register of

 See Qualitex Co. v. Jacobson Products Co., 514 U.S. 159 (1995).   Ex Parte Haig & Haig Ltd., 118 U.S.P.Q. 229 (BNA) (Com’r Pat. & Trademarks July 14, 1958). 26   In the United States Trademark Office, the Supplemental Register (as compared to the Principal Register) exists to give notice to the world that a trademark (and/or trade dress) owner seeks to protect and register a mark that is in the process of developing “secondary meaning” and that once such development (typically over a period of time) occurs then the mark is allowed on the Principal Register once “secondary meaning” is achieved. The decision appeared simply to challenge whether the whisky manufacturing giant had not submitted its trademark on the right 24 25

BARFIELD PRINT.indd 221

13/11/2018 11:52

222  Research handbook on the law of virtual and augmented reality

Figure 7.1  “Pinch” bottle and packaging trademarks maintained by the US Patent and Trademark Office. It was established in 1946 by Subchapter II of the Lanham Act to allow the domestic registration of trademarks that do not meet all the requirements for registration on the Principal Register, so that holders of such marks could register them in another country. Questioning the Examiner’s per se “unregistrable” determination, the Commissioner pointed out the question at issue to be resolved was not whether containers were registrable on which register but whether the container was a registrable trademark or device, the former being a symbol and the latter being an identifier distinguishing Haig & Haig’s products from its competitors; this, as the Commissioner suggested, was a matter of fact that had to be ascertained on an ad hoc case by case basis, not a matter of law as the Examiner purported. On the facts, the subject matter of Haig & Haig’s trademark application was not about the container per se, but about the symbolism of it. To whisky consumers, the bottle alone was the brand. If the foregoing description does not focus the issue, a picture of the bottle will (see Figure 7.1). In reliance of a 1912 decree of injunction entered by the Superior Court of Cook County, Illinois, the Commissioner found that Haig & Haig had purposefully adopted and manufactured a bottle that was “triangular in shape, about eight inches in height, with each of the three sides of the bottle pinched in, or indented toward the center of the same, and so indented and pinched in that indentation occupied the greater part of each of the sides of the bottle.” Accordingly, by placing its whisky in “certain package of original, distinctive and peculiar appearance, or dress,” Haig & Haig’s conformation of the Pinched Bottle had been so successful as a marketing strategy that it had formed the common knowledge of the trade, dealers, consumers, and public at large since as early as 1912, such that there existed “no way of identifying or asking for such brand of product other than by describing the contour or conformation of the container.” The “pinch” case arguably opened a registry. The Examiner made reference to Ex parte The Boye Needle Company, 100 U.S.P.Q. 124 (Com’r., 1953) and Ex parte Minnesota Mining and Manufacturing Co., 92 U.S.P.Q. 74 (Com’r., 1952) in supporting the ruling.

BARFIELD PRINT.indd 222

13/11/2018 11:52

Trade dress in virtual worlds  223 Table 7.1  Key Supreme Court cases ISSUE

KEY POINT

CASE

Preempted? Merely factual? Inherently distinctive? Aesthetic not functional? Purely functional? Artistically functional? Offensively distinctive?

Barred by federal law Barred by public domain Protectable by trade dress Protectable by trade dress Barred by patent law Protectable by copyright First Amendment protected

Bonito Boats Feist Taco Cabana Qualitex TrafFix Star Athletica Matal v. Tam

new era of trade dress protection offered by the Lanham Act—even though “trade dress” is not mentioned in the Act nor in the case itself. What happens, then, when virtual worlds choose elements that create or adopt trade dress? The United States Supreme Court will eventually have to address whether the First Amendment will allow virtual worlds to provide more immediate sensory connection to items that also have valuable goodwill associations. How then would the federal courts, including the Supreme Court, further develop and, in particular, decide the scope of trade dress protection for virtual worlds? In the more than six decades since the “PINCH” case, the Supreme Court has decided a line of cases—to be discussed in more detail presently— of importance to this question; they may be summarized as in Table 7.1. We address each, in turn, below in some detail as each provides (and they collectively provide) a general roadmap to our thesis that in virtual-world cases the expansive trade dress protection principles found in historical law will eventually be subject to limits imposed by fair use and the First Amendment even in commercial contexts. Importantly, though none of these Supreme Court cases directly address virtual-world protection or even the broader category of computer software user interfaces, the cases have important holdings that will be critical to the further development of virtual-world trade dress law,27 and to the degree to which fair use and the First Amendment will play an important and growing role in allowing virtual-world development over time.28 3.1  Bonito Boats v. Thunder Craft Boats In Bonito Boats, Inc. v. Thunder Craft Boats, Inc.,29 the Supreme Court held that a Florida statute protecting boat hull designers from reverse engineering of their boat hulls conflicted with federal patent law and was therefore preempted. There, no patent application had been filed on the design or manufacturing process used by Bonito Boats. When the Florida legislature enacted a statute prohibiting use of 27   See Tom W. Bell, Virtual Trade Dress: A Very Real Problem, 56 Maryland Law Review 384 (1997), available at http://digitalcommons.law.umaryland.edu/mlr/vol56/iss2/4 (arguing that trade dress raises the potential for monopolies to arise from larger companies wrongfully extending trade dress beyond traditional limits). 28   See Dorean M. Koenig, Joe Camel and the First Amendment: The Dark Side of Copyrighted and Trademark-Protected Icons, 11 Thomas M. Cooley Law Review 803 (1994). 29   Bonito Boats, Inc. v. Thunder Craft Boats, Inc., 489 U.S. 141 (1989).

BARFIELD PRINT.indd 223

13/11/2018 11:52

224  Research handbook on the law of virtual and augmented reality the direct molding process to duplicate unpatented boat hulls and forbidding knowing sales of hulls so duplicated, the Bonito Boats hull model in question had been on the market for six years. In its opinion, the High Court held that the Florida statute offered patentlike protection to ideas that were unprotected under federal patent law. In offering protection to ideas that were unpatentable, the Florida statute restricted the public’s ability to exploit unpatented designs and conflicted with the strong public policy favoring free competition in ideas that do not merit patent protection. The High Court stated: From their inception, the federal patent laws have embodied a careful balance between the need to promote innovation and the recognition that imitation and refinement through imitation are both necessary to invention itself and the very lifeblood of a competitive economy.

The Florida statute conflicts with several principles of federal patent law. First, 35 U.S.C. 102(a) and (b) excludes from consideration knowledge already available to the public. Since the boat hulls in question had been sold for six years, they were indisputably already available to the public. Second, along with requirements of novelty and utility, there is a nonobviousness requirement that “extends the field of unpatentable material beyond that which is known to the public under 102, to include that which could readily be deduced from publicly available material by a person of ordinary skill in the pertinent field of endeavor.”30 Bonito Boats emphasizes the Supreme Court’s unwillingness to allow nonpatent ­protection for published functionality: Congress has considered extending various forms of limited protection to industrial design either through the copyright laws or by relaxing the restrictions on the availability of design ­patents.31 Congress explicitly refused to take this step in the copyright laws32 and despite sustained criticism for a number of years, it has declined to alter the patent protections presently available for industrial design.33

Under Bonito Boats, all published unpatented functionality is in the public domain and is available to all for use in future works. Competitive principles require this result. Such competitive principles aid further development and innovation. But does the same hold true when the goodwill is captured in the trade dress elements captured in the design? Put differently, would another result obtain if the boat hulls in Bonito Boats had “pinch” bottle-type goodwill recognition to boating enthusiasts, who, like the whisky drinkers, saw the hulls differently?

30   Id at 134, (citing Graham v. John Deere Co. of Kansas City, 383 U.S. 1 at 15, 15 L. Ed. 2d 545, 86 S. Ct. 684 (1966)). 31   See generally Brown, Design Protection: An Overview, 34 UCLA Law Review 1341 (1987). 32   See 17 USC 101 [17 USCS 101]; HR Rep No 94-1476, p. 55 (1976). 33   Bonito Boats, Inc. v. Thunder Craft Boats, Inc., 489 U.S. at 145 (1989). See also Report of the President’s Commission on the Patent System, S Doc No 5, 90th Cong, 1st Sess, 20–21 (1967); Thomas B. Lindgren, The Sanctity of the Design Patent: Illusion or Reality? 10 Oklahoma City Law Review 195 (1985).

BARFIELD PRINT.indd 224

13/11/2018 11:52

Trade dress in virtual worlds  225 3.2  Feist Publications v. Rural Telephone Service In Feist Publications v. Rural Telephone Service Company, Inc.,34 the Supreme Court held that a publishing company that copied more than 1,000 name, address, and phone number listings from the white pages of a telephone book was not guilty of copyright infringement. The Court emphasized that the individual listings were not copyrightable facts and that the telephone company had not selected, coordinated, or arranged the facts in an original way sufficient to satisfy minimum de minimis copyright requirements. The lack of originality justified the failure of copyright; it did not matter that “sweat of the brow” effort was involved in creating the compilation, nor the precise (exact) copying of that many entries of names, addresses, and phone numbers. A brief summary of the case: Rural Telephone Service was the sole provider of telephone service in its service area. As a telephone company operating in Kansas, it was required by the State of Kansas to collect the telephone listing information and publish it in an annual directory. The white pages of Rural Telephone Service’s directory consisted of a simple alphabetical list of the names of Rural’s subscribers, their towns, and their telephone numbers. Feist Publications, a publishing company making areawide telephone directories, paid for the rights to use listings from telephone companies in 11 different service areas near Rural’s service area. Rural Telephone Service would not allow Feist to pay to use its listings. Because Rural would not allow Feist to license its listings, Feist used them without consent. Feist removed several thousand geographically irrelevant listings, then verified and attempted to obtain additional information on the 4,935 remaining listings for use in its directory. Most of Rural’s listings did not contain street addresses. Ultimately, most of Feist’s listings included the individuals’ names, street addresses, towns, and telephone numbers. However, 1,309 of the 46,878 listings in Feist’s directory were used (that is, precisely copied) directly from Rural Telephone Service’s white pages. Rural sued Feist for copyright infringement. The Supreme Court found that the telephone book as a whole was copyrightable but that the compilation of listings and the individual listings were not copyrightable. The Court’s opinion is useful in defining the scope of protection for works composed of facts and other public domain information. While the facts underlying a compilation cannot be protected by copyright, the author can claim protection of the manner in which the facts are presented. To qualify for copyright protection a work must be original to the author. To be original a work must: (1) be independently created by the author; (2) possess at least some minimal degree of creativity. When neither is present, copyright fails—even in the face of proof of direct copying. Importantly, the Feist Court defines the minimal “creativity” requirement in the negative as to what is not sufficient: “works in which the creative spark is utterly lacking or so trivial as to be virtually nonexistent”; which are “entirely typical,” “garden-variety,” “devoid of even the slightest trace of creativity,”35 “could not be more obvious,” and are “firmly rooted in tradition and so commonplace that it has come to be expected as a matter of course.”36

  Feist Publications, Inc. v. Rural Telephones Service, 499 U.S. at 340 (1991).   Feist Publications, Inc. v. Rural Telephones Service, 499 U.S. at 362 (1991). 36   Feist Publications, Inc. v. Rural Telephones Service, 499 U.S. at 359–62 (1991) (citing 34 35

BARFIELD PRINT.indd 225

13/11/2018 11:52

226  Research handbook on the law of virtual and augmented reality When such a compilation of facts and other public domain information is protected, the copyright is relatively “thin” and limited to the selection and arrangement of such information and only that. The Feist Court explained: This protection is subject to an important limitation. The mere fact that a work is copyrighted does not mean that every element of the work may be protected. Originality remains the sine qua non of copyright; accordingly, copyright protection may extend only to those components of a work that are original to the author . . . [T]he very same facts and ideas may be divorced from the context imposed by the author, and restated or reshuffled by second comers, even if the author was the first to discover the facts or to propose the ideas.37

Overall, the Feist Court makes clear that, at least in the copyright context, the reward is for originality, not effort, and “sweat of the brow” does not necessarily “promote the Progress of Science and useful Arts.”38 But this constitutional limit does not apply directly to trade dress.39 Will that matter, or will some such limitations be imposed eventually in trade dress cases under the Lanham Act? Under Feist, trivial user interfaces will receive only narrow protection; routine or commonplace approaches (such as simple menu selections) may receive no protection whatsoever. However, Feist reaffirms that original creative presentations of even public domain information can be protected by copyright against exact copying—but other Supreme Court cases suggest that even when copyright fails to extend, trade dress law can extend protection even where copyright law may refuse to do so. Copyright attaches not to individual facts but to the selection, organization, and arrangement of compilations of facts when they possess some degree of originality: “The compilation author typically chooses which facts to include, in what order to place them, and how to arrange the collected data so that they may be used effectively by readers.”40 The operative question becomes: What degree of originality? The Court’s answer was “at least some minimal degree of creativity.” How much creativity is “at least some minimal degree of creativity?” What the Court does not define is a clear positive test other than that the work must contain “some creative spark, no matter how crude, humble or obvious it might be,”41 “the work does not need to be novel, but cannot be copied,”42 and the work must contain “the fruits of intellectual labor.”43 Replace the undefined term “creative spark” with the similarly undefined term “inherently distinctive” and the essence of the coming trade dress controversy can be further Bleistein v. Donaldson Lithographing Co., 188 U.S. 239 (1903) (referring to “the narrowest and most obvious limits”)). 37   Id at 348, citing L. Ray Patterson & Craig Joyce, Monopolizing the Law: The Scope of Copyright Protection for Law Reports and Statutory Compilations, 36 UCLA Law Review 719, 800–802 (1989); J.C. Ginsburg, Creation and Commercial Value: Copyright Protection of Works of Information, 90 Columbia Law Review 1865, 1868, and n 12 (1990). 38   Id at 349, citing U.S. Constitution Art. I, 8, cl 8. 39   See footnote 1 supra. 40   Feist Publications, Inc. v. Rural Telephones Service, 499 U.S. at 370. 41   Id at 345, citing M. Nimmer & D. Nimmer, Copyright 1.08[C][1]. 42   Id at 346, citing Sheldon v. Metro-Goldwyn Pictures Corp., 81 F.2d 49, 54 (CA2 1936). 43   Id at 346, citing The Trade-Mark Cases, 100 U.S. 82 at 94 (1884).

BARFIELD PRINT.indd 226

13/11/2018 11:52

Trade dress in virtual worlds  227 understood: The array of different degrees of protection that may arise for different trade dress items in differing virtual worlds means that no “patent claim”-like clarity will ever exist for virtual-world developers wanting to avoid potential litigation minefields.44 Nor will that clarity exist for trade dress owners who want to extend their goodwill from the real-world product and service delivery into deliveries in the virtual world. Given how new the field is—particularly in the context of commercial and consumer transactions—one can expect, at best, a case by case development over an extended period, with likely differences between the federal district and circuit courts—particularly with respect to their interpretation of the leading Supreme Court case in the trade dress area, which we discuss next. 3.3  Two Pesos v. Taco Cabana Plaintiff Taco Cabana operated a chain of restaurants in Texas serving Mexican food; it used what it asserted was a unique style that it claimed was its own distinctive exterior (as well as interior) restaurant trade dress “equivalent to a federally registered trademark.” Figure 7.2 does far more justice. About seven years after the first Taco Cabana restaurant opened in San Antonio, a Two Pesos restaurant opened in Houston with a very similar motif. Within a year, Taco Cabana’s market had expanded such that the markets of the two chains of restaurants overlapped in several cities. Taco Cabana sued Two Pesos for trade dress infringement. In Two Pesos, Inc. v. Taco Cabana, Inc.,45 the Supreme Court addressed the following question in the context of trade dress: Whether the Mexican style appearance of a Mexican food restaurant constitutes protectable intellectual property equivalent to a federally registered trademark, even where the restaurant did not, and cannot, show that its Mexican “look” has secondary meaning in the eyes of its customers, so that members of the public associate the restaurant’s appearance with that particular restaurant?46

The Supreme Court answered the above question in the affirmative and held that the “look” of a restaurant, when inherently distinctive, and even though composed of functional items of furnishings and other leasehold improvements, is protected under the Lanham Act. More generally, when trade dress is inherently distinctive it does not have to acquire secondary meaning to be protected under the Lanham Act. This holding resolved a split between the Second Circuit, which subjected any trade dress protection to the requisite proof of secondary meaning, and the Fifth, Eighth, Eleventh, and other Circuits, which rejected that view and with which the High Court eventually agreed.47

44   Ronald J. Horta, Without Secondary Meaning, Do Product Design Trade Dress Protections Function as Infinite Patents? 27 Suffolk University Law Review 113, 114 (1993) (expansion of trade dress potentially creates “monopolies of unlimited duration to product design features and blunts the role of design patents in intellectual property law”). 45   Two Pesos, Inc. v. Taco Cabana, Inc., 505 U.S. 763 (1992). 46   Taco Cabana International Inc. v. Two Pesos, Inc., 932 F. 2d 1113 (5th Cir. 1991), cert granted, 112 S. Ct. 964 (1992). 47   Compare Laureyssens v. Idea Group, Inc., 964 F.2d 131 (2d Cir. 1992) (both secondary meaning and consumer confusion must be proven to establish trade dress infringement), with Chevron

BARFIELD PRINT.indd 227

13/11/2018 11:52

228  Research handbook on the law of virtual and augmented reality

Figure 7.2  Taco Cabana restaurant “To establish secondary meaning, a manufacturer must show that, in the minds of the public, the primary significance of the product feature or term is to identify the source of the product rather than the product itself.”48 This is not required when trade dress is inherently distinctive. What, then, is the meaning of “inherently distinctive”? The Supreme Court described the inherently distinctive Mexican trade dress as: A festive eating atmosphere having interior dining and patio areas decorated with artifacts, bright colors, paintings and murals. The patio includes interior and exterior areas with the interior patio capable of being sealed off from the outside patio by overhead garage doors. The stepped exterior of the building is a festive and vivid color scheme using top border paint and neon stripes. Bright awnings and umbrellas continue the theme.49

Chemical Co. v. Voluntary Purchasing Groups, 659 F.2d 695 (5th Cir. 1981), cert denied, 457 U.S. 1126 (1982) (acquisition of secondary meaning was not necessary as trade dress at issue was nonfunctional and arbitrary); Stuart Hall Co., Inc. v. Ampad Corp., 51 F.3d 780 (8th Cir. 1995) (if fanciful and arbitrary, no proof of secondary meaning required); and AmBrit, Inc. v. Kraft, Inc., 812 F.2d 1531 (11th Cir. 1986) (protection extends to trade dress that is inherently distinctive or associated with a secondary meaning). 48   Taco Cabana, 505 U.S. 763, 120 L.Ed.2d at 622 n. 4 citing Inwood Laboratories, Inc. v. Ives Laboratories, Inc., 456 U.S. 844 (1982). 49   Id at 621, citing Taco Cabana, 932 F.2d at 1117 (5th Cir. 1991).

BARFIELD PRINT.indd 228

13/11/2018 11:52

Trade dress in virtual worlds  229 The Supreme Court in this respect effectively lowered the threshold, cleared the path ahead, and in a way reserved, if not increased, the likelihood of expanding trade dress protection to the realm of virtual worlds in the near future. Through that expansion, the High Court also paved the way for the coming question of just how far trade dress protection will extend in virtual worlds. “The ‘trade dress’ of a product is essentially its total image and overall appearance.”50 It “involves the total image of a product and may include features such as size, shape, color or color combinations, texture, graphics, or even particular sales techniques.”51 The Taco Cabana holding (and the many cases it has spawned) will be employed by virtual-world authors and their publishers to obtain additional protection for the “look and feel” of virtual worlds. But existing trade dress owners will use Taco Cabana to question whether any use of their protected trade dress in such virtual worlds is actionable as well. Real-world situations will spawn virtual-world simulations; augmented and mixed reality applications will leverage both. In each case, protection will arise when even functional features are presented in an inherently distinctive way, or when they later acquire secondary meaning. Such features, unless purely functional and required to assure fair competition, are subject to trade dress protection under the Lanham Act upon proof of a likelihood of confusion. Such protection extends without proof of access and without proof of copying (though both may be present, as they were in the Taco Cabana case). Will the federal courts apply trade dress principles consistently in virtual-world cases as they do in real-world cases? In other words, will potentially broader judicial “license” be granted to the former over the latter, or will the same or similar tests be applied for “functionality” as the courts have applied in determining whether unprotected “ideas” trump protectable “expression” under Section 102(b) of the Copyright Act? Put differently, will goodwill be protected when visual and other associations are made by virtual-world creators for commercial, quasicommercial or even completely noncommercial purposes? To date, the federal courts have not addressed these questions and commentary has provided no roadmap.52 This is true even as the Supreme Court further expanded goodwill, more than 20 years ago, into just color alone, as we discuss next. Will color alone—this time in a virtual world—be protected by trade dress? 3.4  Qualitex Co. v. Jacobson Products Co., Inc. Three years after Taco Cabana, the Supreme Court decided Qualitex Co. v. Jacobson Products Co.,53 expressly extending trade dress protection to color alone. There, the plaintiff, a dry-cleaning supply business, used a distinctive green-gold color on the dry-cleaning press pads it sold to dry cleaners. A simple photo from the Court’s public domain record of the pads is shown in Figure 7.3. Jacobson Products, a direct rival, began selling pads of a similar color. Qualitex then   Taco Cabana at 621, n.1, citing Taco Cabana, 932 F.2d at 1118 (5th Cir. 1991).   Id at n 1, citing John H. Harland Co. v. Clarke Checks, Inc., 711 F2d 966, 980 (11th Cir. 1983); Restatement (Third) of Unfair Competition 16, Comment a (Tent. Draft No. 2, Mar. 23, 1990). 52   Sean Livesey, Intellectual Property Rights in Virtual Worlds, Richmond Journal of Law & Technology, http://jolt.richmond.edu/2017/02/13/intellectual-property-rights-in-virtual-worlds/. 53   514 U.S. 159 (1995). 50 51

BARFIELD PRINT.indd 229

13/11/2018 11:52

230  Research handbook on the law of virtual and augmented reality

Figure 7.3  Qualitex’s “green-gold” dry-cleaning pad registered its green-gold color with the Patent and Trademark Office and brought this trademark infringement suit against Jacobson, challenging its use of the green-gold color. The Ninth Circuit ruled in favor of Jacobson, holding that color alone is insufficient for Lanham Act purposes because “the better rule is that a trademark should not be registered for color alone . . . [since] under the color depletion theory, no person should have a monopoly on a primary color. We recognize that there are countless shades of colors that could not be depleted, but then, we could well become involved in ‘shade confusion.’” 54 The Supreme Court reversed, holding that “The Courts of Appeals have differed as to whether or not the law recognizes the use of color alone as a trademark55. . . We now hold that there is no rule absolutely barring the use of color alone, and we reverse the judgment of the Ninth Circuit.”56 In explaining the limits to its holding, the High Court went on to advise federal judges (and presumably juries on contested facts) would be able to distinguish color as a repository of protected goodwill from color as a functional element because they will be able to “examine whether its use as a mark would permit one competitor (or a group) to interfere with legitimate (non-trademark-related) competition through actual or potential exclusive use of an important product ingredient.”57 Thus, color alone may suffice. Will judges (and juries), in fact, be capable of evaluating the use of color and other trade dress elements in virtual worlds? Will they be able to determine whether that use (even with direct copying) is consistent with competitive and First Amendment principles?58 How will that be accomplished?59 Can it be accomplished consistently and in ways that   Qualitex Co. v. Jacobson Products Co., Inc., 13 F. 3d 1297 (9th Cir. 1994).  Compare NutraSweet Co. v. Stadt Corp., 917 F. 2d 1024, 1028 (CA7 1990) (absolute prohibition against protection of color alone) with In re Owens-Corning Fiberglas Corp., 774 F. 2d 1116, 1128 (CA Fed. 1985) (allowing registration of color pink for fiberglass insulation) and  Master Distributors, Inc. v. Pako Corp., 986 F. 2d 219, 224 (CA8 1993) (declining to establish per se prohibition against protecting color alone as a trademark). 56   Qualitex Co. v. Jacobson Products Co., Inc., 514 U.S. 159, 161–2 (1995). 57   Qualitex Co. v. Jacobson Products Co., Inc., 514 U.S. at 163–4 (1995). 58   In the patent context, the analogous question of patent claim interpretation is handled pretrial by the Court in the Markman “claim interpretation” phase of the litigation. Markman v. Westview Instruments, Inc., 517 U.S. 370 (1996). Will we see that type of approach used eventually in federal trade dress cases? 59   See, e.g., Jeffrey M. Samuels & Linda B. Samuels, Protection for Color under U.S. Trademark Law, 23 AIPLA Quarterly Journal 129 (Winter 1995) (indepth overview of registrability of color and color alone under the US trademark law); Hillel I. Parness, Curse of the Pink Panther: The Legacy of the Owens-Corning Fiberglas Dissent and Its Role in the Qualitex 54 55

BARFIELD PRINT.indd 230

13/11/2018 11:52

Trade dress in virtual worlds  231 allow lawyers to counsel their clients accordingly, and can that happen given the array of Supreme Court cases that have both broadened and retracted trade dress protection in competitive litigation cases where functionality exists, as we discuss next? 3.5  TrafFix Devices v. Marketing Displays In TrafFix v. Marketing Displays,60 after finding presence of functionality in the features at issue, the High Court rejected the springloaded traffic sign designer’s contention that the dual-spring design had itself become protectable trade dress embodying the goodwill of the business (see Figure 7.4). Comparing the dual-spring features to the colorful combinations presented in Taco Cabana, the Court noted that “the trade dress in those cases did not bar competitors from copying functional product design features. In the instant case, beyond serving the purpose of informing consumers that the sign stands are made by MDI (assuming it does so), the dual-spring design provides a unique and useful mechanism to resist the force of the wind.”61 The unanimous High Court explained: The asserted trade dress consists simply of the dual-spring design, four legs, a base, an upright, and a sign. MDI has pointed to nothing arbitrary about the components of its device or the way they are assembled. The Lanham Act does not exist to reward manufacturers for their innovation in creating a particular device; that is the purpose of the patent law and its period of exclusivity.

Figure 7.4  Wind-resistant springloaded display stands

Supreme Court Appeal, 18 Columbia-VLA Journal of Law & the Arts 327 (1993–4) (analysis of US trademark jurisprudence since more than a century before Qualitex); Paul R. Morico, Protecting Color Per Se in the Wake of Qualitex v. Jacobson, 77 Journal of the Patent and Trademark Office Society 571 (July 1995) (discussion of trademark registrations based on color alone post-Qualitex). 60   TrafFix Devices, Inc. v. Mktg. Displays, Inc., 532 U.S. 23 (2001). 61   TrafFix Devices, Inc. v. Mktg. Displays, Inc., 532 U.S. at 33.

BARFIELD PRINT.indd 231

13/11/2018 11:52

232  Research handbook on the law of virtual and augmented reality The Lanham Act, furthermore, does not protect trade dress in a functional design simply because an investment has been made to encourage the public to associate a particular functional feature with a single manufacturer or seller. The Court of Appeals erred in viewing MDI as possessing the right to exclude competitors from using a design identical to MDI’s and to require those competitors to adopt a different design simply to avoid copying it. MDI cannot gain the exclusive right to produce sign stands using the dual-spring design by asserting that consumers associate it with the look of the invention itself.62

Thus, when virtual-world authors promote and “encourage the public to associate a particular functional feature” with their product or service, they may not achieve any trade dress protection at all; when such functional features become highly popular then, absent patent protection, direct copying is permitted even if actual confusion occurs and even if the first mover’s investment is otherwise being “misappropriated” by competitors. This is a restatement of the original principles set forth in the Bonito Boats case discussed previously. The question then becomes: What is the meaning of “a particular functional feature”? Many thought that question would be decided broadly against intellectual property owners and in favor of competitive principles post-TrafFix but, as discussed next, the opposite occurred last year in the context of what were argued to be functional geometric designs in school uniforms. Though presented in the context of various “original” works registered with the Copyright Office, the case helps explain the extent to which the Court may well delegate factual decisionmaking to some degree to federal administrative agencies. Put differently, registration of trade dress may become crucial; perhaps this is best demonstrated by the role played by the more than 200 copyright registrations (of various geometric designs) in the case we discuss next. First movers who register their trade dress early (and often) may garner more and greater protection for their virtual-world creations than those who fail to do so. 3.6  Star Athletica v. Varsity Brands Though not a trade dress case, in Star Athletica, L. L. C. v. Varsity Brands, Inc.63 the Supreme Court was faced with another application of the above principles, but in the context of the Copyright Act’s bar on functional works, in deciding whether a basic geometric uniform design can be protected as “original” when the design itself is part of a functional article (clothing), as shown in Figure 7.5. The divided High Court upheld protection: We hold that an artistic feature of the design of a useful article is eligible for copyright protection if the feature (1) can be perceived as a two- or three-dimensional work of art separate from the useful article and (2) would qualify as a protectable pictorial, graphic, or sculptural work either on its own or in some other medium if imagined separately from the useful article. Because the designs on the surface of respondents’ cheerleading uniforms in this case satisfy these requirements, the judgment of the Court of Appeals is affirmed.64

  TrafFix Devices, Inc. v. Marketing Displays, Inc., 532 U.S. 23 at 35 (2001) (emphasis added).   Star Athletica, L.L.C. v. Varsity Brands, Inc., 137 S. Ct. 1002 (2017). 64   Star Athletica, L.L.C. v. Varsity Brands, Inc., 137 S. Ct. 1002, 1016 (2017). 62 63

BARFIELD PRINT.indd 232

13/11/2018 11:52

Trade dress in virtual worlds  233

Figure 7.5  Geometric uniform designs The dissent disagreed sharply with a view that the shapes were all common geometric elements of no originality and simply part of the utilitarian object (here, cheerleader uniforms): Were I to accept the majority’s invitation to “imaginatively remov[e]” the chevrons and stripes as they are arranged on the neckline, waistline, sleeves, and skirt of each uniform, and apply them on a “painter’s canvas” . . . that painting would be of a cheerleader’s dress. The esthetic elements on which Varsity seeks protection exist only as part of the uniform design—there is nothing to separate out but for dress-shaped lines that replicate the cut and style of the uniforms. Hence, each design is not physically separate, nor is it conceptually separate, from the useful article it depicts, namely, a cheerleader’s dress. They cannot be copyrighted.65

A concurring opinion suggests that the Copyright Office’s decision to issue registration certificates was case determinative. But one can ask: When even the High Court has sharp disagreement about the scope of protection for works registered as protectable works of original expression, and not just geometric cheerleader uniforms, then how will federal judges and juries reach agreement in cases where utility and functionality conflict with goodwill and advertising investment? And what is to be made of a large, well-known competitor adopting (and arguably thereby potentially depleting) as many of the geometric designs as possible using multiple (more than 200) copyright registrations for a multitude of different “popular” geometric designs? By analogy, in virtual worlds (as well as the real world), will this decision drive trade dress owners to be motivated then to file multiple trademark registrations on their trade dress with a view toward bringing themselves within the line of reasoning of the Supreme Court in Star Athletica? What limits will exist to constrain that effort? Will any limits be imposed to such registration efforts? The next case considered holds that traditional statutory limits may not suffice and federal courts will be empowered to make broader discretionary judgments about the scope of protection permitted, as well as the scope of what will be allowed registration under the Lanham Act.

  Id at 1035.

65

BARFIELD PRINT.indd 233

13/11/2018 11:52

234  Research handbook on the law of virtual and augmented reality 3.7  Matal v. Tam (the “Slants” Trademark Case) In Matal v. Tam,66 a First Amendment case of first impression, the High Court rejected the historical Trademark Office policy of refusing trademark registration of “offensive” marks because The Slants as a rock band’s trademark might be seen as demeaning to Asian Americans. The Patent and Trademark Office did not (and could not) forbid the band from using the mark; it simply denied registration protections for the goodwill that the rock band had already acquired in the word mark. Even this, the High Court held, was viewpoint discrimination—including against allegedly racially offensive viewpoints—and was unconstitutional; Justice Alito (for four justices, as Justice Gorsuch’s appointment was not yet ratified) wrote: The idea that the government may restrict speech expressing ideas that offend . . . strikes at the heart of the First Amendment. Speech that demeans on the basis of race, ethnicity, gender, religion, age, disability, or any other similar ground is hateful; but the proudest boast of our free speech jurisprudence is that we protect the freedom to express “the thought that we hate.”

Justice Kennedy wrote separately, also for four justices, but on this point the opinions agreed because viewpoint discrimination is an “egregious form of content discrimination,” which is presumptively unconstitutional . . . A law that can be directed against speech found offensive to some portion of the public can be turned against minority and dissenting views to the detriment of all. The First Amendment does not entrust that power to the government’s benevolence. Instead, our reliance must be on the substantial safeguards of free and open discussion in a democratic society.

Importantly, the justices made clear that First Amendment speech—even in this “commercial” context and even where some view it as racially offensive—is protected not just against outright prohibition but also against lesser restrictions. In Matal, the First Amendment takes on new, robust power to assure maximum protection of even offensive trade dress content, but it also potentially provides the seeds for additional defenses in a host of litigation contexts generally. More important to the trade dress developments explained in the foregoing, and more specifically: Will this new First Amendment development give new freedoms to those who seek to leverage “immediate associational value” that developers seek in virtual worlds for communication purposes, or will case law developments expanding trade dress for the protection of associational goodwill continue that expansion even in the face of First Amendment arguments? We turn to that question next.

4. CHALLENGES TO EXTENSION OF TRADE DRESS PROTECTION What becomes of trade dress protection as virtual worlds come closer and closer to real-world experiences? What happens when that combines to allow shopping and other 66

  Matal v. Tam, 137 S. Ct. 1744 (2017).

BARFIELD PRINT.indd 234

13/11/2018 11:52

Trade dress in virtual worlds  235 consumer activities to use virtual worlds, including augmented and mixed reality, to enable real-world transactions? Developers of pioneering virtual-world works will seize upon the ability to generate immediate visual associations. The pink sweetener differs from the blue sweetener which also differs from the yellow sweetener, and all of them are distinct from brown sugar and from white sugar, but the trade dress of each conveys meaning as well. Virtual worlds depend on conveying immediate impressions. Will this alone provide greater First Amendment reasons for barring broad trade dress protection or otherwise placing substantial limits on infringement claims? Of course, competitors will continue to seek to expand (and exploit) all published unpatented functionality, including methods of operation, user-learned techniques, and common compatibility. Each new “reality”—indeed, with these mobile works, each new version—presents a new set of facts upon which questions of scope of legal protection and extent of infringement must be examined and reexamined. Where will the “safety valve” to repeated expensive and time-consuming litigation arise and how will that be best implemented by the courts and by industry? Under “fair use” principles—particularly when combined with fee shifting available in federal court litigation—a defendant can impose a fair use defense to the claimed infringement.67 Other defenses too arise from special immunities that Congress has provided to those who follow “take down” procedures that provide complete bars to litigation. These procedures may provide some ability, directly or indirectly by analogy, for the court to screen cases that do not merit judicial attention. Whether that same approach is adopted in virtualworld infringement litigation generally, and in the trade dress cases specifically, remains to be seen. In the near future, counsel will be called upon to provide early opinions about what can and what cannot be expanded upon and what can or cannot exploited in virtual worlds. But one of the largest challenges to the extension of trade dress law will rest on the differing (and overlapping nature of) potential protections offered respectively by case law development seeking to broaden trade dress rights under the Lanham Act as against challenges seeking a broadening role of the First Amendment, as well as statutory fair use available under the Lanham Act. How far will courts go to harmonize policies underlying the latter with the expanding body of trade dress case law seeking to broadly protect an owner’s goodwill? Whether First Amendment values prevail will likely depend on “fair use” factors that may well be quite different in virtual worlds and the context in which the case is presented to the courts. The issue was raised (though not in the United States) in a case of first impression that did not result in litigation, but the relevant facts were present. In the summer of 2015, the famous street artist Banksy produced a real (not virtual, but it could well have been) theme park, which he named “Dismaland” as a parody of the world famous “Disneyland” theme park (Figure 7.6). Videos of it exist and are still available publicly on YouTube,68 as well as on Facebook,69 but the physical world attraction itself, which was in place for several weeks in the United Kingdom, is now closed.

67   Ellii Cho,  Copyright or Trade Dress? Toward IP Protection of Multisensory Effect Designs for Immersive Virtual Environments, Cardozo Arts and Entertainment Journal 823 (2016). 68  www.youtube.com/watch?v=V2NG-MgHqEk. 69  www.facebook.com/dismalandfanpage/.

BARFIELD PRINT.indd 235

13/11/2018 11:52

236  Research handbook on the law of virtual and augmented reality

Figure 7.6  “Dismaland” by Banksy Banksy’s brochure described the event—which now exists “virtually” in many photos and videos, but not yet in any virtual world—as “a chaotic new world where you can escape from mindless escapism . . . This event contains adult themes, distressing imagery, extended use of strobe lighting, smoke effects and swearing. The following items are strictly prohibited: knives, spraycans, illegal drugs, and lawyers from the Walt Disney corporation.” If Banksy had decided (or if he eventually decides) to transform some of the available media into a virtual world available for anyone to “experience” through use of modern VR equipment, will the Disney Corporation have good cause to assert its trade dress (and other intellectual property rights) against this? And if so, will trade dress protection win out over fair use and First Amendment defenses? These questions are not purely academic: Trade dress owners can imagine loss of substantial goodwill in the use (and mockery) of their trade dress (and other intellectual property rights) and the failure to police restrictions on such usage may be viewed as an implied form of consent to the continuing (and perhaps even to more expansive) invasion of those rights. When this type of case is eventually litigated, will courts differ on whether First Amendment freedoms—even in commercial contexts—prevail over well-established trade dress and perhaps even famous trademarks included in that trade dress? Even more broadly, could Banksy register his own trademark rights (in “Dismaland”) and his rights in the associated trade dress? This too is not an academic or hypothetical question, as the Banksy organization did precisely that in the United Kingdom.70 Would the same approach (as was allowed at least for several weeks in the UK) be permitted in the US? A very recent Federal Circuit decision suggests that harmonization will weigh in 70   See footnote 15 supra. One can imagine an eventual AR application available while physically in a Disneyland theme park that compares and contrasts the Dismaland virtual world with the associated real-world theme park that is the subject of the critical commentary. Will those

BARFIELD PRINT.indd 236

13/11/2018 11:52

Trade dress in virtual worlds  237

Figure 7.7  “FUCT” t-shirt designs favor of First Amendment principles: In re Bruneti, _ F.3d _ (Fed. Cir. 2017). Applying post-Tam First Amendment principles, the Federal Circuit in In re Bruneti held that the Lanham Act’s bar against registering immoral or scandalous marks is an unconstitutional restriction of free speech and that Bruneti’s “FUCT” trademark, an image of which can be seen in Figure 7.7, should be allowed registration. Thus, on appeal the court reversed the Trademark Trial and Appeal Board’s holding that “FUCT” is unregistrable for clothing. The Federal Circuit expands here on the Supreme Court’s recent decision in Matal v. Tam: Just as the Lanham Act’s Section 2(a)’s bar on the registration of  disparaging marks  involved governmental viewpoint discrimination in violation of free speech principles, the Federal Circuit held that this principle equally extends to “immoral” or “scandalous” marks as well. The registration of such marks— otherwise barred by the express terms of the Lanham Act—means that even “immoral” and “scandalous” trade dress will also be allowed protection and registration. Will the same apply as well to immoral trade dress embedded in virtual worlds? Will that apply even where the trade dress derives from well-established trade dress owned by traditional real-world owners who have made (and continue to make) substantial brand marketing, advertising, and other investments in causing the public association of the business brand with the inherently distinctive trade dress they developed perhaps over decades? The First Amendment’s prohibitions as announced in Tam and extended now by Bruneti apply to “even regulation of purely commercial speech.” The majority opinion finds First Amendment content discrimination; a concurring opinion notes that a portion of the Lanham Act’s Section 2(a) could and should be limited to bar obscenity, based on the approach of federal criminal law that bars mailing of “obscene, lewd, lascivious, indecent, filthy or vile” material. The majority holds that it is for Congress, not the courts, to revise Section 2(a) and to bring it squarely within the First Amendment. In the meantime, courts will be forced to consider how far these same First Amendment principles provide defenses in virtual-world trade dress litigation. facts change the applicable result, or would they further support the First Amendment expression intended by Banksy?

BARFIELD PRINT.indd 237

13/11/2018 11:52

238  Research handbook on the law of virtual and augmented reality Harmonization may be found in the Ninth Circuit’s approach in recent videogame litigation. In E.S.S. Entm’t 2000, Inc. v. Rock Star Videos, Inc.,71 the Ninth Circuit affirmed that Rock Star’s depiction of the plaintiff’s real-life Los Angeles strip club in the video game Grand Theft Auto: San Andreas (which included both the exterior building design as well as the plaintiff club’s actual logo) did not infringe the plaintiff’s trademark or its trade dress rights because “[n]othing indicates that the buying public would reasonably have believed that [the club owner] produced the video game or, for that matter, that [the video company] operated a strip club.”72 The Ninth Circuit came to this conclusion even after rejecting Rock Star’s fair use defense and instead applied the Second Circuit’s approach from Rogers v. Grimaldi,73 which limits Lanham Act scope “to apply to artistic works only where the public interest in avoiding consumer confusion outweighs the public interest in free expression.”74 Under this approach, Banksy-type artists will have broad ability to argue in favor of free expression trumping even actual consumer confusion in virtual-world trade dress litigation.75 Recognizing and applying fair use defenses means virtual-world trade dress infringement cases will need case by case adjudication, not unlike what was seen at the inception of computer software copyright litigation more than four decades ago.76 In this context, very recent parallel developments that have occurred in the screening of software (and other) patents through administrative procedures in the PTO and the newly established Patent Trial and Appeal Board (PTAB) may well play a role, at least by analogy, in the manner in which the federal courts will seek to screen—at earlier and earlier stages of litigation—noninfringement defenses based on pure functionality, the First Amendment, fair use, and other defenses. We conclude that embedded in the above very recent case law developments is an implicit recognition that First Amendment values are broadening and are increasingly driving respect for defenses to any “infringement” on goodwill that might otherwise have been, and might still be, protected in nonrich media environments. Put differently, the result achieved in the “Pinch” bottle’s garnering of protection and even registration more than 60 years ago may not be achieved in an environment where virtual-world immersions and interactions elect to use depictions that artistically cause immediate sensory connections, just as Banksy did in his “Dismaland” event. Even already registered (and otherwise incontestable) trade dress otherwise historically afforded broad Lanham Act protection, including iconic theme parks and even famous   E.S.S. Entm’t 2000, Inc. v. Rock Star Videos, Inc., 547 F.3d 1095 (9th Cir. 2008).   Id at 1100. 73   Rogers v. Grimaldi, 875 F.2d 994 (2d Cir. 1989). The Second Circuit’s test asks: (1) is the trademark use relevant to the defendant’s work from an artistic standpoint; (2) is the use explicitly misleading? 74   Id at 999. 75  See KP Permanent Makeup v. Lasting Impressions, 543 U.S. 111, 121–2 (2004) (“Since the burden of proving likelihood of confusion rests with the plaintiff, and the fair use defendant has no free-standing need to show confusion unlikely, it follows (contrary to the Court of Appeal’s view) that some possibility of consumer confusion must be compatible with fair use, and so it is”). 76   See Software Copyright Act of 1980 (amending Copyright Revision Act of 1976 based on CONTU Report recommendations); see also Mickey T. Mihm, Software Piracy and the Personal Computer: Is the 1980 Software Copyright Act Effective? 4 Computer Law Journal 171 (1983). 71

72

BARFIELD PRINT.indd 238

13/11/2018 11:52

Trade dress in virtual worlds  239 bottles,77 may not clear the First Amendment challenges that may be imposed as defenses in the virtual-world context. Whether this will include potentially allowing a virtual comparative theme park immersion or using the “Pinch” approach as a common means of conveying the notion of reaching for a specialty whisky in the virtual world is an open question. Post-Tam expansive First Amendment principles may trump whatever dilution or other diminishment of goodwill occurs—even when actual confusion occurs in the communication of the sensory association. More generally, all trade dress owners are subject to the same fair use and First Amendment defenses and virtual-world authors may choose to test the limits of these defenses. There is, of course, room for coexistence of strong First Amendment values as well as strong protection of legitimate goodwill by trade dress protections in virtual worlds. How the courts balance these rights and the due process needs of litigants with the fairness requirements of allowing defendants the earliest consideration of their defenses will be a further matter of case by case development that will unfold in the next wave of litigation, which inevitably will arise from virtual-world expansion and intersection with the real world of commercial transactions. Extension of trade dress protection under the Lanham Act in virtual worlds will be balanced against First Amendment principles and inevitably the Supreme Court will need to state what priorities will best assure that fairness, effectiveness, and efficiencies are promoted throughout the federal courts.

5. CONCLUSION Though the struggle to define the appropriate level of trade dress protection in virtual worlds and other graphical user interfaces will continue to challenge clients, counsel, and the courts, we submit that Lanham Act trade dress protection for virtual worlds, including virtual reality, augmented reality, and mixed reality applications, will remain available for protection of the inherent as well as acquired distinctiveness of nonfunctional human sensory interfaces expressed by software designers and programmers in their new virtualworld products and services in a broad range of coming applications in many diverse fields. Ultimately, goodwill-generating virtual works will obtain protection by federal law against those who “cross the line,” but how to discern where that precise boundary lies in any given case will require an evolution of factual and legal experts and further case law developments—not unlike those occurring post-Feist (in copyright), post-Alice (in patent), post-Qualitex (in trademark), and post-Tam (in First Amendment litigation). These developments, and particularly the latter, will bear directly on the scope of trade dress law developments in virtual worlds. In the meantime, practitioners must become familiar with a growing body of case law and the manner in which overlapping protection is granted (or denied) on particular facts presented in particular cases.

77   Whether “famous” trade dress (such as the historical classic Coca-Cola bottle) may receive even broader protection under the Federal Trademark Dilution Act of 1995 is an open question; the Act defines dilution as “the lessening of the capacity of a famous mark to identify and distinguish goods or services, regardless of the presence or absence of (1) competition between the owner of the famous mark and other parties, or (2) likelihood of confusion, mistake, or deception.”

BARFIELD PRINT.indd 239

13/11/2018 11:52

240  Research handbook on the law of virtual and augmented reality

SOURCES Cases Alice Corp. Pty. Ltd. v. CLS Bank Intern., 134 S. Ct. 2347 (2014) AmBrit, Inc. v. Kraft, Inc., 812 F.2d 1531 (11th Cir. 1986) Apple Computer, Inc. v. Microsoft Corporation, 799 F. Supp. 1006, 1022–1026 (N.D. Cal. 1992) Bleistein v. Donaldson Lithographing Co., 188 U.S. 239 (1903) Bonito Boats, Inc. v. Thunder Craft Boats, Inc., 489 U.S. 141 (1989) Chevron Chemical Co. v. Voluntary Purchasing Groups, 659 F.2d 695 (5th Cir. 1981) E.S.S, ENTER’T 2000 v. Rock Star, 547 F.3d 1095 (9th Cir. 2008) Ex parte The Boye Needle Company, 100 U.S.P.Q. 124 (Com’r., 1953) Ex parte Minnesota Mining and Manufacturing Co., 92 U.S.P.Q. 74 (Com’r., 1952) Feist Publications v. Rural Telephone Service Company, Inc., 499 U.S. 340 (1991) In re General Mills IP Holdings II, LLC, 124 USPQ2d 1016 (TTAN 2017) Graham v. John Deere Co. of Kansas City, 383 U.S. 1 at 15, 15 L. Ed. 2d 545, 86 S. Ct. 684 (1966)) Inwood Laboratories, Inc. v. Ives Laboratories, Inc., 456 U.S. 844 (1982) KP Permanent Makeup v. Lasting Impressions, 543 U.S. 111 (2004) Laureyssens v. Idea Group, Inc., 964 F.2d 131 (2d Cir. 1992) Lotus Development Corp. v. Borland International, Inc., 799 F. Supp. 203, 216–19 (D. Mass. 1992) Markman v. Westview Instruments, Inc., 517 U.S. 370 (1996) Matal v. Tam, 582 U.S. ___ (2017) Moroccanoil Israel Ltd v. Aldi Stores Ltd [2014] EWHC 1686 (29 May 2014) Qualitex Co. v. Jacobson Products Co., Inc., 13 F. 3d 1297 (9th Cir. 1994) Qualitex Co. v. Jacobson Products Co., Inc., 514 U.S. 159 (1995) Reckitt & Colman Ltd v. Borden, Inc., 1 All ER 873 (1990) Rogers v. Grimaldi, 875 F.2d 994 (2nd Cir. 1989) Sony Corp. of Am. v. Universal City Studios, Inc., 464 U.S. 417 (1984) Stuart Hall Co., Inc. v. Ampad Corp., 51 F.3d 780 (8th Cir. 1995) Taco Cabana International, Inc. v. Two Pesos, Inc., 932 F. 2d 1113 (5th Cir. 1991) TrafFix Devices, Inc. v. Mktg. Displays, Inc., 532 U.S. 23, 33 (2001) Two Pesos, Inc. v. Taco Cabana, Inc., 505 U.S. 763 (1992) Volkswagenwerk Aktiengesellschaft v. Church, 411 F.2d 350 (9th Cir.), supplemented, 413 F.2d 1126 (9th Cir. 1969) White v. Pauly, 580 U.S. ___ (2017)

Statutes Defend Trade Secrets Act of 2016 (Pub.L. 114–53, 130 Stat. 376, enacted May 11, 2016, now 18 U.S.C. § 1836, et seq Model Trademark Bill (MTB) Software Copyright Act of 1980 (amending Copyright Revision Act of 1976 based on CONTU Report recommendations) Trademark Act of 1946, ch. 540, 60 Stat. 427 (codified as amended at 15 U.S.C. §§ 1051–127 (1994)) Uniform Deceptive Trade Practices Act (UDTPA) U.S. Constitution art. I, 8, cl. 8 U.S.C. § 1127 (1994) (United States Code) 17 U.S.C. 108 (Limitations on exclusive rights: reproduction by libraries and archives)

BARFIELD PRINT.indd 240

13/11/2018 11:52

PART III ISSUES OF CONSTITUTIONAL AND CRIMINAL LAW

BARFIELD PRINT.indd 241

13/11/2018 11:52

8.  The First Amendment, video games, and virtual reality training Marc Jonathan Blitz

1. INTRODUCTION One of the great benefits of virtual reality is that it is, in certain respects, unreal. In virtual reality (“VR”), the physical environment in which one seems to be immersed is not really there. Consider the use of VR to fly a virtual plane through turbulent air. The plane that a VR user is piloting is a computer-generated illusion, produced within a head-mounted display that rests over her eyes. So too is the rough weather pushing against the illusory plane and challenging the pilot’s skills. A vivid simulation of a crash or engine failure may be unsettling for the pilot, even when she knows that it is only a simulation.1 But its consequence is simply that the pilot may have to reset the experience and learn how to better confront the challenge in the next simulation. This aspect of VR allows it to serve as a laboratory of sorts for individuals to practice all types of conduct before they engage in its real-life equivalent. Soldiers can train for battle in circumstances where their lives aren’t at stake. Surgeons can conduct virtual surgeries in order to prepare themselves for operating on real patients, where mistakes would have deadly consequences. Psychotherapy patients can confront the objects of their phobias—whether these objects are spiders, confined places, or heights—under circumstances where those objects are virtual rather than physically real. In short, an individual can simulate a variety of day-to-day activities—driving vehicles, engaging in athletic maneuvers, facing audiences, or performing challenging occupational tasks—in ways which do not actually put their safety and financial security at risk. They can also use VR to do repeatedly, in a virtual setting, what they may only have one chance to do correctly in the physical world—and thus hone their skills or routines to a greater extent than would be the case without the simulation. While this rehearsal for real-world action has great benefits, it can also be a component of harmful behavior. First, even though VR is an illusion, it may cause physical harm when the simulation is inaccurate in important respects: When real-world dangers facing a pilot, police officer, or surgeon are not accurately captured by a VR simulation, the person learning from that simulation may be ill-prepared to face these dangers in real-world challenges—or she might develop habits that, while unproblematic in the VR setting, are unhelpful or even dangerous outside of it.2 And even where VR simulation is accurate

1   Many users of virtual reality have reported feeling more intense fear in simulations of danger than they would if they simply saw a movie showing the same danger. One journalist, for example, speaks of the fear she felt with the “wind blowing against my face as I faced the world from the height of a skyscraper” in a simulation of a tightrope walk between World Trade Center towers. Laura Hudson, “It’s Not Real. It’s Not Real. It’s Not Real,” Slate, May 2, 2016. 2   See Shelly Wood, So You Think You’re an Interventional Cardiologist? Let the Simulators

242

BARFIELD PRINT.indd 242

13/11/2018 11:52

First Amendment, video games, and virtual reality training  243 and effective, it may cause harm when it is in the wrong hands. A virtual rehearsal session might, for example, be a trial run for an act of crime or terrorism. Daniel Gerstein, former Undersecretary for Science Technology at the Department of Homeland Security and current RAND Corporation analyst, has warned that terrorists may wish to train with “virtual rehearsals of an attack”:3 They may use a “virtual display of a building” to practice an act of terrorism on that site, or seek out “lifelike training scenarios to prepare recruits for a suicide bombing attack or a kidnapping.”4 While serving as Director of National Intelligence, James Clapper issued a report stating that terrorists will likely take advantage of “virtual environments” as well as other emerging technologies.5 As VR use expands, it thus may potentially be an object of government regulation, and with such regulation comes an interesting constitutional question: To what extent are the virtual encounters we use to hone skills, or otherwise reshape mental functioning, experiences insulated by the US Constitution from government control? Various scholars have argued that our mental autonomy is an object of constitutional protection. The Constitution, Supreme Court Justices have said, protects our “freedom of thought”6 or “freedom of mind.”7 It secures for each individual a realm of “autonomy” in which to live and develop their lives free from government interference.8 This constitutionally protected zone of autonomy clearly includes the right to shape one’s beliefs—or, for that matter, the beliefs of others—with words or other forms of expression. The First Amendment right to freedom of speech gives individuals opportunities to obtain new knowledge, or begin to master a new skill, by reading a book about such a subject or watching a video about it. Individuals can also write books or create videos that help others learn new information, or master new skills. But does such constitutional protection for freedom of speech—or any other source of constitutional protection—also extend to use of virtual training environments, and perhaps also some physical training environments, that instill new habits or hone skills without any interpersonal communication? When a person practices maneuvers in a real plane, or practices firing a real machine gun, this is generally not First Amendment “speech.” Do the same flying maneuvers in a virtual plane, or practicing shooting with a virtual machine gun, count as protected speech? If not, is this constitutionally protected activity of some other sort? There is one argument in favor of First Amendment protection that it makes sense to consider before engaging in deeper analysis. One might say that to the extent that virtual training is simply a more practically oriented form of virtual gaming, then it clearly does count as constitutionally protected expression. In Brown v. Entertainment Merchants Association, the Supreme Court held that video games are protected speech: Decide, Medscape, May 19, 2005, which notes “some of the foremost experts in the field [of cardiology] are sounding alarms over just how simulators are being used, pointing out that some have the ability to teach bad habits as well as good.” 3   Eric Niler, VR Is a National Security Threat, Seeker, Feb. 12, 2016. 4   David Roos, Terrorists Could Use Virtual Reality to Prepare for Attacks, How Stuff Works, Jun. 6, 2016. 5   Id. 6   See Palko v. Connecticut, 302 U.S. 319, 326–27 (1937); United States v. Schwimmer, 279 U.S. 644, 654–5 (1929) (Holmes J dissenting). 7   See W. Va. State Bd. of Educ. v. Barnette, 319 U.S. 624, 637 (1943). 8   Stanley v. Georgia, 394 U.S. 557, 566 (1969).

BARFIELD PRINT.indd 243

13/11/2018 11:52

244  Research handbook on the law of virtual and augmented reality Like the protected books, plays, and movies that preceded them, video games communicate ideas—and even social messages—through many familiar literary devices (such as characters, dialogue, plot, and music) and through features distinctive to the medium (such as the player’s interaction with the virtual world). That suffices to confer First Amendment protection.9

Virtual reality is, one might claim, best known as an extension of the more familiar video game technology that has been used for decades: It allows us to immerse ourselves in the game—to generate the feeling that we are inside of it—instead of watching it unfold in the two-dimensional space of a monitor screen. After all, although VR has multiple uses (including training, tourism, scientific analysis, or previewing of real estate offerings or building designs), the use that is most familiar to modern day consumers is video game play. It is VR video consoles—such as Oculus Rift, HTC Vive, and PlayStation VR—that are most responsible for making virtual reality a part of modernday consumers’ lives. It was the hype surrounding such gaming technologies that led certain commentators in the past few years to predict that VR was about to move from laboratories and specialized training facilities into the “mainstream.”10 As one article making this claim emphasizes, “[v]ideo games seem to be an ideal medium for a technology that requires users to place a device on their heads and ‘live’ inside a 3D virtual world”—since video games typically invite players to act within an imaginary world, and VR technology lets them experience such a world more vividly.11 Of course, the same technology can also be used for the similar but more passive experience of watching a movie. Even before the rise of modern VR gaming, certain pioneers of VR tried to develop technology that would make movie audiences feel immersed inside of a film’s action: Fred Waller invented a curved movie screen in the 1930s;12 in the 1950s, Morton Heilig created a screen inside of a helmet that made movies feel even more immersive;13 and movie theaters have for decades now experimented with providing viewers with “3D glasses” that can make film action seem to jump off the screen. One might thus argue that the First Amendment question I posed about VR training is an easy one: First, the imaginary driving, flying, or sexual encounter one experiences with VR technology is First Amendment speech for the same reason that driving, flying, or other nonspeech conduct counts as expression when we see it in a movie plot, or when we steer video game vehicles or cause a video game avatar to have a sexual encounter. In each of these cases, we are watching, or playing a role in, a visually rich story. We are watching fictional characters drive, fly, or have sex in movies, or we are playing fictional characters who do such things in a video game. Second, even if we indulge in such fiction not simply for entertainment, but for a more practical purpose—such as gaining new knowledge, learning a new skill, or refining an existing skill—that does not make the video game activity any less expressive. People often play video games not simply to enjoy themselves, but to challenge themselves and get better at meeting the challenge. Thus, even if one’s goal in using a flight simulator game is not simply to enjoy an hour’s worth of gaming but   Brown v. Entm’t Merchants Ass’n, 564 U.S. 786, 790 (2011).   Don Reisinger, Why Virtual Reality Is About to Go Mainstream, Fortune, Oct. 7, 2015, at http://fortune.com/2015/10/07/virtual-reality-mainstream/. 11   Id. 12   Howard Rheingold, Virtual Reality: The Revolutionary Technology of Computer-Generated Artificial Worlds—and How It Promises to Transform Society 355 (First Touchstone ed. 1992). 13   Id.  9 10

BARFIELD PRINT.indd 244

13/11/2018 11:52

First Amendment, video games, and virtual reality training  245 rather to get better at the flying tasks featured in the game, the same First Amendment protection should apply. There is nothing wrong with such an answer as a first-cut response to the question of what status VR has in First Amendment law. But the second of the two points I have made above is subject to challenge: Just because flying a virtual plane in a video game counts as protected expression under current free speech doctrine, that does not mean the same is true of all use of a flight simulator. As Kevin Saunders notes, “[t]he interactivity of a flight simulator is a qualitatively different experience, and one more likely to lead to a response in actual flight, than reading a flight manual, watching one of the Airport movies, or reading the most engaging descriptions penned by pilots.”14 Saunders actually argues that this is true not only of flight simulators or other training technology, but also of video games: they can prepare individuals (and especially impressionable young game players) to take sometimes harmful or dangerous real-world actions, and not simply to immerse themselves in a fictional adventure. He thus rejects the Supreme Court’s holding in Brown v. Entertainment Merchants Association that video game play should generally count as protected speech despite its interactivity.15 But even if we accept the Court’s extension of First Amendment protection to video games, that holding alone does not clearly require that courts go further. Just because video game technology is typically used to tell stories or generate interactive fictional adventures, and this is First Amendment speech (according to the Supreme Court), does not mean that other uses of the same technology also count as First Amendment speech: If a video game-like interface is used, for example, to make a shopping experience more vivid, that does not mean that the purchase of, and payment for, products using that interface would become First Amendment speech. There may be protected expressive elements in the communications preceding or surrounding that purchase—elements that would likely count as “commercial speech,” and thus receive some protection against restriction. But the selling and buying itself would almost certainly count as economic “conduct” rather than First Amendment “speech,” even when accomplished with words, or a visual medium akin to that used in video games. Scholars have already made similar observations. Jack Balkin, for example, has argued that the constitutional status of what occurs in a virtual world depends on how it is designed and how it is used. On one hand, many virtual worlds are designed to tell, and let game players participate in, the construction of a story.16 Where virtual-world activity counts as “collective storytelling,” says Balkin, it should be protected by the First Amendment. On the other hand, virtual worlds may also be a platform that “offers entertaining ways for people to shop,”17 or other forms of “real-world commodification of virtual worlds.”18 Balkin’s focus is on

14   Kevin W. Saunders, Regulating Youth Access to Violent Video Games, 2003 L. Rev. Mich. St. U. Det. C.L. 51 71–2 (2013) (emphasis added). 15   Kevin W. Saunders, A Comparative Look at Children and Free Expression, 22 Transnat’l L. & Contemp. Probs. 455, 494 n.244 (2013). Justice Breyer makes a similar point in his dissent in Brown v. Entertainment Merchants Association. See text accompanying notes 21–23 infra. 16   Jack Balkin, Virtual Liberty: The Freedom to Design and Freedom to Play in Virtual Worlds, 90 Va. L. Rev. 2043, 2043–50 (2004). 17   Id at 2068. 18   Id at 2070.

BARFIELD PRINT.indd 245

13/11/2018 11:52

246  Research handbook on the law of virtual and augmented reality use of two-dimensional virtual worlds, such as Second Life, but his distinction intuitively applies to—and provides a First Amendment framework for—three-dimensional virtual reality as well. It is unclear, however, which side of this line virtual training falls on. On the one hand, for reasons I have already mentioned, training can have harmful effects in the real world and if, as Balkin argues, government should be able to regulate virtual-world behavior when it has “harmful effects in the real world,”19 then perhaps training should likewise be outside the scope of the First Amendment. On the other hand, training uses some of the same elements that make video games protected speech: individuals engage in roleplay within fictional scenarios, and wield simulated objects to confront challenges presented in a phantom environment. Justice Stephen Breyer briefly explored this question while dissenting in Brown v. Entertainment Merchants Association. In Brown, the Court had found that California violated the First Amendment when it forbade the sale of violent video games to minors (after finding, as noted above, that video games were protected speech).20 Breyer agreed that video games were protected by the First Amendment, but pointed out (echoing Saunders’s arguments) that they were different from reading books or watching movies— because playing a violent video game involving shooting is not only gaming, it is “an interactive, virtual form of target practice.”21 Elaborating on this point, he stressed: [V]ideo games are excellent teaching tools. Learning a practical task often means developing habits, becoming accustomed to performing the task, and receiving positive reinforcement when performing that task well. Video games can help develop habits, accustom the player to performance of the task, and reward the player for performing that task well. Why else would the Armed Forces incorporate video games into its training?22

Breyer then argued that even though video games are “speech,” they are speech which government must have the power to restrict and regulate when the skill and habit development is used not for a “beneficial purpose,” such as training soldiers, but rather “put to less desirable ends” in circumstances where “harm can ensue.”23 Breyer’s point seems even more relevant when we are thinking about the status of VR experiences that are not simply games that happen to instill skills and habits, but rather VR training modules whose primary function is to provide target practice, or otherwise transform thinking or behavioral patterns for purpose of enabling some future nonspeech conduct. This chapter will thus use VR training as a way to more closely explore the question of when VR activity receives First Amendment or other constitutional protection. It will begin by briefly exploring a more abstract question: How do courts, legal scholars and others address the more general challenge of marking the line between “speech” or “expression” that is protected by the First Amendment and nonspeech “conduct” that is not? It will then ask how the methods that courts and scholars have proposed to delineate the First Amendment’s scope might apply to this question and whether and when the First     21   22   23   19 20

Id at 2063. Brown v. Entm’t Merchants Ass’n, 564 U.S. 786, 790 (2011). Id at 849 (Breyer J dissenting). Id at 850. Id.

BARFIELD PRINT.indd 246

13/11/2018 11:52

First Amendment, video games, and virtual reality training  247 Amendment protects the use of VR to rewire one’s mental processes or behavior. Drawing on some specific illustrations of how VR is used for this purpose, it will then look at three possible arguments for extending First Amendment or other constitutional rights to cover at least some such uses of VR technology. One is that virtual training shares many of the elements of video game play, so it should ultimately (at least in many cases) receive the free speech protection courts have extended to the latter. A second emphasizes that even when training is interactive, a player’s interaction with a VR world generates knowledge acquisition of a kind that has generally been protected by First Amendment law: A virtual dry run of a complex surgery, for example, not only lets surgeons hone the fine motor skills they will need, but also allows them to simultaneously become familiar with (and able to navigate) aspects of human anatomy, and, perhaps, the anatomy of a particular patient. A third approach, drawing not only on First Amendment speech protection but also on the liberty protection one finds in the due process clauses of the Fifth and Fourteenth Amendments, argues that First Amendment protection for many uses of virtual reality is best understood as a variant of protection not for freedom of speech but for freedom and autonomy of thought. The Constitution, in other words, might assure that it is typically up to each person, not the government, to decide how to shape her own mental capacities. Such mental autonomy might allow her not only to seek knowledge of the world, or form her own ideas about it, but also to generate the “muscle memory” that allows her to more effectively interact with it and navigate through it. In short, I will briefly suggest here that there are at least three plausible bases for extending First Amendment protection to use of VR activity to hone one’s skills and reshape one’s mental processing and behavior. The constitutional protection can cover such VR activity: (1) because it is, like video game play, an expressive medium (or perhaps another, more practical use of the same expressive medium we use when we play video games); (2) because government should be barred from thwarting our use of VR (or other technology) to acquire knowledge the government wants to keep us from acquiring; or (3) because the Constitution gives us a right to exercise autonomy not just in acquiring knowledge, but in shaping mental capacity and character more generally. I will also briefly consider another possible implication of these arguments, especially the latter two approaches: if the First Amendment protects nonverbal training in VR environments on the grounds that it gives us new knowledge, or allows us to reshape our mental operations, perhaps it also protects some use of mechanized training environments, outside of virtual reality, that do so: If VR is protected because it allows us to hone skills or develop habits in role playing, or to generate knowledge in doing so, other methods of doing so may be entitled to the same First Amendment status when they serve the same function.

2. FIRST AMENDMENT COVERAGE AND EMERGING TECHNOLOGIES In order to understand why the First Amendment’s free speech clause might cover (at least some) virtual reality training, it is useful to begin with a more basic question: How, more generally, do we know when an activity counts as “expression” protected by the First Amendment? In many cases, there is little need to ask this question—particularly

BARFIELD PRINT.indd 247

13/11/2018 11:52

248  Research handbook on the law of virtual and augmented reality when people engage in a verbal activity, that is, when they communicate by talking or in writing. When, for example, somebody yells out a message, or holds up a sign supporting or opposing the current President of the United States, or perhaps expresses such support or opposition on a Web post or blog site, she is unquestionably engaged in constitutionally protected speech. That does not mean that all speaking or writing counts as protected “speech.” The writing in which a doctor engages when she fills out a prescription, for example, will likely count as the practice of medicine and be subjected (without First Amendment limits) to laws governing that professional activity. But political or religious discourse is speech. So are conversations about everything from sports loyalties, to the origin of life, to fashion preferences. Matters become more complicated, however, when we turn to the protection that the First Amendment extends to nonverbal activity, such as the artistic expression that one finds in painting or instrumental music, or in expressive conduct, such as burning a flag, draft card, or effigy as an act of protest. The Supreme Court has made clear that the First Amendment protects artistic expression, indeed, even art that is abstract or avant-garde: It “unquestionably” protects the “painting of Jackson Pollock, music of Arnold Schoenberg, and [nonsensical] Jabberwocky verse of Lewis Carroll.”24 However, this does not mean that any activity receives First Amendment protection as soon as someone regards it as “art”: If a performance artist decides to make a short video about carrying a fake bomb on an airport security line, they will likely have no First Amendment defense when they are arrested for violating airport security rules. Nor will hackers likely have a First Amendment defense when they claim that the sabotage they have engaged in by inserting malware on another computer was a way of expressing themselves (to protest the company they are sabotaging, or perhaps just to demonstrate skill and cleverness by tackling a very challenging hacking target). How then do we tell what type of activity is and is not “expression” covered by the First Amendment? A number of possible approaches to answer this question are helpfully and carefully set forth by Alan Chen, Mark Tushnet, and Joseph Blocher in their book, Free Speech Beyond Words.25 First, courts might try to define the boundaries of the free speech clause’s scope in a way that serves the underlying purposes of free speech protection. If free speech protection essentially erects an antigovernment forcefield of sorts around only certain kinds of human activity (whatever counts as “speech” under the First Amendment), then it is useful to understand why it does so. Why, more specifically, might the Constitution protect freedom to speak, but not give us a broader liberty to engage in other activities? As Chen, Tushnet, and Blocher write, scholars generally try to answer this question by drawing upon the “various normative theories that have been proposed to justify and explain the First Amendment.”26 If we know what ends the First Amendment is supposed to achieve, that will help us figure out what kind of activity (beyond clear cases of speech) it is supposed to protect.

  Hurley v. Irish-Am. Gay, Lesbian, Bisexual Group of Boston, 515 U.S. 557, 568 (1995).   Mark V. Tushnet, Alan K. Chen, and Joseph Blocher, Free Speech Beyond Words: The Surprising Reach of the First Amendment (Kindle Edition, NYU Press 2017). 26   Id. 24 25

BARFIELD PRINT.indd 248

13/11/2018 11:52

First Amendment, video games, and virtual reality training  249 One simple account of such First Amendment values is that expression is protected because it is an action that is free of the harm that comes with many types of nonspeech conduct, like driving a car, operating heavy machinery, or trading stocks. But this explanation is too simple. Certain activities that we think of as “nonspeech” conduct are unlikely to be harmful: for example, a solitary basketball practice session on a driveway court, or the 3D printing of a lamp. Certain kinds of “speech,” such as cutting words from family members or peers, can cause emotional pain and leave deep psychological scars. A harmbased test for what is covered by freedom of expression thus seems both too broad and too narrow. Consequently, we need other candidates to explain why speech is distinctively deserving of rights protection. As many other scholars have noted, the key candidates for such a principle are: (1) the notion that free speech generates a “marketplace of ideas,” or competition which moves individuals closer to an understanding of the truth; (2) the proposal, most closely associated with Alexander Meiklejohn, that the free exchange of ideas is a precondition for democratic deliberation and governance;27 and (3) the idea that freedom of speech is an essential component of individual autonomy, or an essential precondition for individual “self-development.”28 Whichever account of First Amendment purposes one adopts, such an account provides a basis for defining First Amendment coverage in ways that could extend to certain nonverbal activities. For example, in explaining why courts assume that instrumental music counts as First Amendment speech even when it lacks a message, Alan Chen explains how instrumental music may—like expression—support individual autonomy: The power of music to express, elicit and shape human emotions makes it a powerful tool, perhaps as powerful as the spoken word, for individuals both to communicate their feelings to others, and also to generate certain feelings in their own minds.29 Moreover, as Chen makes clear, because distinctive forms of music are often central to particular cultures, “autonomy over the formation of one’s cultural, religious, or moral identity” might also be “advanced by protection against government control over the range of musical expressions of that identity.”30 Second, one might try to delineate the boundaries of First Amendment coverage in more modest ways, either because First Amendment purposes are difficult to identify with clarity (and courts and scholars disagree about what they are), or because identifying a general First Amendment purpose (such as promoting autonomy) will rarely—by itself—answer specific questions about what speech or nonverbal conduct is and is not protected. Too many activities, other than expression, advance individual autonomy or self-development: We exert control over our lives, for example, when we exercise to improve physical fitness, take certain medications, or travel to other countries. The autonomy rationale is thus overinclusive. As Daniel Halberstam points out, the same problem plagues some variants of the claim that free speech exists to promote democratic governance: “Some forms of 27   Alexander Meiklejohn, The First Amendment Is an Absolute, 1961 Sup. Ct. Rev. 245, 255 (1961). See also James Weinstein, Participatory Democracy as the Central Value of American Free Speech Doctrine, 97 Va. L. Rev. 491, 513 (2011); Robert Post, I, 97 Va. L. Rev. 477, 482 (2011). 28   See, for example, Martin H. Redish, The Value of Free Speech, 130 U. Pa. L. Rev. 591, 620 (1982); C. Edwin Baker, Human Liberty and Freedom of Speech (Oxford University Press 1989). 29   Tushnet, Chen, and Blocher, supra note 25, at 1296. 30   Id.

BARFIELD PRINT.indd 249

13/11/2018 11:52

250  Research handbook on the law of virtual and augmented reality conduct, such as, for example, working for a unit of government, working in a regulated industry, or working in a sector of the economy that is not currently being regulated as it should be, serve to create an awareness of the quality of government officials and government regulation and, thus, would appear to further the political speech principle as any speech about the conduct of government officials would.”31 One more modest method of defining First Amendment coverage is to proceed on the assumption that speech clause covers not all forms of human conduct that advance individual autonomy, democracy, or truthseeking, but rather a subset of such activity. It is only when we engage in self-development, democratic deliberation, or truthseeking in certain ways that we receive the First Amendment’s shielding. As Chen, Tushnet, and Blocher put this point, “[m]aybe only ‘speech’ acts that further the relevant value can or should be protected.”32 But this raises a question: “[W]hat acts count as ‘speech’ and why”?33 The most familiar way to limit the realm of free speech in this way is to assume that there is some essential characteristic of much speaking or writing that entitles it to First Amendment protection—and that courts are justified in protecting (and maybe obligated to protect) nonverbal conduct when the latter shares this defining characteristic of speech. An example of such an approach is the “Spence test,” which the court used in Spence v. Washington to determine if a person is justified in insisting that certain nonverbal conduct in which she has engaged, such as hanging a flag upside down or burning a draft card, is—under the specific circumstances—being used as a vehicle of communication, and thus deserves to be treated as First Amendment “speech.”34 Under the Spence test, displaying a flag upside down, burning a draft card, or engaging in some other typically nonexpressive form of conduct might nonetheless temporarily count as expressive when it performs one of the key functions often performed by speaking or writing—that is, conveying a “particularized message” to an audience in such a way that the audience will understand that it is a message and understand what it says.35 The idea behind the Spence test is that one can, at least in some circumstances, separate the function of spoken or written language from the act of speaking or writing, and find (perhaps with some creativity) a way to perform that communicative function without using words—for example, by protesting the Vietnam War with an upside down American flag (as Spence, the defendant, did in Spence v. Washington) rather than with words.36 The Spence test has been subject to harsh scholarly criticism, however, on the grounds that the characteristic it identifies as the underlying and defining function of speech—namely, communication of a “particularized message” to an audience likely to understand it—is neither a necessary characteristic of First Amendment expression nor a characteristic whose presence is, by itself, sufficient to justify counting conduct as First Amendment speech. It is not necessary because one can easily think of examples 31   Daniel Halberstam, Commercial Speech, Professional Speech, and the Constitutional Status of Social Institutions, 147 U. Penn. L. Rev. 772, 814 (1999). 32   Tushnet, Chen, and Blocher, supra note 25, at 1296. 33   Id. 34   Spence v. Washington, 418 U.S. 405, 410–11 (1974). 35   Id. 36   Id.

BARFIELD PRINT.indd 250

13/11/2018 11:52

First Amendment, video games, and virtual reality training  251 of protected First Amendment expression that lack any such particularized message. Indeed, the Supreme Court did so in Hurley v. Irish-American Gay, Lesbian and Bisexual Group of Boston, Inc.: In finding that parades are expressive, Justice Souter noted that it is pointless to search for such a particularized message in the “unquestionably shielded painting of Jackson Pollock, music of Arnold Schoenberg, or Jabberwocky verse of Lewis Carroll.”37 So the First Amendment staunchly protects artistic expression where a particularized message is absent. And as Robert Post points out, free speech doctrine often fails to provide any protection to conduct where a particularized message is present and likely to be understood by onlookers—for example, in an act of vandalism where the damage someone does to a wall or other structure is accompanied by a clear message of hatred or violence.38 The Spence test is thus, at best, incomplete. One can respond to the Spence test’s flaws by continuing to search for some other defining element of First Amendment expression—one that might be found in abstract art or music as well as spoken or written language, and might be absent from clearly unprotected acts such as vandalism. But scholars such as Post argue that such a search is fruitless: There is no characteristic we can find in “speech as such” that justifies distinguishing it from other “nonspeech” conduct as an object specifically deserving of constitutional protection.39 How then might we define First Amendment “speech” if such speech has no common defining element? One answer is that what makes a form of conduct “speech” is not something internal to it, but rather a feature of how Americans regard or understand it. For example, commercial cinema was once treated as nonspeech by courts. In a 1915 case, Mutual Film Corp. v. Industrial Comm. of Ohio, the Supreme Court rejected a claim that screening of a motion picture in a movie house was protected speech: “[T]he exhibition of moving pictures,” it said, “is a business, pure and simple, originated and conducted for profit, like other spectacles, not to be regarded, nor intended to be regarded by the Ohio Constitution, we think, as part of the press of the country, or as organs of public opinion.”40 In Joseph Burstyn, Inc. v. Wilson, 37 years later, the Supreme Court came to a different conclusion regarding motion pictures’ First Amendment status: “It cannot be doubted,” it stated, “that motion pictures are a significant medium for the communication of ideas. They may affect public attitudes and behavior in a variety of ways, ranging from direct espousal of a political or social doctrine to the subtle shaping of thought which characterizes all artistic expression.”41 Nor, said the Court, was First Amendment coverage ruled out by the profitseeking nature of those who created or screened the film. The same profitseeking motives lay behind the distribution of “books, newspapers and magazines,” all of which were clearly within the First Amendment scope. What likely changed between the times at which each of these cases was decided—apart from significant changes in First Amendment law after World War I—was not a matter of the inherent characteristics of cinema that made it more expressive: It was, although silent 37   See Hurley v. Irish-Am. Gay, Lesbian, Bisexual Group of Boston, 515 U.S. 557, 569–70 (1995). 38   Robert Post, Recuperating First Amendment Doctrine, 47 Stan. L. Rev. 1249, 1252 (1995). 39   Id at 1279–80. 40   Mutual Film Corp. v. Industrial Comm. of Ohio, 236 U.S. 230, 244 (1915). 41   Joseph Burstyn, Inc. v. Wilson, 343 U.S. 495, 500–501 (1952).

BARFIELD PRINT.indd 251

13/11/2018 11:52

252  Research handbook on the law of virtual and augmented reality and technologically more primitive, just as much a device for storytelling in 1915 as it was in 1952. What changed was instead that Americans (including judges) had come to view cinema differently. They had come to view it as a cultural practice by which individuals tell stories, and sometimes educate viewers. Robert Post suggests that something like this provides the best guide to First Amendment coverage. First, he says, courts should not ask whether particular acts of speech fall within the coverage of the First Amendment, but should rather ask that coverage question of social practices: “‘speech as such,’” he says, “has no constitutional value, and so the Court’s project [in attempting to identify its essential characteristics] has been doomed from the start. Constitutional value inheres instead in specific forms of social order.”42 Moreover, what matters is not only that a particular medium can potentially serve as a social practice in which communication takes place but whether it is a “recognized” or “accepted” medium for the communication of ideas.43 In other words, it is not the inherent nature of the social practice in question, but whether it has come to be accepted by social convention as a medium for intellectual exchange, and perhaps for conveying emotions as well as ideas. Looking at social convention is therefore a third means of defining First Amendment coverage, apart from looking at fundamental First Amendment purposes or trying find some essential characteristic of speech (or a practice involving) that applies to all acts of speech deserving constitutional protection. In a sense, this method still focuses on an essential shared characteristic of certain social practices that include verbal or nonverbal conduct—namely, the characteristic of being an integral part of a medium or method of expressing ideas that has been recognized by social convention as requiring insulation against state control. Determining whether that characteristic is present, of course, requires courts to look not only at the speech activity in question but also at its history, and at sociological understanding of the role it plays in contemporary life. As Chen, Tushnet, and Blocher write, the challenge here “is to identify the social practices and conventions that constitute human expression and communication.”44 A fourth possibility is to call off the search for any shared characteristic that might endow certain acts or social practices with First Amendment value (whether it is an inherent feature of such conduct or lies in social convention), and instead proceed by analogy. Once we recognize a particular social practice as having First Amendment value, we might recognize other practices as candidates for entry into the realm of the First Amendment on the basis of a “family resemblance” to it. Thus, Mark Tushnet suggests that abstract art may count as First Amendment speech, in spite of its lack of words, because it has enough in common with political communication to merit First Amendment protection: “Artworks are sometimes intended to communicate relatively precise messages; they are sometimes the object of suppression because of their assumed political content; they contribute something to the development of a democratic culture; and perhaps more.”45 That perhaps is enough to find that, if political speech is protected, abstract art should be as well—and that it is better to have courts protect the entire medium of abstract art

 Post, supra note 38, at 1279.   Id at 1255, 1259. 44   Tushnet, Chen, and Blocher, supra note 25, at 2065–7. 45   Id., at 2065–7. 42 43

BARFIELD PRINT.indd 252

13/11/2018 11:52

First Amendment, video games, and virtual reality training  253 rather than asking, in each case, if it is political enough or worries government enough to merit protection in that particular case. There may also be other ways of defining First Amendment coverage by analogy: Even without using Tushnet’s “checklist” method of establishing family resemblance, we might have the sense that, for example, if we admit movies and graphic novels to the realm of First Amendment protection, we cannot easily exclude video games, which are in some respects like interactive movies, and in others like computer-generated interactive graphic novels. We need not be able to identify some essential characteristic that is shared by video games, movies, and comics—and that ties them to all other forms of protected speech—to find that they all belong in the realm of First Amendment protected activity. We can just proceed by noting that the resemblance between video games and the already protected categories of movies and comic books is strong enough to treat them the same way. Such analogies may well arise from social developments rather than judicial opinions: Even before judges recognized resemblances between video games and films, for example, people interested in immersing themselves in fictional universes came to realize that they could do so with a video game console and not only by watching movies. It is useful to understand how each of these models might provide a foundation for extending First Amendment coverage to VR training. If we learn to drive by steering a phantom car, or learn to fly a plane by piloting an illusory plane, why might doing so be protected speech when its real-life equivalent (learning with a physical car or plane) would not be? First, accounts of First Amendment purposes may do this on the ground that VR experiences help expose us to truth in ways that might be otherwise impossible, impractical, or unsafe (for example, by allowing us to enter and explore a cell nucleus, or visit a volcano or other feature of the natural environment far from where we live, or conduct surgery on a virtual patient). All of these arguments, however, seem to prove too much: It is not only a virtual reality road trip that can let us explore new natural habitats, visit and immerse ourselves in the lives of other Americans, or train ourselves in a new skill (like driving a truck). A real-life road trip can serve all of these purposes as well. Nor does the Spence test provide an answer. If virtual reality driving, or a ride on a VR plane, seems intuitively to count as speech, the Spence test seems like a poor candidate to explain that intuition, much for the same reason it seems like a poor candidate for explaining why the First Amendment should cover abstract art. Where is the particularized message in our sense of speeding down a road or flying through the skies? In some cases, the presence of a message may help show that certain virtual training clearly does count as First Amendment speech. The Supreme Court made this clear in Holder v. Humanitarian Law Project. In that case, American organizations and individuals wished to train members of the Partiya Karkeran Kurdistan (or PKK) “to use international law to resolve disputes peacefully.”46 However, the State Department had classified PKK as a “foreign terrorist organization” and a federal statute enacted by Congress barred Americans from providing “material support” to any group that was so classified, including any “expert advice” or “training.”47 When the organizations and individuals

  Holder v. Humanitarian Law Project, 561 U.S. 1, 2 (2010).   Id at 13–15.

46 47

BARFIELD PRINT.indd 253

13/11/2018 11:52

254  Research handbook on the law of virtual and augmented reality challenged this material support statute as a restriction on their speech, the government argued that the training of terrorist groups was not speech at all, but conduct.48 While the Court ultimately upheld the material support statute, as a constitutionally permissible speech limit, before doing so, it squarely rejected the government’s argument that training members of terrorist groups was entirely nonexpressive. The organizations were not simply “imparting a specific skill” to the PKK, they were doing so with speech. “The conduct” that made them liable under the material support statute “consists of communicating a message.”49 However, this judicial analysis does not tell us whether—and when—virtual training can count as First Amendment expression where it lacks any such message. Problems likewise arise when we turn to social conventions as key to defining First Amendment coverage. For a technology as new as VR, relying on social conventions is of questionable value. How do we define the contours of whatever “social practice” is relevant here? Although virtual reality technology is decades old, it is only now becoming familiar to consumers. Is use of VR in a video game, with a particular goal, part of the same social practice as use of VR to better experience a particular environment in Google Earth?50 Is use of VR to prepare for a particularly difficult real-world challenge—such as an upcoming surgery—the same as using it to improve visual acuity or develop general shooting skills? Are either or both of these uses of VR to prepare or shape one’s capacities part of the same social practice as use of VR (or less immersive technology) in video game play? It seems unlikely that sociological study can provide a clear answer to such questions. Indeed, it is not obvious that there are clear answers to these questions: There may be no consensus, nor any other objective test, of what kinds of actions count as being part of a single social practice and it may thus be a matter of choice whether we consider different kinds of VR roleplay as being part of a single social practice, with a single First Amendment status, or as a set of varied social practices, each of which may be treated differently under First Amendment law. Family resemblance or other analogy-based arguments might help here: Just as a person can sometimes gain admission to an exclusive club because of his or her close connection with those who are already members, perhaps use of VR for other skill development can gain admission to the “exclusive club” of First Amendment protected speech because of its close connection to video games. This is a variation of the argument I described in the introduction: Since the Supreme Court held, in Brown v. Entertainment Merchants Association,51 that video games count as protected speech, perhaps it is true (1) that this should be true of VR video games as well as those we watch on a flat screen and (2) that it is untenable to exclude from that protection very similar uses of VR for training, even where it lacks all of the characteristics of a game. If the technology we use to immerse ourselves in such skill development is the same as the Oculus Rift or HTC Vive consoles that individuals use to play video games (or create 3D artwork), perhaps it should be treated the same way for First Amendment purposes.   Id at 26.   Id. 50   I have previously explored this challenge in Marc Jonathan Blitz, A First Amendment for Second Life: What Virtual Worlds Mean for the Law of Video Games, 11 Vand. J. Ent. & Tech. L. 779, 807–8 (2009). 51   Brown, 564 U.S. 786, 790 (2011). 48 49

BARFIELD PRINT.indd 254

13/11/2018 11:52

First Amendment, video games, and virtual reality training  255 However, simply raising such a possibility does not tell courts how to analyze it. How do courts decide when an analogy between an unclear case and an already covered First Amendment activity is strong enough to extend First Amendment coverage to the unclear case? Is it simply a matter of the analogy’s strength (however one might define that) or are there other assessments we can make in deciding if such an analogy justifies treating new cases (and not just the already covered one) as within the First Amendment’s scope? Part of the problem here is that reasoning by analogy is, as Cass Sunstein argues, frequently a part of an “incompletely theorized mode of argument,”52 where the argument from an analogy (this new activity is like an already covered First Amendment activity and thus should be covered as well) substitutes for a more theoretically complete argument about First Amendment coverage (for example, this activity has those characteristics which, given the fundamental purposes of First Amendment free speech doctrine, justify treating it as covered by that doctrine). But if one does without deeper First Amendment theory, this raises the question of where one obtains criteria that allow courts to draw on more than i­ ntuition in ­determining  whether  ­analogies are close enough to merit similar constitutional treatment.

3. VIRTUAL TRAINING: THREE BASES FOR FIRST AMENDMENT PROTECTION 3.1 Virtual Reality Training as a Variation of Video Games (and Immersive Story Telling) Apart from this high-level look at how First Amendment arguments about coverage might apply to VR, and, in particular, use of it for training, it is useful to take a closer look at three more specific approaches—and, in doing so, to consider certain aspects of virtual reality training more closely. First, it is useful to look more closely at whether use of VR for training—in a realistic flight simulator, for example—might in some cases count as First Amendment activity on the ground that it is essentially a variant of video game play, and video game play should (as the Supreme Court itself has found) count as protected expression. As noted earlier, the Court said that video games count as speech because they communicate ideas through “familiar literary devices (such as characters, dialogue, plot, and music) and through features distinctive to the medium (such as the player’s interaction with the virtual world).”53 One finds certain elements of narrative, role playing, and dialogue in some virtual training as well. Consider how some police departments have recently used VR to train police officers to make difficult judgment calls about when to use force. The VR system they use is a variant of a Cave Automatic Virtual Environment (“CAVE”) which generates an artificial reality not with a head-mounted display (as Oculus Rift, HTC Vive and other VR gaming systems do) but rather by projecting images onto the walls of a room. One article

  Cass Sunstein, Legal Reasoning and Political Conflict 63, 67 (1996).   Brown v. Entm’t Merchants Ass’n, 564 U.S. 786, 790 (2011).

52 53

BARFIELD PRINT.indd 255

13/11/2018 11:52

256  Research handbook on the law of virtual and augmented reality describes the setting as a “dark, circular room almost entirely surrounded by screens.”54 A controller projects on to these screens scenarios featuring actors who interact with, and sometimes threaten or attack, the trainee, and “appear to respond to [his] commands.”55 To make the simulation more realistic, the technology also adds haptics, that is, a feeling or touch or force produced by events portrayed in the simulation: The trainee carries a physical (not virtual) handgun that has been modified so that it fires blanks, or infrared signals, rather than real bullets; and his chest is outfitted with a receptor which produces a small electric shock to give the trainee a sense that he has been hit by a “bullet” in the simulation.56 For the purposes of this chapter, what is significant about this VR police training system is that—like a video game—it consists of stories, interactions with characters, and goals the trainee has to aim to achieve. In one scenario, for example, the trainee confronts an individual who is illegally dumping debris off of a truck. When asked by the trainee to stop, the individual raises a rock and prepares to throw it at the trainee. Other scenarios feature a “pimp yelling at and verbally threatening a sex worker who seem[s] strung out,” “a group of kids who look stoned . . . blasting metal in the parking lot of a mall,” complaints about neighbors playing loud music, and a man’s ex-girlfriend blocking his driveway with her SUV and yelling at him.57 In each case, the trainee becomes a character who enters into conflict, or faces a potential attack, and has to make judgments about whether and how he can calm his interlocutors and how to notice and defend against physical attacks. As in a video game, the experience is interactive: How characters respond to the trainee depends on the trainee’s actions. The scenarios can be altered as they unfold by a supervisor running the computer system. As one article describes it, the training system is like “a good choose-your-own-adventure novel” in which the trainer can manipulate what happens next.58 Some VR training, of course, lacks such a narrative or interpersonal component. When a pilot-in-training flies a plane in a virtual flight simulator, it is possible he will encounter (and perhaps fight) enemy aircraft or try to evade attacks from other sources. But he may also simply focus on steering his plane through difficult weather. Still, he will be facing the kind of challenge that could be the basis for a video game, movie, or literary story (for example, the one in Antoine de Saint-Exupery’s book Night Flight59). The history of many types of virtual training technology also provides support for treating such technology as analogous to video gaming, because the same technology is often used for both purposes. Decades before the rise of computerized virtual reality, flight simulators provided a type of simulation and practice for pilots, beginning in the 1910s and 1920s. Edwin Link invented the “Pilot Maker” in the 1920s—a device that, as one article describes it, “looked like a toy aircraft with short wooden wings and fuselage

54   Chris Hayes, Could Virtual-Reality Training Be the Key to Fewer Police Shootings, Hive, Mar. 21, 2017. 55   Id. 56   Id. 57   Id. 58   Issie Lapowski, The Virtual Reality Sim That Helps Teach Cops When to Shoot, Wired, Mar. 30, 2015. 59   See Antoine de Saint-Exupery, Night Flight (Houghton Mifflin Harcourt 1974).

BARFIELD PRINT.indd 256

13/11/2018 11:52

First Amendment, video games, and virtual reality training  257 and was fixed to a universal joint mounted on a platform which could be made to pitch and roll using bellows activated by an electric pump.”60 Link’s flight simulator made its greatest impact after six Pilot Makers were purchased by the US military in 1943.61 His first customers didn’t purchase the Pilot Maker for training, however, but for entertainment at amusement parks.62 Flight simulator programs developed for computers have served the same double purpose. They are used to prepare for flying airlines, but they too have often been used as video games. And there are other examples of virtual simulations developed for gaming being used for training (or vice versa). Before developing some of the sophisticated VR simulations it currently uses to train soldiers in specific tasks or prepare them to fight in specific battles, the military adapted commercial video games for training: The US Army “modified the Atari tank battle game Battlezone to more closely mimic the gunner controls of the Bradley Infantry Fighting Vehicle,”63 the Marines modified a version of Doom II, and the Defense Advanced Research Projects Agency (or DARPA) encouraged commercial video game developers to create games that could train soldiers.64 The military also worked with a developer from Nintendo to develop the Multipurpose Arcade Combat Simulator, an arcade-like device used for training soldiers in shooting skills.65 Not surprisingly, modernday developers of virtual training also try to make modifications to repurpose their training technology for gamers, or to make a single VR environment serve both purposes simultaneously. For example, the company STRIVR develops VR technology used to train NFL quarterbacks.66 It captures 360-degree videos, filmed from behind a quarterback, and then transfers these video clips to a VR scenario that NFL quarterbacks use to plan plays and develop skills that will be useful in various game situations.67 Its VR training is also used by football kickers and professional hockey players. But apart from training professional athletes, STRIVR has also adapted its technology to give fans a chance to participate in the action—or to try playing the role of athlete themselves.68 Given the similarities, and the longstanding connection, between video games and virtual training, one might argue there is no basis for distinguishing the First Amendment status of the two: Both place individuals in the middle of an unreal environment; generally populate it with characters or (in the case of a flight simulator) nonhuman adversaries, such as threatening weather or equipment failures; and then challenge the VR user to summon (and develop) her skills in a way that will overcome these obstacles and achieve certain goals. In short, it is a technological adaptation of the storytelling one finds in 60   Jeff Newman, Virtual Becoming Reality: The Next Training Revolution, Naval Aviation News, Apr. 4, 2016. 61   Id. 62   Id. 63   Simon Parkin, How VR Is Training the Perfect Soldier, Wearable, Dec. 31, 2015. 64   Id. 65   Blake Stillwell, 6 Military Video Games Used to Train Troops on the Battlefield, Military. com, May 13, 2016. 66   Kurt Wagner, This Virtual Reality Startup Is Training NFL Quarterbacks to Be the Next Tom Brady, Recode, Feb. 3, 2017. 67   Id. 68   Id.

BARFIELD PRINT.indd 257

13/11/2018 11:52

258  Research handbook on the law of virtual and augmented reality books and movies, but with a twist that allows an individual to act as the active decisionmaker within the story rather than a passive audience watching it from “outside.” In fact, one might argue that VR training technologies are not simply analogous to virtual video games, but rather are part of the same “medium of expression” or “medium of communication.” Consider how this works when courts extend First Amendment protection to movies. Post argues that, in approaching the medium of cinema, courts have assumed that if “medium were constitutionally protected by the First Amendment, each instance of the medium would also be protected.”69 As a consequence, courts will likely protect an avant-garde film even where it lacks the narrative elements we find typically find in a movie. And one might add that in protecting film, it protects instances of communication that serve very different purposes: First Amendment protection for film and video extends to blockbuster action films, arthouse avant-garde animation, historical documentaries, and perhaps even instructional videos. Perhaps protection for video games could similarly include “games” whose primary purpose is to develop specific skills rather than to entertain. In any event, whether VR training is a kind of VR video game, or merely a use of VR that is analogous to such a video game, the approach I have been exploring here would bring it within the First Amendment’s coverage on the grounds that it too involves roleplaying, storytelling, and generation, reshaping, or viewing of visual, auditory, or other creations. It may well be true that VR training systems are far more likely than VR games to be part of a contractual or other fiduciary relationship where individuals rely upon the accuracy or effectiveness of the system. If so, the free expression protection accorded to VR training may often have to leave room for product liability, medical malpractice, or breach of contract actions. Imagine, for example, a situation where a surgical team uses VR technology to rehearse an unusually complicated surgery on a virtual recreation of a particular patient’s anatomy: Modern VR systems are being developed not only to help surgeons train for general skills (such as how to perform certain kinds of surgery, like temporal bone surgery), but also to incorporate “individual patients’ pre-op CT scans” to create “patient specific models” that “simulate the anatomy of a specific patient.”70 If there is an inaccuracy in such a patient-specific model, and the doctor using the VR system to rehearse the surgery is unaware of it, this could have detrimental consequences for a surgical procedure that is a matter of life and death. This might mean that First Amendment protection for VR use or design in such circumstances has to be weakened, or possibly even suspended entirely.71 This differential First Amendment protection for VR training—strong when it is used for entertainment or personal interest, but weak (or absent) where health, safety, or other crucial interests are at stake—would mirror the way First Amendment doctrine appears to apply to maps and charts. Since maps and charts communicate ideas, government cannot suppress them simply in order to prevent  Post, supra note 38, at 1253.   Hadley Leggett, Virtual Reality Could Keep You From Being a Surgical Guinea Pig, Wired, Aug. 10, 2009. 71   See Frederick Schauer, The Boundaries of the First Amendment: A Preliminary Exploration of Constitutional Salience, 117 Harv. L. Rev. 1770 (2004). Schauer observes here that product liability suits typically do not raise First Amendment issues even when the use of the product at the center of the suit includes expressive activity. 69 70

BARFIELD PRINT.indd 258

13/11/2018 11:52

First Amendment, video games, and virtual reality training  259 communication of these ideas. Certain maps, for example, take sides in territorial disputes among nations: Maps might differ in how they label the West Bank of the Jordan River, the Kurdish region of Iraq, Tibet, or Crimea—taking one side or the other in debates about who has a right to exercise sovereignty over such territory. Or they may communicate information about the different effects of climate change, or tax policy, in particular areas. If the US government wished to censor some of these maps because it objects to the ideas conveyed by them, it would almost certainly be violating the free speech clause of the First Amendment.72 Other restrictions on creating or disseminating information conveyed by maps may likewise run afoul of the First Amendment: One court found such a violation when a state barred dissemination of maps showing the locations of celebrities’ homes in Los Angeles.73 But this does not mean that all mapping information counts as First Amendment speech, or all regulation of maps violates free speech law: As Frederick Schauer points out, where maps and charts are relied upon by pilots to navigate safely, or by a power company technician to avoid danger while installing a cable, the map- or chartmaker can be sued for errors without any First Amendment problems being raised.74 The information in such maps is not there to express a particular stance on a political issue, or to convey information that someone wishes to explore as a matter of personal interest. It is rather information that workers facing dangerous tasks rely on to provide the accurate information they need for these tasks. The same might well be true of VR training content: Where it is there to entertain or inform personal interest, it may receive the same First Amendment protection as video game content does (under the Supreme Court’s holding in Brown v. Entertainment Merchants Association). Where it is relied upon to perform a job successfully, it might be treated as conduct rather than speech, or as expression that receives lower protection under First Amendment doctrine. 3.2  Virtual Reality Training as a Form of Knowledge Creation There is a second approach one might take to First Amendment protection for VR training, and that is to protect such training not simply on the ground that it offers a stage for narrative and roleplay, but also as it allows for the acquisition of knowledge—including both information about how the world looks and operates underneath the appearance it presents to us, and knowledge about how it sounds and feels. It is useful to briefly elaborate how this approach differs from the first one I described. I suggested above that the First Amendment might protect VR training, as well as VR gaming, as a platform for interactive storytelling, but remove or reduce that protection when such a storytelling platform is relied upon to accurately and effectively prepare certain types of experts (or others playing a professional role) for particular tasks, such

72   See Marc Jonathan Blitz, The Right to Map (and Avoid Being Mapped): Reconceiving First Amendment Protection for Information-Gathering in the Age of Google Earth, 14 Colum. Sci. & Tech. L. Rev. 115, 124–5 (2013). 73   Welton v. City of Los Angeles, 556 P.2d 1119, 1121 (Cal. 1976). 74   See Schauer, supra note 71 at 1765, 1803–04 and n. 173 (2004). For similar reasons, Lemley and Volokh argue that where VR or AR is justifiably relied upon by drivers or others using it to navigate, the VR or AR creator might be liable for inaccuracies. See Mark A. Lemley and Eugene Volokh, Law, Virtual Reality, and Augmented Reality, 166 U. Pa. L. Rev. (2018), 1105–6.

BARFIELD PRINT.indd 259

13/11/2018 11:52

260  Research handbook on the law of virtual and augmented reality as surgery. This approach is a variation of Jack Balkin’s argument for how the First Amendment should apply to virtual worlds: Protect them when they are stages for ­“collective storytelling” or imaginative roleplay, but leave them unprotected when they are stages for buying and selling, or other activity where financial welfare or physical health is at stake.75 On this view, moreover, the treatment of particular virtual training platforms may not just be a binary choice between extending or withholding First Amendment protection. It may be the case that certain elements of a single VR training platform are protected and others are not, or that certain government restrictions on how such a VR platform is designed or used count as First Amendment violations and others do not. Where government restricts use of a VR platform in order to prevent it from conveying a message or telling a story, then it may violate the First Amendment. Where government’s only concern, by contrast, is to prevent VR users from receiving inaccurate information from a VR system on which they rely to prepare them for a real-world task, its regulation may not be a matter of First Amendment law at all (so long as it is narrowly tailored to the goal of assuring accuracy). Moreover, perhaps there are aspects of a VR training system other than its accuracy in such circumstances that would fall outside the First Amendment’s coverage. In an early video game case, the Seventh Circuit Court of Appeals suggested—even as it extended free speech protection to the storytelling features of modern video games—that some elements of a first-person video game may lack such protection: If video games “lacked any story line and were merely animated shooting galleries,” government restriction, said Judge Posner for the Court, “might survive a constitutional challenge.”76 In other words, where a simulation is not an interactive story but merely a virtual version of a discrete physical activity, isolated from any larger narrative—when one is firing a weapon not to fend off a virtual adversary in a scenario, but just to engage in repeated shooting—then one is not engaged in expression. Justice Breyer made a similar point in his dissent in Brown, when he stressed that “[v]ideo games combine physical action with expression” and “[w]ere physical activity to predominate in a game, government could appropriately intervene.”77 But we might take a different approach to the First Amendment status of such narrativefree virtual simulations: We might find that they presumptively count as First Amendment expression so long as they increase our understanding of, and familiarity with, the world around us. In a prior article on virtual maps (and globes), I suggested that this might be a basis for First Amendment protection. Where individuals wish to “explore an environment” and can learn about it in a way that avoids imposing physical or other harms of the kind government is empowered to regulate, the First Amendment should protect their right to “observe the world free from government-imposed blinders.”78 Courts have found that the First Amendment entails a right not only to speak, but also to receive information contained in speech, and this might, I argued (drawing on prior arguments by Thomas Emerson for a “right to know” and with it a right “to see” or “listen”79), also justify a right   See Balkin, supra note 16 at 2043–50.   Am. Amusement Mach. Ass’n v. Kendrick, 244 F.3d 572, 579–80 (7th Cir. 2001). 77   Brown v. Entm’t Merchants Ass’n, 564 U.S. 786, 847 (2011) (Breyer J dissenting). 78  Blitz, The Right to Map, supra note 72, at 122, 182–3. 79   See id at 143 (citing Thomas I. Emerson, Legal Foundations of the Right to Know, 1976 Wash. L.Q. 1, 2 (1976)). 75 76

BARFIELD PRINT.indd 260

13/11/2018 11:52

First Amendment, video games, and virtual reality training  261 to receive information from one’s environment, at least where it could be obtained in ways that avoid harm to others. One way in which individuals can explore the world without threatening others’ interests is to explore a “phantom copy of the environment”—as they do when they immerse themselves in a virtual simulation, where any impacts they have on virtual persons or property can be undone by resetting the program.80 Jane Bambauer has set forth a similar, but more ambitious, argument for extending First Amendment protection not only to our observation of, and harmfree interaction with, surrounding environments (physical and electronic), but more generally to all processes in which we engage to obtain or generate knowledge. The First Amendment, she argues, implicitly includes a “right to create knowledge,” one that “protect[s] observation and thought.”81 Indeed, even where observation or other knowledge generation does entail a risk of harm, she argues, the First Amendment should still protect it, but with a level of scrutiny that lets government address the harm without unnecessarily restricting the generation or flow of knowledge.82 Taking such a First Amendment approach as our foundation, we might argue that virtual training merits protection not merely because it, like video games, immerses us in an interactive story, but for a more fundamental reason: It allows us to learn about the world by observing and interacting with a simulation of it. Physicians can learn about human anatomy not simply by reading about it in a book, but by dissecting or manipulating virtual human organs. One medical VR training program, for example, allows physicians to “explore and manipulate a lifelike human heart.”83 This happens to have tremendous value in preparing physicians for specific tasks, such as heart surgeries. But it also allows physicians to learn about elements of the heart’s anatomy and physiology that were not known to them before. Moreover, knowledge is gained not merely through visual perception, but also through use of our other senses. We learn about the world through our sense of touch, and also when our memory registers how our body feels as it moves in and interacts with the world. For this reason, it is important for many VR training programs to mimic not just what we would see in a certain scenario, but how the world would feel to us. Consider a doctor’s description of a virtual surgery training program developed for simulating surgery on the temporal bone, which includes the challenge of operating near major arteries: Adequate preparation for this surgery, he says, requires “an integration between different senses.”84 A trainee needs not only to see the parts of the head and neck she is operating on, but also to “feel the resistance of the drill.”85 They need to experience this as well as the “realistic sounds and squirting virtual blood.”86 All of these sensations—sight, sound, and touch—provide invaluable information to a surgeon as he conducts an operation, so a VR training program missing any one of these would be incomplete. The knowledge

  Id at 122.   Jane Bambauer, Is Data Speech? 66 Stan. L. Rev. 57, 60–61 (2014). 82   Id at 105–9. 83   Jennifer Jolley, Doctors are Saving Lives with VR, USA Today, Jul. 28, 2017. 84   Hadley Leggett, Virtual Reality Could Keep You From Being a Surgical Guinea Pig, Wired, Aug. 10, 2009. 85   Id. 86   Id. 80 81

BARFIELD PRINT.indd 261

13/11/2018 11:52

262  Research handbook on the law of virtual and augmented reality we gain is obtained not only from the light that enters our eyes, but from all of the other sensory information our body absorbs from the world (and, sometimes, processes and retains outside of consciousness). Indeed, it is more than the external sensory inputs our body receives from the world, but the more global feeling we have of being in a particular environment. As one writer notes, “one of the most distinctive features of virtual environments is that they not only present us with images, but a sense of ‘your visceral presence in the world.’”87 One might argue that this approach to defining the First Amendment’s scope sweeps in too much. Government would be virtually paralyzed if the free speech protection covered not just the words we read, or artistic expression for which we serve as an audience, but everything we see and feel. Virtually every government measure places some limits on the ways in which we can experience the world, and in so doing places limits on the interactions we can have with it, or knowledge we can obtain from it. As the Supreme Court itself said in 1965, “there are few restrictions on action which could not be clothed by ingenious argument in the garb of decreased data flow.”88 On the view of the First Amendment considered here, this might mean that any restriction on action could be understood as a restriction on knowledge generation, and thus as a violation of the Constitution. But there are at least two defenses we might raise against such an argument. One of them is set forth by Bambauer in her First Amendment article: First, we might understand free speech doctrine as protecting us not against any restriction that happens to place limits on what we see or experience (as that would make virtually every restriction a matter of First Amendment law), but only against those restrictions that government imposes for the purpose of restricting our knowledge of the world.89 Where government stops us from observing events not to protect others’ interests but specifically to keep us in ignorance of a certain feature of the world, it is pursuing a motive normally viewed as illegitimate in a society committed to freedom of speech and thought, namely, the motive of controlling an individual’s understanding of her world. Second, observation of, and interaction with, virtual reality may well deserve stronger insulation than our encounters with physical reality. This is something I stress in Chapter 10 of this volume, on the contrast between our interactions with physical and virtual versions of public space. Because in VR we are often interacting with phantom images (or other stimuli) that are sensible only to us and other consensual users of the virtual space, government has less basis to claim that interference with such environments is necessary to protect others. In regulating virtual space, government is extending into a realm of private sensation, not social coordination of different interests that occur in the physical world. In other words, in VR we experience the world without needing to examine, or disturb, its physical substrate: Because a virtual surgery involves cutting open an illusory body, it does not cause any more physical damage than passive observation does in other circumstances. Thus, if government were to restrict such uses of virtual technology, and someone raised a free speech objection against such restriction, it is not likely that the free speech action would simply be an attempt to thwart legitimate government regulation by “cloth[ing]”

  Laura Hudson, “It’s Not Real. It’s Not Real. It’s Not Real,” Slate, May 2, 2016.   Zemel v. Rusk, 381 U.S. 1 (1965). 89  Bambauer, supra note 81, at 63, 87. 87 88

BARFIELD PRINT.indd 262

13/11/2018 11:52

First Amendment, video games, and virtual reality training  263 it in “the garb of decreased data flow.”90 The virtual experience consists entirely of data flow (without the physical substrate that normally gives rise to it), so it is hard to see why government would limit our use—except to limit our access to the data inside of it, whether our means of sensing and registering that data is visual, auditory, or tactile.91 Of course, there are circumstances in which government is justified in limiting our access to data. Privacy laws generally aim to do so: They prevent us from gaining information about other people’s private affairs, or gaining it in ways that would violate others’ privacy (for example, by intercepting private phone communication or emails, or installing a hidden camera in their homes). Similarly, when security officials have raised concerns about terrorists’ use of virtual reality, the concerns they have emphasized regard a terrorist’s ability to gain information about, and thus prepare for attacks on, certain targets, or to find out how to evade certain counterterrorism measures. This, however, does not mean that the information conveyed by virtual reality must go unprotected by the First Amendment, but rather that any such protection must leave room for government to pursue legitimate interests, such as protecting privacy or guaranteeing security against terrorism (or other forms of crime). This does not mean it would be an easy or straightforward task for courts to balance speech rights and interests in privacy or security. Scholars have elsewhere explored how courts might begin to think about such a challenge in the context of image capture.92 It is simply worth mentioning that even if there must be First Amendment limits on the experience and imagery gained from virtual experience, this does not mean there can be no First Amendment coverage. Moreover, in some circumstances, it may be possible (and better) for government to monitor use of virtual training technology, at least where they have grounds to suspect possible criminal activity, rather than to restrict use of it altogether. 3.3 Virtual Reality Training as an Exercise of Mental Autonomy and Self-Transformation Finally, there is a third possible basis for extending First Amendment protection for virtual training. It is similar to the one I have just considered, but it extends First Amendment coverage not only to the steps we take to gather information (whether visual, auditory, or tactile) but also to the other ways that we use VR to transform our mental functioning. Consider the ways that video games can­­—and according to some studies do—enhance aspects of our visual and cognitive processing. According to cognitive scientists C. Shawn Green and Daphne Bavelier, “action-video-game playing is capable of altering a range of visual skills” including “attentional capacity.”93 Video game play also appears to enhance contrast sensitivity, that is, the ability to distinguish objects from other objects in the field

  Zemel, 381 U.S. 1 (1965).  Blitz, The Right to Map, supra note 72, at 190. 92   See Margot Kaminski,  Privacy  and  the  Right  to  Record,  97  B.U.  L.  Rev.  167,  202  (2017), Justin Marceau and Alan Chen, Free Speech and Democracy in the Video Age, 116 Colum. L. Rev. 991 (2016), Marc Jonathan Blitz, James Grimsley, Stephen Henderson, and Joseph Thai, Regulating Drones Under the First and Fourth Amendments, 57 W. & Mary L. Rev. 49 (2015); Blitz, The Right to Map, supra note 72, at 190–201. 93   C. Shawn Green and Daphne Bavelier, Action Video Game Modifies Visual Selective Attention, Nature 29; 423 (6939) 534 (2009). 90 91

BARFIELD PRINT.indd 263

13/11/2018 11:52

264  Research handbook on the law of virtual and augmented reality of view on the basis of shade or color (for example, the ability to see a gray circle when it is imposed on a gray background only slightly lighter or darker). The studies reaching these findings did not focus on immersive games in VR. But virtual reality scenarios can also alter cognitive function, and it is for this reason that VR has recently drawn interest from clinical psychologists. Psychologists have experimented with using it to treat post-traumatic stress disorder, by exposing individuals to VR simulations of the events that generated the trauma. They have also used VR to help treat phobias, such as fear of heights or of flying. When a person thus transforms their perceptual capacity or mental functioning, they are not simply adding to their store of knowledge. Their brain is certainly changing in certain ways—as it does when it stores a new memory, whether that is a visual, auditory, or tactile one. But the brain is not only changing in a way that adds new content to the record it keeps of past experience, but also in the way it processes certain kind of stimuli, or the fear it generates in response to certain kinds of external conditions. In other words, such uses of VR do not only add to our collection of facts and remembered experiences; they also add to our repertoire of habits and capacities. One might argue that this aspect of a virtual reality training experience is simply a variation of the knowledge-creating aspect of VR which I just suggested may be protected by First Amendment protection for perception or other receipt of information. Consider again the VR training used by many police departments (and described earlier in this chapter): When the company VirTra designs this police simulation technology it uses CAVE VR technology—where the virtual images appear on screen surrounding a room rather than within googles—in large part because it wants the trainee to “learn proper muscle memory,” by engaging in movements akin to those she would engage in in real life (not those that occur with goggles that would be absent from acts in the physical world). Here, the VR experience is not only generating visual, auditory, or tactile memories—it is also creating, giving the trainee new habits of thinking or movement. It is training him to react (perhaps unconsciously) in a certain way when confronted by certain circumstances. This mental change, one can argue, is like a remembered visual, auditory, or tactile experience: It still involves generation, and later retrieval, of memory. Even when it is “muscle memory” rather than memory entailing consciously remembered sensory experiences, it is still memory stored by the brain for future retrieval and use. Psychologists and neuroscientists even describe it as such: It is “procedural memory,” which allows us to perform certain tasks, rather than the kind of “declarative memory” we can consciously bring to mind when we recall “episodic memories” of life experiences, or “semantic memory” of facts about the world.94 This is captured in the way we describe our capacities to perform physical actions that we have learned: We speak of “remembering” or “knowing” how to play the piano, drive a car, or do a backward somersault. Moreover, some types of visual or auditory memory are similar to procedural memory in that they are also “implicit”—rather than consciously generated or recalled—and can thus, without the person’s awareness, change (and improve) how a person responds to particular challenges.95

94   Daniel L. Schacter, Searching for Memory: The Brain, the Mind, and the Past, 16–17 (Basic Books 1996). 95   Id at 182–90.

BARFIELD PRINT.indd 264

13/11/2018 11:52

First Amendment, video games, and virtual reality training  265 However, one might argue that extending the First Amendment or other constitutional rights protection to the generation of muscle memory is, at best, highly questionable. In the first place, when memories—such as certain procedural memories—inherently work by triggering certain bodily action, one might argue that this distinguishes them from the speech that the First Amendment protects (which we use to communicate, or give clearer form to, sensations and concepts) and is more closely connected to nonspeech conduct (like driving or fighting). Is some of the brain rewiring involved in creation of these behavioral habits nonetheless protected by the First Amendment, or some other part of the Constitution? One might argue that, if it is, that is not because self-transformation of this sort is itself protected by the First Amendment, but rather because we are protected when we transform ourselves by speaking, serving as an audience for expression, or acquiring knowledge. In other words, it is books, movies, video games, or virtual reality learning that are protected, not the self-transformation we happen to generate with such means of expression or information-gathering. When we instead use a sequence or pattern of nonspeech action (such as repeated actual or virtual shooting or driving) to transform our “muscle memory,” this—one might argue—is more like using gym equipment to develop our bodily strength, or practicing baseball to develop our hand–eye coordination. That may be right. And Justice Breyer seems to draw upon such an argument in his Brown dissent when he states that, although the narrative and artistic qualities of video games endow them with the status of First Amendment speech, the habit forming and habit reinforcing that often accompanies video games’ interactive experience must be subject to government regulation.96 But it is also plausible that the Constitution provides at least some protection for exercises of mental autonomy itself. Perhaps government should be limited not only when it restricts what we say or believe, but also when it tries to prevent us from changing the way we understand and process the world, and the way we transform our minds and bodies to interact with it. It is best to understand this argument as having two components: First, one might argue, the “freedom of thought” guaranteed by the Constitution protects our ability to generate imagery—visible only to each individual—and this should extend not only to natural generation of imagined experiences (say, in a daydream) but also to artificial generation of private experience (with VR). Second, this freedom to control what we see, hear, and feel without government restriction continues to insulate us when we use VR technology not merely to generate a particular experience at a particular time, but rather to shape ourselves with these experiences (either by increasing our knowledge of the world, or our capacity to interact with it). The first of these two claims has been supported by various scholarly arguments and also finds some support (albeit less clearly spelled out support) in Supreme Court and other opinions. Free speech scholars have suggested that mental freedom lies at the root of the First Amendment’s protection for speech. Seana Shiffrin, for example, argues that one of the central values supported by free speech law is “the individual agent’s interest in the   See text accompanying notes 21–2 supra; Brown, 564 U.S. 786, 790 (2011).

96

BARFIELD PRINT.indd 265

13/11/2018 11:52

266  Research handbook on the law of virtual and augmented reality protection of the free development and operation of her mind.”97 Christian Halliburton has likewise suggested that mental freedom is at the core of the Constitution’s right to be let alone: In the American constitutional order, he writes, it makes sense to understand “the mind [as] a safe harbor from which all outsiders can and should be excluded should we so desire.”98 The Supreme Court has sometimes made statements that seem to support such protection for mental autonomy. It has said that the First Amendment protects “freedom of thought”99 or “freedom of mind”100 and more specifically protects against government measures that are premised solely on the desirability of controlling individuals’ mental lives.101 It has also said that a person’s “fantasies” are “his own and beyond the reach of government.”102 The Seventh Circuit has added that this does not mean that any regulation that affects thoughts implicates First Amendment protections: “only governmental regulations aimed at mere thought, and not thought plus conduct” can violate “the First Amendment’s freedom of mind mandate.”103 I have argued in a previous article that such protection of freedom of thought should encompass not only our ability to generate fantasies through daydreaming or other use of our natural mental capacity to generate internal imagery, but also an individual’s use of VR to experience a private simulated world.104 The freedom of thought courts have found in First Amendment law, I argued, “insulates against government control not only the realm of communication and debate, but also the realm of ‘fantasies,’”105 and this “is precisely the kind of realm that VR allows us to enter when we put on a headmounted display and instantly transport ourselves to an hallucinatory streetscape, city, or universe—a streetscape, city, or universe that appears only to us, and constitutes no part of the external material realm that the state is responsible for governing and policing.”106 For experiences such as driving a car or flying a plane, for example, VR brings them within the “coverage of freedom of thought” by “taking them out of the social realm where legitimate government power holds sway and into a private perceptual realm where, as in a dream or daydream, the government generally lacks any authority to issue commands.”107 Mark Lemley and Eugene Volokh similarly emphasize that “[t]he contents of one’s own VR sensory feed are very close to the contents of one’s thoughts and fantasies,”  97   See Seana Valentine Shiffrin, A Thinker-Based Approach to Freedom of Speech, 27 Const. Comment. 283, 285 (2011); see also Seana Valentine Shiffrin, Speech Matters: On Lying Morality and The Law 92 (2014).  98   Christian M. Halliburton, Letting Katz Out of the Bag: Cognitive Freedom and Fourth Amendment Fidelity, 59 Hastings L.J. 309, 314 (2007).  99   Palko v. Connecticut, 302 U.S. 319, 326–7 (1937). 100   See Wooley v. Maynard, 430 U.S. 705, 714 (1977) (quoting W. Va. State Bd. of Educ. v. Barnette, 319 U.S. 624, 637 (1943). 101   See Stanley v. Georgia, 394 U.S. 557, 566 (1969). 102   Paris Adult Theater I v. Slaton, 413 U.S. 49, 67 (1973). 103   Doe v. City of Lafayette, Ind., 377 F.3d 757, 765 (7th Cir. 2004). 104   See Marc Jonathan Blitz, Freedom of 3D Thought: The First Amendment in Virtual Reality, 30 Cardozo L. Rev. 1141, 1070 (2008). 105   Id at 1150. 106   Id. 107   Id at 1154. See also id at 1179–85 (considering analogies between virtual reality and dreaming, and hypothetical attempts to regulate either activity).

BARFIELD PRINT.indd 266

13/11/2018 11:52

First Amendment, video games, and virtual reality training  267 and people should no more be punished for the contents of their “sensory feed” than the content of their thoughts and fantasies.108 Such a “freedom of sensescape” might well be narrower than a broad right to freedom of thought: That we have a right which lets us use VR to generate visual stimuli or other sensations does not necessarily mean we have a constitutional right to change our mental state, or our mental operations, in other ways—for example by using a neurofeedback device that not only provides us with visual or other stimuli—but taps into our mental activity in order to allow us to visualize and better control it.109 Nor does it mean individuals necessarily have a right to create VR that functions like such a neurofeedback device, or that in other ways enables people not merely to perceive the visions and sounds of their choice, but also to systematically reshape the way their brain works. Moreover, such arguments that VR experience is covered by First Amendment freedom of thought—or perhaps a narrower “freedom of sensescape”—do not clearly answer the question of what happens when government regulates VR not in order to stop a person from imagining or perceiving something, but rather in order to stop the development of habits or tendencies that will cause harmful action in the world. After all, government officials might argue, people can build a custom-designed sensescape—with visual, auditory, and tactile sensations of their choice—not merely with VR, but also in a more traditional manner, by interacting with the physical world in certain ways. We can do so, for example, by entering a room designed to have a certain appearance, only sounds piped in through an audio channel, and walls or objects with a certain texture. More generally, we exercise some control over our sensescape every time we decide to take a certain action: Our sensescape—the collection of visual, auditory, tactile, olfactory, and internal bodily sensations—is the medium through which our experience of the external world is constituted. Presumably, if government officials restrict our actions in the physical world solely in order to prevent us from seeing things they don’t want us to see, or hearing things they don’t want us to hear, they would likely be committing the same “sin” under the Constitution as they commit when they seek to repress use of VR to explore or learn about the environment it depicts. However, this does not mean that government is (or should be) in all cases without power to control the way we use our environment to reshape our sensescape. They are certainly not without regulatory power when their concern is not our sensescape itself, but rather effects that our physical action has on other aspects of the physical world, and especially other individuals’ physical welfare. Nor are they without power to control and limit some of the inputs we might want to use for our sensescape: Privacy and copyright law, for example, may prevent us from obtaining certain types of images in certain ways. A more difficult question is whether government can control the way we generate certain sensations we experience when its concern in doing so is not simply to prevent us from seeing, hearing, or feeling something, but to prevent us from reshaping the way our mind and body function. This brings us to the second of the two claims I set forth earlier:

108   Mark A. Lemley and Eugene Volokh, Law, Virtual Reality, and Augmented Reality, 166 U. Pa. L. Rev. (2018), 1117. 109   See Marc Jonathan Blitz, Freedom of Thought for the Extended Mind, 2010 Wis. L. Rev. 1049, 1075–7.

BARFIELD PRINT.indd 267

13/11/2018 11:52

268  Research handbook on the law of virtual and augmented reality that freedom of thought included in the First Amendment gives us a right not only to conjure up a certain idea, perception, or other experience, but also to use such experiences for another more transformative exercise of mental autonomy—one that does not merely generate a thought, but transforms the self that does the thinking. In a recent article examining the distinction between speech and action, Frederick Schauer raises some questions relevant to this inquiry. He explores how a certain thought can be “harmful” in that it “increases the likelihood of harmful action,” and the reasons why such harmfulness raises challenges for an absolute prohibition on state interference in (or punishment of) thought.110 Our legal system, he points out, allows government to punish, or subject to civil liability, actions that, while harmless in isolation, make a subsequent harm more likely—such as driving at 20 miles above the speed limit (even if one hasn’t crashed or in any way affected other drivers) or possessing heroin (that one hasn’t ingested or sold). A satisfactory justification for freedom of thought, he says, requires explaining why a just legal system can punish such actions on the basis of the harm they make more probable, but can never punish thinking even where it similarly increases the likelihood of future harm. This challenge seems even more important when “the thought” that the government wishes to regulate is not pure thought, but rather a sequence of virtual perceptions that is generated only in order to be interwoven with physical conduct (like shooting an unloaded weapon or manipulating a steering wheel that controls a virtual car or plane). Where such VR technology, or perhaps other more traditional forms of training technology, has been specifically developed to, in Justice Breyer’s terms, enhance individuals’ performance of a future “practice task” by developing and reinforcing certain “habits” necessary to carry it out, then the latter type of activity is not just about thought, perception, and expression—it is about developing capacities for future physical action.111 One might respond that, if the government’s concern is future criminal action, the onus should be on government to show that the private VR experience it wants to prevent is closely linked to some specific future harm (and not an abstract fear that harm may arise in the future). This requirement is a familiar part of First Amendment doctrine. Government can no longer limit speech as “incitement” on the grounds that it has “bad tendencies.”112 It has to show that such speech is both directed to and likely to incite unlawful action.113 More generally, as the Court said in Ashcroft v. Free Speech Coalition, the First Amendment demands that government demonstrate more than “a remote connection between speech that might encourage thoughts or impulses” and the harmful action such thoughts or impulses may generate.114 One might impose similar requirements when government targets the perception or other thought enabled by VR. Lemley and Volokh appear to take a similar stance when they argue that the “freedom of sensescape” should shield our use of VR even where “VR images . . . may lead to bad 110   See Frederick Schauer, On the Distinction Between Speech and Action, 65 Emory L.J. 427, 442–9 (2015). 111   Brown v. Entm’t Merchants Ass’n, 564 U.S. 786, 840 (2011) (Breyer J. dissenting). 112   See David M. Rabban, The Emergence of Modern First Amendment Doctrine, 50 U. Chi. L. Rev. 1205, 1210 (1983). 113   Brandenburg v. Ohio, 395 U.S. 444 (1969). 114   535 U.S. 234 (2002).

BARFIELD PRINT.indd 268

13/11/2018 11:52

First Amendment, video games, and virtual reality training  269 behavior.”115 None of this directly answers Schauer’s question, which is why the familiar First Amendment doctrine I have just set forth is justified in the first place, if and when thought or speech carries just as much, and just as serious, risk of harm as does nonspeech action. But we might at least take the position that, if we find reason to adhere to this doctrine with respect to speech and artistic expression, there is also good reason to extend it to the stream of perceptions we can generate with VR. It is also possible that, if the Constitution shields the way we shape our minds with VR technology, such shielding is provided not by First Amendment doctrine (on freedom of thought or of speech) but rather by other constitutional provisions. One possibility, for example, is that certain VR training might be constitutionally protected where the skill that is the focus of the training is itself integral to performing a constitutionally protected activity: Where VR training enables someone to learn to engage in expressive action (such as photography or videography) for example, it may be protected by the First Amendment because of the expression it makes possible. Where VR training is “an interactive, virtual form of target practice,” of the kind Justice Breyer describes in his dissent in Brown, it may in some circumstances be protected by the Second Amendment’s right to keep and bear arms—in part, for one’s own self-defense.116 A right to protect oneself with arms would likely include a right to learn how to do so. As the Seventh Circuit court said in striking down certain Chicago restrictions on location and use of shooting ranges, the “core right to possess firearms for protection wouldn’t mean much without the training

115   Mark A. Lemley and Eugene Volokh, Law, Virtual Reality, and Augmented Reality, 166 U. Pa. L. Rev. (2018), 1117. Where a VR environment is developed or used specifically as part of a plan to commit an act of terrorism or other crime, then its creation or use may well be unprotected: As the Fourth Circuit Court of Appeals made clear in the case of Rice v. Paladin Enterprises, Inc., 128 F.3d 233 (1997), “speech—even speech by the press—that constitutes criminal aiding and abetting does not enjoy the protection of the First Amendment.” In that case, the Fourth Circuit refused to extend First Amendment protection to the publication of a book called “Hit Man: A Technical Manual for Independent Contractors” which provided detailed instructions for contract killers, and did so (the court found) with the intention that they be used to commit murders. If even a book is treated as outside of the protection of the First Amendment when it is created and published to aid and abet criminal activity, it is almost certainly the case that VR developed for this purpose would also be. What courts could not assume, under current doctrine, is that VR technology which is not developed for crime, but happens to be useful for it, would similarly fall outside of the First Amendment’s protection. Moreover, while courts have generally refused to extend any First Amendment protection to crime abetting speech, some scholars have argued that this issue is more complex than courts have recognized. Eugene Volokh, for example, raises the concern that a simple exclusion of all “crime facilitating speech” from First Amendment covers not only a hitman manual like that in Rice v. Paladin, but also “newspaper articles that mention copyright-infringing Web sites, academic articles that discuss computer security bugs, and mimeographs that report who is refusing to comply with a boycott.” Crime-Facilitating Speech, 57 Stan. L. Rev. 1095, 1104 (2005). He thus argues that crime-facilitating speech should only be excluded from the First Amendment’s scope when “(1) it’s said to a person or a small group of people when the speaker knows these few listeners are likely to use the information for criminal purposes, (2) it’s within one of the few classes of speech that has almost no noncriminal value, or (3) it can cause extraordinarily serious harm (on the order of a nuclear attack or a plague) even when it’s also valuable for lawful purposes.” Id, 1106. One could arguably defend applying a similar rule to uses of VR that might promote criminal activity. 116   Brown v. Entm’t Merchants Ass’n, 564 U.S. 786, 840 (2011) (Breyer J dissenting).

BARFIELD PRINT.indd 269

13/11/2018 11:52

270  Research handbook on the law of virtual and augmented reality and practice that make it effective.”117 Restrictions on use of VR shooting ranges could well be problematic for the same reason. And the Constitution may protect not only restrictions on training oneself in certain skills (such as production of speech or use of firearms) but also restrictions on honing of one’s skills more generally. Such protection is arguably included in the personal liberty guaranteed by the due process clauses of the Fifth and Fourteenth Amendments. The contours of this liberty are not entirely clear. The language of these amendments itself speaks generally only of individuals’ right against being deprived of “liberty” without “due process of law.”118 But the court has for decades now read this language to bar government from interfering in certain spheres of “intimate conduct,” no matter what procedures government might follow when it engages in such interference.119 For example, government cannot generally ban individuals from using contraceptives no matter what procedures it follows in imposing such a ban. In other circumstances, the liberty protection is weaker: Government may be able to interfere in certain realms of human action so long as it makes a certain kind of factual showing that such interference is necessary, and avoids causing certain kinds of harm. While the courts have not specifically stated that the due process clause protects mental autonomy, they have said that government might violate the Fourteenth or Fifth Amendments when it requires a prisoner or mental institution inmate to take antipsychotic drugs without adequate medical justification.120 Its focus in these cases has been the protection of bodily autonomy, and more specifically a person’s right to be free from unwanted medical treatment imposed by the state. But it is likely that many people would intuitively find psychotropic drugs—which change a person’s mental operation—to be far more intrusive than say, a medical procedure that simply involves withdrawing blood, closing a wound, or addressing some other bodily harm. As Sarah Stoller and Paul Root Wolpe put it, we “feel a more intimate connection to the activity in our brains than to the activity in our blood vessels.”121 And to the extent that we view brain transformation as being a transformation in the nature of our mind and person, then it is likely that this intuition comes with a sense that state control should be excluded, or at least strictly limited, in this sphere. Thus, if a state’s regulation of VR is intended to stop us from being able to develop our minds in a certain way, such a restriction might well violate the Fourteenth or Fifth Amendments even if it does not violate the First. It might violate the personal liberty that is central to a free society. Or, to use language the Court has previously used to explain the nature of such constitutionally protected liberty, individuals’ freedom to control and reshape the way their minds work might be “implicit in the concept of ordered liberty.”122   Ezell v. City of Chicago, 846 F.3d 888, 892 (7th Cir. 2017).   Id at 241. See U.S. Const. amend. V (“No person shall . . . be deprived of life, liberty, or property, without due process of law”), amend XIV (“nor shall any State deprive any person of life, liberty, or property, without due process of law”). 119   Lawrence v. Texas, 539 U.S. at 574 (2003). 120   See Washington v. Harper, 494 U.S. 210 (1990); Riggins v. Nevada, 504 U.S. 127 (1992); Sell v. United States, 539 U.S. 166 (2003). See also Rodney J.S. Deaton, Neuroscience and the In Corpore-ted First Amendment, 4 First Amendment L. Rev. 181, 183 (2006). 121   Sarah E. Stoller and Paul Root Wolpe, Emerging Neurotechnologies for Lie Detection and the Fifth Amendment, 33 Am. J.L. & Med. 359, 371 (2007). 122   Palko v. Connecticut, 302 U.S. 319 (1937). 117 118

BARFIELD PRINT.indd 270

13/11/2018 11:52

First Amendment, video games, and virtual reality training  271 It is, after all, hard to imagine a system of constitutional liberty that permits government to bar us from changing our thought process by rehearsing actions, or experiencing scenarios, in our “mind’s eye.” Moreover, the Fifth and Fourteenth Amendments, unlike the First, also protect certain aspects of our bodily freedom—including freedom from physical restraint and a right to reject unwanted medical treatment and other external interference in the decisions we make about our body. Thus, even where the mental change we generate with VR is change that reshapes our “muscle memory” as well as our beliefs or other knowledge, it may still fall under the coverage of a Fifth and Fourteenth Amendment shield for personal autonomy. It is also perhaps relevant for the constitutional status of VR training that the due process clause has been held, in certain cases, to protect parents’ decisions about the education of their children (for example, the decision to instruct them in a foreign language),123 and might also cover some aspects of individuals’ decisions about how to educate themselves—even where such education involves nonspeech conduct. In Meyer v. Nebraska, the Supreme Court struck down a state law that forbade teaching any person in a secondary school any language other than English. The Court found that this state measure intruded into a zone of autonomy established by the due process clause of the Fourteenth Amendment, which included not only the right of parents to control the education of their children, but also the right of an individual herself “to acquire useful knowledge.”124 Modern due process law is, in certain respects, quite different from the due process law of that era. And it is unclear whether the protection recognized by the Court in Meyer for rights to “acquire useful knowledge” might extend to modern forms of training—including training in physical as well as intellectual tasks. But as the courts decide what, if any, constitutional provisions might insulate VR training from government control, it may be that these cases, as well as the more recent ones I have discussed dealing with mental and bodily autonomy, could be relevant for courts deciding what VR activities are insulated by the due process clause against government interference. All of these possibilities might open the way not only to constitutional protection for VR training, but also to training that involves interaction with the physical world. After all, if the Constitution protects individuals from government interference as they acquire knowledge or reshape their mental functioning, why should it cease to offer any such protection when individuals do so within a physical flight simulator rather than a purely virtual one? Indeed, much of the VR training I have described above already includes physical elements. The police training I described earlier, for example, uses a CAVE VR system to project virtual characters onto the walls of a room. But the modified gun carried by trainees isn’t virtual. It is a physical gun they can hold and aim, and from which they can fire blanks or infrared energy.125 Flight simulators have often projected virtual images within a physical model of an airplane cockpit. As Joshua Fairfield writes, it may be the   Meyer v. Nebraska, 262 U.S. 390 (1923).   Id. Another 1920s decision, Pierce v. Society of the Sisters of the Holy Names of Jesus and Mary, 268 U.S. 510 (1925), similarly barred government from interfering in parents’ decisions about education (in this case, decisions to send children to private rather than public schools) on due process grounds. 125   Chris Hayes, Could Virtual-Reality Training Be The Key to Fewer Police Shootings, Hive, Mar. 21, 2017. 123 124

BARFIELD PRINT.indd 271

13/11/2018 11:52

272  Research handbook on the law of virtual and augmented reality case that for a long time it will be easier to “build fantasy worlds on top of real-world physicality” than to “import kinetics (i.e., the sensation of touch and balance) into a virtual world,” and so rather than recreating an entire training or other scenario within VR, VR technology may instead project virtual characters or objects onto “real objects, such as a blank-faced robot or a moving floor panel.”126 The presence of such physical elements doesn’t provide reason to deny VR experience the status of First Amendment expression. And training with similar physical models of a real environment may likewise merit First Amendment protection even in the absence of any virtual elements. If, for example, training of the kind used by police includes not only a physical (modified) gun, but also a real-life actor rather than a virtual character, it is still serving the same functions that I have considered above as a basis for providing First Amendment protection to VR: (1) It immerses individuals in a roleplaying narrative (although the practical purpose for which they do so could conceivably make a difference); (2) it gives individuals knowledge of the world they did not have before the training; or (3) it lets them not only increase their knowledge, but also develop mental habits and capacities they lacked prior to the training. It is thus possible that if the First Amendment protects VR on any of these bases, it would likewise protect mechanized simulations for the same reasons. Like VR, traditional video games, as Justice Breyer noted, include a “physical action” as well as “expression,” and yet they are still within the First Amendment’s coverage.127 Moreover, as I have argued in earlier work, the same expressive elements that give video games protection, might be found to justify protection for nondigital equivalents (like a mechanized rather than virtual pinball game).128 It is also possible that the First Amendment might protect use of VR training without protecting all brick and mortar methods of achieving the same methods. When VR environments create characters and objects largely out of light, sound, or tactile sensations divorced from any physical substrate, the elements of the interactive story they present are, as a general matter, more isolated than are practice sessions in the physical world. VR simulations are, in this sense, closer to the realm of private thought and fantasy than are simulations which consist solely of physical props. To the extent that First Amendment law must draw imperfect distinctions between different media of expression, and sometimes accord these media different treatment based on their typical features, it may presumptively protect VR versions of conduct of a kind that is unprotected when it occurs in the brick and mortar world. Moreover, as noted earlier, courts may have more reason, in applying First Amendment law, to be highly skeptical of government justifications for restricting use of a training or learning technology, such as VR, where the physical risks that often justify restrictions on expression in the shared public space are less likely to exist. As the court said in Stanley v. Georgia, government “cannot constitutionally premise legislation on the desirability of controlling a person’s private thoughts,”129 and as noted previously, this may mean that courts will at least be skeptical of government restrictions that aim to prevent individuals from modifying their beliefs, or 126   Joshua A.T. Fairfield, Mixed Reality: How the Law of Virtual Worlds Govern Everyday Life, 27 Berkeley Tech. L.J. 55, 77–8 (2012). 127   Brown, 564 U.S. at 840. 128  Blitz, A First Amendment for Second Life, supra note 50, at 809–11. 129   Stanley v. Georgia, 394 U.S. 557, 566 (1969).

BARFIELD PRINT.indd 272

13/11/2018 11:52

First Amendment, video games, and virtual reality training  273 their mental operations more generally.130 In physical-world activity, thought is generally expressed only in action, and government can plausibly claim it is aiming to regulate the action rather than individuals’ intellectual life. In VR, by contrast, this argument is harder to make: Where there is little physical activity taking place (other than sitting in a chair with a head-mounted display over one’s eyes), or sitting or standing before a CAVE projection, there is less for the government to regulate apart from the imagery, sound, and tactile sensations that the VR user experiences.131 Whether this distinction between VR and brick and mortar training justifies different First Amendment treatment for each is unclear. My purpose in this chapter has been primarily to raise such questions and sketch possible answers, rather than to defend specific conclusions about how to resolve them.

4. CONCLUSION When many people think of virtual reality, they tend to think of its most familiar application: to provide a new, immersive way to play video games. The most familiar VR products—Oculus Rift, HTC Vive, and PlayStation—are video game consoles. There are an increasing number of virtual reality arcades in movie theaters and malls, allowing individuals who do not own their own consoles to challenge themselves in VR adventures. But virtual reality is also increasingly finding a role in other, more serious endeavors: It is being used to train soldiers for battle, athletes for games, and physicians for complex surgeries. It is being used in psychotherapy to treat PTSD and anxiety. While it is very likely that the VR video games will receive the same First Amendment protection that the Supreme Court (and other federal courts) have made clear applies to their less immersive equivalents, it is less clear what courts will say about the constitutional status of virtual reality training. The latter is intended not only to provide, and immerse an audience within, a vivid story. It is intended to mimic—and improve upon—physical training methods (such as use of a shooting range, or physical flight simulators) which seem unlikely to count as First Amendment speech themselves. Nonetheless, VR training may well be First Amendment speech, or some other activity that is insulated from government control by constitutional liberty protection. I have here explored how courts or scholars might reach this conclusion using methods they have previously considered for establishing the boundaries of First Amendment coverage, and also by understanding virtual reality training (or some subset of it) as falling within one of more of three categories that might qualify for First Amendment or other constitutional protection: (1) as a variant of the same immersive storytelling one finds in video games; (2) as a means of obtaining knowledge about the world (both with vision and other senses); and (3) as a means individuals use to reshape the operation of their minds. Future work on this topic might look at these possible arguments more closely, considering them in light of the improvements that will undoubtedly arise as engineers develop and enhance virtual reality training technology.

  See text accompanying notes 90–91 supra.   Id.

130 131

BARFIELD PRINT.indd 273

13/11/2018 11:52

274  Research handbook on the law of virtual and augmented reality

SOURCES Am. Amusement Mach. Ass’n v. Kendrick, 244 F.3d 572, 579–80 (7th Cir. 2001). Brandenburg v. Ohio, 395 U.S. 444 (1969). Brown v. Entm’t Merchants Ass’n, 564 U.S. 786 (2011). Ezell v. City of Chicago, 846 F.3d 888, 892 (7th Cir. 2017). Holder v. Humanitarian Law Project, 561 U.S. 1, 2 (2010). Hurley v. Irish-Am. Gay, Lesbian, Bisexual Group of Boston, 515 U.S. 557, 569–70 (1995). Joseph Burstyn, Inc. v. Wilson, 343 U.S. 495, 500–501 (1952). Lawrence v. Texas, 539 U.S. at 574 (2003). Meyer v. Nebraska, 262 U.S. 390 (1923). Mutual Film Corp. v. Industrial Comm. of Ohio, 236 U.S. 230, 244 (1915). Palko v. Connecticut, 302 U.S. 319, 326–7 (1937). Paris Adult Theater I v. Slaton, 413 U.S. 49, 67 (1973). Pierce v. Society of the Sisters of the Holy Names of Jesus and Mary, 268 U.S. 510 (1925). Rice v. Paladin Enterprises, Inc., 128 F.3d 233 (1997). Riggins v. Nevada, 504 U.S. 127 (1992). Sell v. United States, 539 U.S. 166 (2003). Spence v. Washington, 418 U.S. 405, 410–11 (1974). Stanley v. Georgia, 394 U.S. 557, 566 (1969). United States v. Schwimmer, 279 U.S. 644, 654–5 (1929) (Holmes J dissenting). W. Va. State Bd. of Educ. v. Barnette, 319 U.S. 624, 637 (1943). Washington v. Harper, 494 U.S. 210 (1990); Wooley v. Maynard, 430 U.S. 705, 714 (1977) Zemel v. Rusk, 381 U.S. 1 (1965).

BARFIELD PRINT.indd 274

13/11/2018 11:52

9.  Virtual reality, haptics, and First Amendment protection for sexual sensation Brooke Lewis*

1. INTRODUCTION Imagine the last time you watched a scary movie. When the music intensified as a shadow lurked around the corner, you likely felt your heart race, your muscles tighten, and your breath quicken. Or maybe you have watched your favorite stand-up comic. As he or she told all of your favorite jokes, your body convulsed with laughter. And I am sure we all remember a time when a sad song made us cry. Needless to say, much of the way in which we experience the information communicated to us has a very physical component. In fact, many of the thoughts and feelings we experience are inseparable from the physical sensations that accompany them. The movies, performances, and songs imagined above all represent communications protected in the US by the First Amendment.1 The First Amendment prohibits the government from “abridging the freedom of speech,”2 and in doing so provides constitutional protection to communications that effectuate physical responses and sensations. Because physical sensation is integral to how we communicate, it is no surprise that modern media now uses “haptic technology” to enhance the way we express ourselves. The term “haptic” refers generally to anything “relating to touch” and, in the context of digital technology, refers specifically “to tactile sensations and the sense of touch as a method of interacting with computers and electronic devices.”3 While haptic technology is not exactly new, it is advancing in ways that are likely to transform several media platforms. For instance, those building the future of virtual reality are hoping to use haptics to provide users with sensations of touch that are coordinated with visual and audio experiences. In fact, virtual and augmented reality media that engages our sexuality has

*  The author would like to thank Ashley Lewis and Professors Ellen P. Goodman, Alan K. Chen, Marc Jonathan Blitz, Woodrow Barfield, and James Weinstein for their helpful comments and suggestions. 1   Joseph Burstyn, Inc., 343 U.S. 495 (1952) (affording First Amendment protection to film); Ward v. Rock Against Racism, 491 U.S. 781 (1989) (finding that music is protected by the First Amendment); Edelman v. Croonquist, No. 09-1938, 2010 WL 1816180 (D.N.J. May 4, 2010) (finding that standup comedian Sunda Croonquist’s jokes about her mother-in-law being a racist were protected by the First Amendment and not defamatory). 2   U.S. Const. amend. I. 3  Dictionary.Com, http://dictionary.reference.com/browse/haptic (last visited Feb. 24, 2016). In many ways, “haptics is to touch what optics is to sight.” Joel N. Shurkin, Haptic Technology Makes You Feel Things That Aren’t There, Inside Science (Dec. 18, 2014) www.insidescience.org/ content/haptic-technology-makes-you-feel-things-aren%E2%80%99t-there/2421.

275

BARFIELD PRINT.indd 275

13/11/2018 11:52

276  Research handbook on the law of virtual and augmented reality already gravitated toward some of the most sophisticated forms of haptic technology. One example is online pornography’s use of “teledildonics.” “Teledildonics” does not have a single, precise definition but can generally be understood as “the . . . use of virtual reality to mediate sexual interaction between computer users operating in different places.”4 Significantly, while traditional online pornography allows a customer to both see and hear portrayals of sex, pornography that uses teledildonics allows that customer to experience tactile simulations of sex as well. Today, online pornography has integrated the use of teledildonics into standard pornographic films as well as live online pornographic webcam performances.5 In the future, it may use such tactile simulations to make the immersive experiences generated by virtual reality seem even more like sexual encounters in the physical world. This chapter explores whether the First Amendment protects not only the visual and auditory components of pornographic and other films, or more recently VR and AR pornography, but also the use of haptic technology to generate sensations of touch.6 To do so, it begins with, and focuses on, a particular use of haptics—namely pornography’s use of haptic technology in teledildonics—and asks if it should receive protection as “expressive conduct” under the First Amendment. The First Amendment protects conduct as “speech” when conduct contains an expressive component.7 For example, the First Amendment protects pornographic films as “expressive conduct” because, while the performance of sexual acts is pure conduct, the production of film is presumed to be expressive.8 However, although the communicative qualities of film have been sufficient to protect the making of pornography, it is not clear that pornography’s “expressive conduct” classification will extend to its use of teledildonics in virtual and augmented reality, or in other technologies that use haptics to reproduce sexual sensations. 4  Oxford Dictionaries, www.oxforddictionaries.com/us/definition/american_english/­teledil​ donics; see also Erin Mckelle Fischer, What Are Teledildonics, Anyway?, Bustle (July 14, 2015) www.bustle.com/articles/97086-what-are-teledildonics-heres-how-theyre-changing-sexuality-video (providing an overview of teledildonics). 5   Currently leading in this endeavor is Kiiroo, an Amsterdam-based company that produces teledildonics. Founded in 2013, Kiiroo’s two leading products are the “Onyx” and the “Pearl.” The Pearl is a vibrator with a touch-sensitive surface that sends signals to the corresponding Onyx. The Onyx is a male masturbator with contracting rings designed to simulate the acts being performed on the Pearl. Kiiroo has partnered with a pornographic webcam company called “Flirt4Free” to offer paying customers the ability to experience sensations of touch that are synchronized with live pornographic performances. A female pornographic webcam performer on the Flirt4Free website will perform live sex acts on the Pearl that are viewed through a computer by paying male customers. The sensations of those sex acts are then experienced in real time by every male customer wearing an Onyx. 6   Apart from generating sensations of touch, haptic technology can also allow a person to respond to what she feels or otherwise experiences in a virtual world or other environment: It uses force to act upon that environment, such force being generated by an algorithm directing the performance of a physical device. 7   See Spence v. State of Wash., 418 U.S. 405 (1974) (affording First Amendment protection to the display of an American flag with a peace sign on it); Texas v. Johnson, 491 U.S. 397 (1989) (affording First Amendment protection to the act of burning the American flag). 8   People v. Freeman, 46 Cal.3d 419, 426−7 (1988) (finding that the making of pornography, as a nonobscene film, is protected by the First Amendment as expressive conduct).

BARFIELD PRINT.indd 276

13/11/2018 11:52

VR, haptics, First Amendment protection for sexual sensation  277 When using teledildonics, a pornographic actor is performing sexual acts that are instantaneously being “felt” by the audience. Because the performance of sexual acts in the physical world is generally “conduct,” not First Amendment “speech,” establishing the “expressive” component required for First Amendment protection will depend on the communicative value of the sensations of touch being conveyed. In this context, the sensation of touch being conveyed to the audience can be called “sexual sensation” because it is intended to imitate the feeling and effectuate the physical responses associated with sex. Therefore, the central question presented is whether the simulations of sexual sensation produced and distributed to audiences via teledildonics contain the communicative value necessary to make the use of teledildonics count as expressive conduct. Interestingly, understanding the communicative value of sexual sensation raises its own distinct questions about how to distinguish pure conduct from protected expression. On the one hand, the experience of sexual sensation can directly induce the physical responses often used to interpret one’s own thoughts and emotions. In this sense, the distribution of sexual sensation resembles protected expression insofar as it produces the physical responses used to access and understand certain messages. But on the other hand, the communication of sexual sensation also resembles unprotected conduct because it bypasses the cognitive interpretation that would otherwise be required for a person to experience those physical responses. An individual’s cognitive and intellectual engagement with a given expression has traditionally been valued by the First Amendment, and its absence might preclude protection. Therefore, determining whether the use of teledildonics is expressive conduct requires us to understand not only the communicative value of sexual sensation itself, but also the role that its subsequent physical responses might have in either enhancing or undermining communicative value. This chapter focuses on the potential communicative value of sexual sensation to determine whether the use of teledildonics is expressive conduct, but in doing so considers how future work might resolve larger First Amendment questions raised by haptics in virtual reality: Once we understand whether and when the First Amendment might protect emerging uses of haptics to convey sexual sensation, we can begin to explore whether such protection might also extend to other tactile sensations of touch generated in virtual or augmented reality, or other technology. In section 2, this chapter briefly discusses the use of haptics in virtual reality to underscore the broader applicability this analysis may have in other contexts. Next, in section 3, it discusses the traditional framework for understanding when First Amendment protection applies to a particular kind of human activity and when it does not. After that section’s rejection of the traditional framework, section 4 outlines the theoretical justifications for providing First Amendment protection. These theoretical justifications go on to serve as the primary triggers for providing First Amendment protection throughout this analysis. Then section 5 discusses the various arguments in favor of finding sexual sensation to have communicative value; where a given argument would further a theoretical justification, the First Amendment applies. Last, in section 6, this chapter discusses the counterarguments for providing First Amendment protection to sexual sensation. It is here that this chapter resolves some of the problems that physical sensation, and sexual sensation in particular, present for the First Amendment’s distinction between protected speech and unprotected conduct. Namely, it argues that the elements of mass communication employed by virtual pornography will allow the First Amendment to

BARFIELD PRINT.indd 277

13/11/2018 11:52

278  Research handbook on the law of virtual and augmented reality protect the communicative aspects of sexual sensation produced through teledildonics, while continuing to deny protection for real-life sexual encounters.

2.  HAPTIC TECHNOLOGY IN VIRTUAL REALITY Generally speaking, haptic technology, or “haptics,” uses force and tactile feedback to “take advantage of [the] user[’s] sense of touch by applying forces, vibration and motion to the user.”9 In virtual and augmented reality, force feedback can be used to explore remote or virtual objects, while tactile feedback can be used to provide “surface properties such as roughness, smoothness and temperature.”10 Haptics has been used in entertainment for decades and is already a standard feature in both video games and cellphones. Cellphones, for example, deliver vibrations to alert users to a new message, while video game controllers allow players to “feel” the forces exerted against their avatars.11 On the horizon in media technology is the ability to provide audiences with a virtual reality experience, allowing users to interact with computer simulated environments.12 In addition to providing realistic sounds and visuals, virtual reality media platforms may also use haptics to create lifelike sensations of touch. Although this chapter focuses on analyzing teledildonics, a form of haptics used in virtual pornography, several other forms of haptic technology are being integrated into virtual reality media. Haptics use has long been a feature of virtual reality design. In 1967, Frederick Brooks incorporated haptic technology into VR technology in order to let chemists feel as well as view virtual molecules.13 And in the 1980s, NASA made gloves capable of conveying haptics a part of the package of virtual reality technology they used in their VIEW workstation.14 Today’s consumer-oriented VR companies are likewise weaving haptics into their technology. The Oculus Rift, for example, is a virtual reality headset that can be paired with Oculus Touch, a set of handheld motion controllers that use haptics to make the user feel the objects he or she appears to be touching.15 The Oculus Rift can also be paired with haptic gloves,16 foot cuffs, and harnesses.17 Furthermore, a company called AxonVR  9   B. Divya Jyothi & R.V. Krishnaiah, Haptic Technology—A Sense of Touch, International Journal of Science & Research, www.ijsr.net/archive/v2i9/MjMwOTEzMDQ=.pdf. 10   Id. 11   Richard Moss, Haptic Technology: The Next Frontier in Video Games, Wearables, Virtual Reality, and Mobile Electronics, Gizmag (Jan. 15, 2015), www.gizmag.com/haptic-tech​-vr-weara​ bles-games-sightlence/35616/. 12   Anupam Alur, Pratik Shrivastav, & Aditya Jumde, Haptic Technology: A Comprehensive Review of its Applications and Future Prospects, International Journal of Computer Science & Information Technologies, www.ijcsit.com/docs/Volume%205/vol5issue05/ijcsit2014050508.pdf. 13   Mario Gutierrez, F. Vexo, & Daniel Thalmann, Stepping into Virtual Reality 6 (Spring, 2008). 14   Id. 15   Josh Constine, Oculus Preview “Oculus Touch” Handheld Motion-Tracking Haptic Controllers, Tech Crunch (June 11, 2015), https://techcrunch.com/2015/06/11/oculus-touch/ 16   Josh Constine, Zuckerberg Shows Off Oculus for Typing in VR, Tech Crunch (Feb. 9, 2017), https://techcrunch.com/2017/02/09/oculus-gloves/. 17   Adi Robertson, We’ve Seen Virtual Reality. It’s Time to Touch It, The Verge (June 13, 2014), www.theverge.com/2014/6/13/5805628/at-e3-virtual-reality-goes-beyond-goggles.

BARFIELD PRINT.indd 278

13/11/2018 11:52

VR, haptics, First Amendment protection for sexual sensation  279 is marketing a full body suit that employs haptic technology to immerse the user’s entire body in virtual reality.18 With these ambitious attempts to incorporate haptics into virtual reality, it is likely that the same or similar First Amendment questions raised by the use of teledildonics in virtual pornography will arise in other forms of virtual reality too. Accordingly, although this chapter’s analysis is factually specific to the use of teledildonics and the communicative value of sexual sensation, the framework outlined below may be useful in determining if other sensations produced through haptics communicate ­messages that warrant First Amendment protection.

3. THE TRADITIONAL FRAMEWORK To determine whether activity counts as speech, courts have mostly relied on the Supreme Court’s Spence test, asking whether the activity contains a “particularized message” likely to be understood by the audience in the context in which it occurs. However, the “particularized message” requirement of the Spence test, both before and after its liberalization by courts, provides insufficient guidance for analyzing the First Amendment protection afforded to nonpropositional expressions, such as sexual sensation. Therefore, in finding the traditional framework for affording First Amendment protection unhelpful, this chapter adopts a nontraditional approach better suited for nonverbal, nonpropositional forms of expression. 3.1  Categorical Inclusions From time to time, the Supreme Court announces what is essentially a “categorical inclusion,”19 meaning that “when the medium is used, the communicator is presumptively a speaker.”20 In other words, the medium so regularly communicates messages protected by the First Amendment that its use categorically triggers some level of constitutional protection. Categorical inclusions can be thought of as “where speech happens.”21 When the Supreme Court is confronted with a new medium, the Court must determine whether that medium has the characteristics necessary to categorically implicate the First Amendment.22 In some instances, the Court has found that over time a medium can develop the characteristics needed to be recognized as a categorical inclusion.23 An examination of the various categorical inclusion cases can shed light on the characteristics needed for a particular medium to categorically implicate the First Amendment. In 1915, the Supreme Court first confronted film as a medium under the First Amendment in Mutual Film Corporation v. Industrial Commission of Ohio.24 In Mutual

  Axon VR, http://axonvr.com/#powered-by-haptx.   The Supreme Court has also created categorical exclusions to the First Amendment. These exclusions include incitement, false statements of fact, obscenity, and child pornography. Tim Wu, Machine Speech, 161 U. Pa. L. Rev. 1495, 1509−10 (2013). 20   Id. at 1511. 21   Id. 22   Id. at 1512. 23   Id. 24   236 U.S. 230 (1915). 18 19

BARFIELD PRINT.indd 279

13/11/2018 11:52

280  Research handbook on the law of virtual and augmented reality Film, the Supreme Court denied First Amendment protection to film, finding that “the exhibition of moving pictures is a business pure and simple, originated and conducted for profit, like other spectacles, not to be regarded, nor intended to be regarded . . . as part of the press of the country, or as organs of public opinion.”25 However, in 1952 the Supreme Court reversed itself in Joseph Burstyn, Inc. v. Wilson and identified film as a medium categorically entitled to First Amendment protection. The Court stated: It cannot be doubted that motion pictures are a significant medium for the communication of ideas. They may affect public attitudes and behavior in a variety of ways, ranging from direct espousal of a political or social doctrine to the subtle shaping of thought which characterizes all artistic expression. The importance of motion pictures as an organ of public opinion is not lessened by the fact that they are designed to entertain as well as inform.26

In this way, over the course of these 37 years, the Supreme Court observed film as having developed the ability to “affect public attitudes and behavior” in a way that makes it a “significant medium for the communication of ideas.”27 Many years later, the Supreme Court observed and responded to similar developments in the world of video games. In 1982, in America’s Best Family Showplace Corp. v. City of New York, the Eastern District of New York denied First Amendment protection to video games, reasoning that “a video game, like a pinball game, a game of chess, or a game of baseball, is pure entertainment with no informational element.”28 By 2011, however, the Supreme Court had afforded First Amendment protection, finding that “[l]ike the protected books, plays, and movies that preceded them, video games communicate ideas—and even social messages—through features distinctive to the medium (such as the player’s interaction with the virtual world).”29 Because the Court observed   Id. at 244.   343 U.S. 495, 501 (1952). However, it is important to note that categorically providing a medium with protection does not by itself determine the contours of the government’s ability to regulate. See Burstyn, 343 U.S. at 502–3 (“[i]t does not follow that the Constitution requires absolute freedom to exhibit every motion picture of every kind at all times and all places . . . Nor does it follow that motion pictures are necessarily subject to the precise rules governing any other particular method of expression. Each method tends to present its own peculiar problems”); see also Red Lion Broad. Co. v. FCC, 395 U.S. 367, 368 (1969) (“Although broadcasting is clearly a medium affected by a First Amendment interest, differences in the characteristics of new media justify the differences in the First Amendment standards applied to them”). 27   Burstyn, 343 U.S. at 502–3. However, as Tim Wu points out, although “[t]he earliest commercial films were silent, were usually under ten minutes, and rarely had plot or characters,” by 1915 the film industry had already transformed in a way that clearly communicated messages and ideas. Therefore, the Mutual Film decision, even in its own time, was indefensible. Wu, supra note 19, at 1512 n.80. 28   Am.’s Best Family Showplace Corp. v. City of New York, 536 F. Supp. 170, 174 (E.D.N.Y. 1982). 29   Brown v. Entertainment Merchants Ass’n, 131 S. Ct. 2729, 2733 (2011). By 2011 video games had transformed substantially, with many using “the full range of literary devices found in film and novels, as well as other devices that more traditional media cannot mimic.” Wu, supra note 19, at 1513–14. For example, the Japanese video game Ico features a child born with horns who has been exiled from his village and placed in a prison from which he must escape. Id. Other well-known video games, such as Halo, which features a “plot not unlike a science fiction movie,” and Grand Theft Auto, which “communicates a kind of humorous glorification of the low-level criminal that is not unlike the films of Quentin Tarantino,” utilize traditional literary devices. Id. 25 26

BARFIELD PRINT.indd 280

13/11/2018 11:52

VR, haptics, First Amendment protection for sexual sensation  281 that video games developed the ability to transmit ideas, video games were recognized as a medium protected by the First Amendment.30 As has been made clear by Burstyn and Brown, whether or not a new medium will be made a categorical inclusion turns on the character of the “ideas” or “information” purportedly being communicated. Are teledildonics, when distributing sexual sensation to audiences, providing a platform for the exchange of information?31 Do teledildonics allow the performer to communicate a story or otherwise affect public opinion through the transmission of sexual sensation? The answers to these questions require a firm understanding of the communicative value of sexual sensation. Because the communicative value of sexual sensation has not yet been explored, and the exact process by which a medium becomes a categorical inclusion is not well defined, we do not have enough information to reliably predict whether or not a court would make teledildonics a medium that categorically receives First Amendment protection. 3.2 The Spence Test In cases not involving a medium that has been categorically afforded First Amendment protection, determining whether or not the First Amendment applies becomes more complicated. Spence v. Washington is,32 however, often regarded as providing the “test” for when First Amendment protection applies, particularly in situations involving conduct with an arguably expressive component.33 Because the use of teledildonics undeniably involves conduct normally considered to be nonexpressive when undertaken in the real world, namely the nonverbal sensation which accompanies sexual contact, the relevant inquiry is whether the use of teledildonics qualifies as expressive conduct under the Spence test. On May 10, 1970, a college student named Harold Spence hung an upside down American flag with a large peace sign affixed to both sides outside the window of his apartment.34 As a result of this display, which appeared to be an act of protest against the Cambodian war and the killings at Kent State University,35 Spence was prosecuted and convicted under a Washington statute that prohibited “the exhibition of a United States flag to which is attached or superimposed figures, symbols, or extraneous material.”36 However, the Supreme Court reversed Spence’s conviction on First Amendment grounds.

30   Id.; see also Reno v. Am. Civil Liberties Union, 521 U.S. 844, 853 (1997) (holding that the internet receives First Amendment protection because “[t]he web is . . . comparable, from the reader’s viewpoint, to both a vast library including millions of readily available and indexed publications and a sprawling mall offering goods and services. From the publishers’ point of view, it constitutes a vast platform from which to address and hear from a worldwide audience of millions of readers, viewers, researchers, and buyers”). 31   See M. Lombard and M. Jones, Presence and Sexuality, Presence Conference, October 13–15, 2004. Available at: https://ispr.info/presence-conferences/previous-conferences/presence-2004/. 32   418 U.S. 405 (1974). 33   See Wu, supra note 19, at 1510 (“Spence v. Washington serves a general guide to the line between speech and conduct”). 34  Spence, supra note 7, at 405. 35   Id. at 410. 36   Id. at 405.

BARFIELD PRINT.indd 281

13/11/2018 11:52

282  Research handbook on the law of virtual and augmented reality In doing so, the Court stated that First Amendment protection applies where “[a]n intent to convey a particularized message [is] present, and in the surrounding circumstances the likelihood [is] great that the message [will] be understood by others.”37 This language has been referred to as the “Spence test.” In sum, the Spence test states that the First Amendment applies where “1) the speaker-actor intends for the conduct to express a particularized message; and 2) that message would be understood by others.”38 Following the Spence decision, however, the “particularized message” requirement of the Spence test has been substantially liberalized, and in some jurisdictions eliminated. In 1995 the Supreme Court decided Hurley v. Irish-American Gay Lesbian and Bisexual Group of Boston, a case involving the free speech rights associated with the selection of parade participants.39 In Hurley the Court found that the petitioner’s choice to exclude the Irish-American Gay Lesbian and Bisexual Group of Boston from the St Patrick’s Day Evacuation Parade was protected speech under the First Amendment.40 In reaching its decision, the Court reasoned that “a narrow, succinctly articulable message is not a condition of constitutional protection, which if confined to expressions conveying a ‘particularized message’, would never reach the unquestionably shielded painting of Jackson Pollock, music of Arnold Schoenberg, or Jabberwocky verse of Lewis Carroll.”41 Based on this language, many lower courts have interpreted the Hurley decision to have eliminated, or at the very least substantially liberalized, the particularized message requirement of the Spence test.42 For example, in Troster v. Pennsylvania State Department of Corrections, the Third Circuit found that the particularized message requirement of the Spence test was no longer necessary for First Amendment protection.43 Instead, the court focused its attention on the language in Spence which stated that to receive First Amendment protection, the

37   Id. at 410−11. This approach was later affirmed in Texas v. Johnson where the court afforded First Amendment protection to the act of burning an American flag at a Republican convention because it conveyed a particularized message that was likely to be understood by others. See Johnson, supra note 7, at 420 (finding that “Johnson was convicted for engaging in expressive conduct”). 38   Angelica M. Sinople, Note, “No Saggy Pants”: A Review of the First Amendment Issues Presented by the State’s Regulation of Fashion in Public Streets, 113 Penn. St. L. Rev. 329, 339 (2008). For the purposes of this analysis, we can easily dispense with the “likely to be understood by others” requirement. Factors which a court will consider when deciding if the message is likely to be understood include “the timing of the conduct, the political or social conditions surrounding the conduct, the viewer’s personal knowledge, and, perhaps, the social position of the speaker-actor.” Id. at 347. Based on these factors, the sexual sensation received by an end user, if communicative, is likely to be understood. 39   515 U.S. 557 (1995). 40   Id. 41   Id. at 567. 42   See James M. McGoldrick, Jr, Symbolic Speech: A Message From Mind to Mind, 61 Okla. L. Rev. 1, 72 (2008); Sinople, Saggy Pants, 113 Penn St. L. Rev. at 341 (discussing how the 3rd and 11th Circuits have liberalized the particularized message requirement). 43   65 F.3d 1086, 1090 (3d Cir. 1995) (stating that the message requirement was “no longer viable” after Hurley); see also Tenafly Eruv. Ass’n v. Borough of Tenafly, 309 F.3d 144, 160 (3d Cir. 2002) (“Spence . . . set signposts rather than requirements, and that its two factors can no longer be viewed as its only criteria”).

BARFIELD PRINT.indd 282

13/11/2018 11:52

VR, haptics, First Amendment protection for sexual sensation  283 activity need only be “sufficiently imbued with elements of communication.”44 Pursuant to Spence, in determining whether an activity is “sufficiently imbued with elements of communication,” the Third Circuit considered “the nature of [the] activity, combined with the factual context and environment in which it was undertaken.”45 Furthermore, the Eleventh Circuit, while not going as far as elimination, substantially liberalized the Spence test in finding that while the activity still needs to communicate a message, the message need not be particularized.46 In Holloman ex. rel. Holloman v. Harland the Eleventh Circuit stated that Hurley “liberalized” the particularized message requirement, meaning that the court must now only “ask whether the reasonable person would interpret it as some sort of message, not whether an observer would necessarily infer a specific message.”47 Together, the Troster and Holloman decisions are representative of how some circuit court interpretations of the Hurley decision have eliminated or substantially liberalized the particularized message requirement of the Spence test. 3.3  Rejecting the Traditional Framework With the particularized message requirement of the Spence test being eliminated or substantially liberalized by the Hurley decision, it is difficult to understand when conduct communicates a message that is protected by the First Amendment. However, the Hurley decision and its various interpretations by lower courts correctly identify and address what was always inherently wrong with the Spence test; that is, much of what the First Amendment seeks to protect does not have what would be considered a “particularized message.” As Hurley points out, if First Amendment protection was given only to communications that conveyed a particularized message, the First Amendment would not protect the “painting of Jackson Pollock, music of Arnold Schoenberg, or Jabberwocky verse of Lewis Carroll.”48 Courts and commentators unanimously agree that these types of artistic expression, although lacking a particularized message, are “unquestionably shielded” by the First Amendment.49 Yet, neither the Spence test, nor any court applying it, has been able to fully and clearly explain why.50 The lack of explanation for providing First Amendment protection to nonverbal, nonpropositional artistic expression is problematic for the First Amendment analysis of

  Id. (citing Spence, 418 U.S. at 409–10).   Id. 46   Holloman ex. rel. Holloman v. Harland, 370 F.3d 1252, 1270 (2004). 47   Id.; see also Blau v. Fort Thomas Pub. Sch. Dist., 401 F.3d 381, 388 (6th Cir. 2005) (holding that the particularized message requirement still stands but that the particularized message need not be “succinctly articulable”). 48   Hurley, 515 U.S. at 567. 49   Id. 50  In National Endowment of the Arts v. Finley, the artists argued that art is entitled to First Amendment protection specifically because it is ambiguous and does not have any particular message. However, the Supreme Court declined to analyze why art, in the absence of a particular message, should receive First Amendment protection and presumed that the art was expressive. See Brief for Respondent at 14, National Endowment for the Arts v. Finley, 524 U.S. 569 (1998) (No 97−371). 44 45

BARFIELD PRINT.indd 283

13/11/2018 11:52

284  Research handbook on the law of virtual and augmented reality sexual sensation—conveyed through teledildonics or another medium—because sexual sensation also does not likely contain any particularized message. Nonetheless, just like abstract art and instrumental music, both of which have First Amendment value because of their appeals to “the part of human existence involving our senses, intuition, feelings, and vision,”51 sexual sensation may provide a type of experience that furthers the purposes of the First Amendment. In response to the lack of explanation for why some nonverbal, nonpropositional expression receives First Amendment protection, many scholars have sought answers.52 For example, Alan K. Chen, author of Instrumental Music and the First Amendment, has endeavored to understand why instrumental music, although both nonverbal and nonpropositional, receives First Amendment protection without question.53 Because instrumental music and sexual sensation would both be regarded as nonverbal, nonpropositional forms of communication, Chen’s approach to understanding the First Amendment protection afforded to instrumental music is helpful in analyzing whether sexual sensation should be afforded First Amendment protection. Adopting an approach similar to Chen’s, I will first outline the theoretical justifications for providing First Amendment protection, drawing heavily upon Chen’s analysis of those justifications.54 In place of the Spence test, these theoretical justifications will be used to understand when the First Amendment is implicated. Next, I will discuss the arguments for finding sexual sensation to have communicative value. I will first revisit the Spence test, exploring possible arguments that acts which generate sexual sensation (whether technologically sophisticated acts, such as use of haptic devices, or more traditional and primitive means of generating such sensations) might count as expressive because they have—or supplement—particularized messages of the kind the Spence test requires for a nonverbal act to count as “expressive.” However, after finding such arguments lacking, I will explore arguments for treating sexual sensation as expressive even where it lacks any particularized message, or any close link to such a particularized message. In both its Spence test-based analysis, and its exploration of alternative theories, this chapter will analyze how the potentially communicative aspects of sexual sensation fit within the theoretical justifications for providing First Amendment protection.

  Edward J. Eberle, Art as Speech, 11 U. Pa. J. L. & Soc. Change 1, 4 (2007−08).   See, e.g., Eberle, supra note 51; Joseph Blocher, Nonsense and the Freedom of Speech: What Meaning Means for the First Amendment, 63 Duke L.J. 1423 (2014); Sheldon H. Nahmod, Artistic Expression and the Aesthetic Theory: The Beautiful, the Sublime and the First Amendment, 1987 Wis. L. Rev. 221 (1987). 53   Alan K. Chen, Instrumental Music and the First Amendment, 66 Hastings L.J. 381, 423 (2015) (stating that a nonpropositional expression is one that “does not intend (or does not always intend) to convey, nor can it be understood to convey, a particular, identifiable message”). 54   See generally id. (outlining the theoretical justifications for First Amendment protection and then analyzing how the arguments for the communicative value of instrumental music do and do not further those justifications). 51 52

BARFIELD PRINT.indd 284

13/11/2018 11:52

VR, haptics, First Amendment protection for sexual sensation  285

4. THE THEORETICAL JUSTIFICATIONS 4.1 Democracy One theoretical justification for providing First Amendment protection is that such protection promotes, or enables, the deliberation necessary for democratic self-governance (the “democracy” or “democracy-based” justification). Alan Chen writes that this is “Perhaps the most widely discussed and accepted theory underlying the constitutional protection of speech.”55 Alexander Meiklejohn is its most well-known proponent. To Meiklejohn, the democracy-based justification initially meant that the First Amendment only protects speech that is explicitly political.56 However, in response to criticisms that such a narrow construction of the First Amendment would exclude most art and literature, Meiklejohn later clarified that he believed the arts are protected by the First Amendment because they help generate qualities that are necessary to intelligent selfgovernment: “knowledge, intelligence, sensitivity to human values,” and “capacity for sane and objective judgment.”57 Defending a view akin to Meiklejohn’s earlier, narrower view of the First Amendment’s scope, Robert Bork vigorously argued in 1971 that only expressly political speech should be protected by the First Amendment.58 Bork contended that “[t]here is no basis for judicial intervention to protect any other form of expression, be it scientific, literary, or that variety of expression we call obscene or pornographic.”59 Notably, Bork argued that limiting First Amendment protection to only expressly political speech was the only sure way to achieve the First Amendment’s principal objective of “the discovery and spread of political truth.”60 Bork did not believe that artistic expression promoted democracy and considered it to fall outside the First Amendment’s scope.61 Modern First Amendment theorists continue to advocate for the democracy-based justification but, unlike Bork, set forth theories that include artistic expression within the scope of the First Amendment, even where it lacks expressly political content.62 Robert Post, for example, argues for a broad conception of the democracy justification—one that protects not only communications that inform decisions about candidates and issues,

 Chen, supra note 53, at 403.   Joseph Blocher, Nonsense and the Freedom of Speech: What Meaning Means for the First Amendment, 63 Duke L.J. 1423, 1453 (2014) (citing Alexander Meiklejohn, Free Speech and its Relation to Self-Government 94 (1948) (“The guarantee given by the First Amendment is not, then, assured to all speaking. It is assured only to speech which bears, directly or indirectly, upon issues with which voters have to deal—only, therefore, to the considerations of public interest”)). 57   Alexander Meiklejohn, The First Amendment Is an Absolute, 1961 Sup. Ct. Rev. 245, 255 (1961) (stating that “the First Amendment does not protect ‘a freedom to speak.’ It protects those activities of thought and communication by which we ‘govern.’ It is [thus] concerned not with a private right, but with a public power, a governmental responsibility.”). 58   Robert H. Bork, Neutral Principles and Some First Amendment Problems, 47 Ind. L.J. 1 (1971)). 59   Id. at 20. 60   Id. at 25. 61   Id. at 29. 62   Id. 55 56

BARFIELD PRINT.indd 285

13/11/2018 11:52

286  Research handbook on the law of virtual and augmented reality but also the general process of forming public opinions.63 In contrast to Meiklejohn, who argued for protecting speech on certain subjects on the grounds it was necessary for self-government, Post argues that the First Amendment must protect public discourse and not any specific topics of such discourse: “Public discourse,” he writes, “merits unique constitutional protection because it is the process through which the democratic ‘self,’ the agent of self-government, is itself constituted.”64 This requires protecting all “speech acts and media of communication” crucial to formation of public opinion. Artistic expression as well as expression on political topics takes place in public discourse, and receives the protection that the First Amendment provides for it.65 Post argues that in this way, the democracy-based justification allows the First Amendment to protect artistic expression.66 4.2 Truth A second theoretical justification for providing First Amendment protection is to further the search for truth (the “Truth” or “truth-based” justification). Although the truth-based justification overlaps with the democracy-based justification to some extent, the truth-based justification focuses on “the idea that expressive freedom promotes the search for more general truths beyond the world of politics and governing.”67 These truths might include “religious or spiritual truths, to the extent they inform personal and collective understandings of the universe” and “matters of science and morality . . . in which unregulated discourse is important to produce a better truth through the evolution of ideas.”68 The truth-based justification has antecedents in the political theory of writers such as John Milton and John Stuart Mill. Mill, for example, argued that censorship of speech is unjustified because “it is robbing the human race; posterity as well as the existing generation” of the chance to find truth from the free exchange of ideas.69 The truthbased justification also received doctrinal support in Justice Holmes’s dissent in Abrams v. United States. In Abrams, the Supreme Court upheld the defendants’ convictions under the Espionage Act for, among other things, conspiring to “unlawfully print, write, and publish . . . disloyal, scurrilous and abusive language about the form of government of the United States.”70 In his dissent, Justice Holmes disagreed with the majority decision’s conclusion that the convictions were constitutional. He stated that “the ultimate good   Robert Post, Participatory Democracy and Free Speech, 97 Va. L. Rev. 477, 483 (2011).   Robert Post, Meiklejohn’s Mistake: Individual Autonomy and the Reform of Public Discourse, 64 U. Colo. L. Rev. 1109, 1128 (1993). 65  Post, Participatory, Democracy and Free Speech, supra note 63 at 483. 66   Id. at 486. 67  Chen, supra note 53, at 408. 68   Id. at 409. 69   John Stuart Mill, On Liberty and Other Essays 17, 21 (John Gray ed., 1991). 70   Abrams v. United States, 250 U.S. 616, 617 (1919). The defendants were also convicted of conspiring to publish “language ‘intended to incite, provoke, and encourage resistance to the United States in said war’”, and conspiring to “‘unlawfully and willfully, by utterance, writing, ­printing and publication to urge, incite and advocate curtailment of production of things and products, to wit, ordnance and ammunition, necessary and essential to the prosecution of the war.’” Id. 63 64

BARFIELD PRINT.indd 286

13/11/2018 11:52

VR, haptics, First Amendment protection for sexual sensation  287 desired is better reached by free trade in ideas—that the best test of truth is the power of the thought to get itself accepted in the competition of the market.”71 Here, Justice Holmes essentially uses the truth-based justification for First Amendment protection to find that the convictions violated the Amendment. 4.3 Autonomy A third theoretical justification for providing First Amendment protection is that it furthers or enables exercise, or development, of individual autonomy (the “autonomy” or “autonomy-based” justification). As Alan Chen points out, theorists have conceptualized autonomy in a number of ways.72 For example, Thomas Scanlon “examines the notion of autonomy in terms of the individual’s freedom to engage in self-determination.”73 To Scanlon, “[s]tate interference with the ability to experience the universe of competing ideas seriously compromises such autonomy.”74 Advocating for an even broader version of the autonomy-based justification, Martin Redish argued that free speech values were designed to achieve “individual self-realization.”75 According to Redish, self-realization can “refer either to development of the individual’s powers and abilities . . . or to the individual’s control of his or her own destiny through making life-affecting decisions.”76 And Edwin Baker argued that the autonomy-based justification was grounded in the concept of legitimacy, meaning that “a legitimate legal order must fully respect (among other things, e.g. equality) both individual and collective autonomy—both non-political and political speech.”77 Others make similar arguments. Seana Shiffrin argues that much of First Amendment law “only makes sense if the individual mind and the autonomy of its operation . . . are valued and treated with respect.”78 Christina Wells likewise makes a case that important aspects of First Amendment doctrine are based not on furthering just any conception of autonomy, but rather one that is “ric[h] and comple[x]”: “one based on the rights and responsibilities of personhood that is generally associated with Immanuel Kant” and that is “not about atomistic individuals but about social creatures entitled to respect for their dignity.”79 The most prominent example of the autonomy-based justification in First Amendment case law is Stanley v. Georgia. In Stanley, the Supreme Court overturned a man’s conviction for possessing obscene materials inside his own home, even though those materials  Chen, supra note 53, at 630.   Id. at 410. 73   Thomas Scanlon, A Theory of Freedom of Expression, 1 Phil. & Pub. Aff. 204, 216 (1972). 74   Id. 75   Martin H. Redish, The Value of Free Speech, 130 U. Pa. L. Rev. 591, 593 (1982). 76   Redish argued that the autonomy-based justification is a stronger approach to First Amendment protection because the other justifications, such as democracy and truth, are subsumed by self-realization. Id. at 411. 77   C. Edwin Baker, Human Liberty and Freedom of Speech (1992). 78   Seana Shiffrin, A Thinker-Based Approach to Freedom of Speech, 27 Const. Comment. 283, 288 (2011). 79   Christina E. Wells, Reinvigorating Autonomy: Freedom and Responsibility in the Supreme Court’s First Amendment Jurisprudence, 32 Harv. C.R.-C.L. L. Rev. 159, 161 (1997). 71 72

BARFIELD PRINT.indd 287

13/11/2018 11:52

288  Research handbook on the law of virtual and augmented reality could otherwise be regulated outside the home.80 Justice Marshall wrote: “If the First Amendment means anything, it means that the State has no business telling a man, sitting alone in his own house, what books he may read or what films he may watch. Our whole constitutional heritage rebels at the thought of giving government the power to control men’s minds.”81 In this way, the Supreme Court forcefully acknowledged the component of First Amendment protection that allows for freedom of thought and autonomous experiences.

5. THE COMMUNICATIVE VALUE OF SEXUAL SENSATION Having established the theoretical justifications for providing First Amendment protection (democracy, truth, autonomy), we must seek to understand the communicative value of sexual sensation and how those communicative characteristics align with the theoretical justifications. Whether or not sexual sensation implicates the First Amendment depends on both the argument that it has communicative value and the theoretical justification being applied. Where a given argument for communicative value would further a theoretical justification, the First Amendment applies.82 This chapter will outline three main arguments for finding that sexual sensation has communicative value. The first is the cognitive claim, which offers two possible bases for finding that it has such value: (1) that sexual sensation is communicative in enhancing, or in association with, a particularized message; or (2) that sexual sensation is communicative in containing its own particularized message. The two remaining arguments for finding that sexual sensation has communicative value find that it can have such value even in the absence of any particularized message, or any link to such a message. The second argument for finding sexual sensation to have communicative value, for example, is that even where sexual sensation lacks a particularized message of its own, it may be an expression of culture and identity. The third argument is that sexual sensation counts as First Amendment speech (again, even in the absence of a message) because it is an expression of emotion.83 As will be explained, while the enhancement and association components of the cognitive claim are weak and mostly inapplicable to sexual sensation, the two other arguments for finding sexual sensation to have communicative value (on the basis that it expresses culture and identity or emotion) are supported by at least one of the theoretical justifications. Accordingly, this analysis concludes there are arguments which could successfully show that sexual sensation has communicative value entitled to First Amendment protection.

  Stanley v. Georgia, 394 U.S. 557 (1969).   Id. at 565. 82   See supra note 53 (citing Chen’s analysis). 83   Chen makes similar arguments for finding that instrumental music is communicative. See Chen, supra note 53, at 417−31 (analyzing instrumental music as “evocative/associative of cognitive thought,” “as enhancing other communicative messages,” as “cognition,” “as an expression of culture, religion, and other social values,” and “as an expression of emotion”). 80 81

BARFIELD PRINT.indd 288

13/11/2018 11:52

VR, haptics, First Amendment protection for sexual sensation  289 5.1  Cognitive Claims 5.1.1  Sexual sensation associates with or enhances a particularized message Even if sexual sensation does not itself contain a particularized message, it may be communicative by either associating with or enhancing a particularized message. For example, some nonverbal, nonpropositional forms of expression can be communicative by associating with particularized messages. In provocative or controversial titles, for example, the expressive effect stems not just from the title’s propositional content, but from its ability to elicit a strong emotional reaction.84 Nonverbal, nonpropositional expressions can also have an enhancing effect by combining with particularized messages to deliver the information with more force or more emotion than would otherwise have been possible.85 However, neither the enhancement nor the association claims are likely to be successful at establishing the communicative value of sexual sensation. To begin, enhancement and association claims are inherently weak; that is, if an expression’s communicative value is only derivative of some verbal, propositional form of communication, then the expression’s independent communicative value may be de minimis or nonexistent.86 When sexual sensation occurs alone without any accompanying propositional content, it will be undeserving of First Amendment protection (at least on this account of why sexual sensation may have such protection). Moreover, beyond this inherent weakness, enhancement and association arguments face an additional challenge when applied to sexual sensation because pornography likely does not present a particularized message for sexual sensation to enhance or associate with. When pornography is nonobscene under the Supreme Court’s test for obscenity,87 it receives First Amendment protection not because of the character of its message, but because it is a nonobscene film and therefore categorically receives First Amendment protection.88 In other words, pornography does not receive First Amendment protection because of its sexual content, but in spite of it.89 Given this fact, it would be difficult  Chen, supra note 53, at 418.   For example, “the melodic or rhythmic components of musical expression may sometimes interact in meaningful ways with lyrics or other verbal messages, combining to express something that is both greater than and distinctive from, lyrics alone.” Id. at 421. 86   Cf. id. at 419–20, 422 (explaining the weaknesses of the enhancement and association claims when applied to instrumental music). 87   The Supreme Court found in Miller v. California that obscene speech is that which, “taken as a whole, appeal[s] to the prurient interest in sex, which portray[s] sexual conduct in a patently offensive way, and which, taken as a whole, do[es] not have serious literary, artistic, political, or scientific value.” 93 S. Ct. 2607, 2615 (1991). 88   See, e.g., Burstyn, 343 U.S. at 502 (“[W]e conclude that expression by means of motion pictures is included within the free speech and free press guaranty of the First and Fourteenth Amendments”). However, many scholars argue that pornography should receive less protection than other types of expressions. See, e.g., Christopher M. Shultz, Content-Based Restrictions on Free Expression: Reevaluating the High versus the Low Value Speech Distinction, 41 Ariz. L. Rev. 573 (1999) (noting how Cass Sunstein “argues that certain kinds of pornography should be, as obscenity already is, classified as low-value speech so that government regulation of its content may be facilitated”). 89   See Burstyn, 343 U.S. 495 at 502; Ashcroft v. Free Speech Coalition, 535 U.S. 234, 248 84 85

BARFIELD PRINT.indd 289

13/11/2018 11:52

290  Research handbook on the law of virtual and augmented reality to use an enhancement or an association claim to establish the communicative value of sexual sensation. 5.1.2  Sexual sensation, itself, contains a particularized message All other areas of this chapter presume that sexual sensation is most likely a nonverbal, nonpropositional form of expression, akin to instrumental music or abstract art. Thus, it remains true that courts are most likely to find that sexual sensation is nonpropositional, meaning it does not contain a particularized message. Nonetheless, there are viable arguments that at least some, if not most, actions that trigger sexual sensation contain a particularized message. Accordingly, before considering how conveyance of sexual sensation might be First Amendment expression even when it lacks a message, or any link to a message, this section will first explore the possibility that sexual sensation is typically communicative because it contains a particularized message.90 For instance, the experience of sexual sensation may more accurately and precisely describe emotions and feelings that remain elusive to language.91 In this sense, sexual sensation contains a particularized message, even if that message cannot be reduced to words. Although it may seem inaccurate to use the word “message” to describe such communications of feelings or physical states, some courts applying the Spence test clearly seem to treat such communications as satisfying the “particularized message” prong of that test. The Seventh Circuit, for example, has suggested that the court might view the generation of “sensual” or “erotic” reactions as particularized messages. In Miller v. Civil City of South Bend, for example, it applied the Spence test to determine if nude dancing implicated the First Amendment. Because the court found that nude dancing was expressive conduct under Spence, we can infer that the message of “eroticism and sensuality” communicated by the dancer was regarded as a particularized message by the Seventh Circuit.92

(2002) (noting “an essential First Amendment rule: The artistic merit of a work does not depend on the presence of a single explicit scene); 2 Ranthkopf’s The Law of Zoning and Planning § 24:10 (4th ed.) (“When embodied in forms of ‘speech’, such as movies, books, and theatrical productions, sexually explicit expression is held to be a part of the ‘speech’ and protected by the First Amendment. As generic vehicles for communicating information, opinions, and artistic ideas in our society, these forms of speech are deemed presumptively protected by the First Amendment despite their sexually explicit content and without inquiry as to whether these forms of ‘speech’ in any particular instance actually convey some ‘message’ or ‘speech’ worthy of First Amendment protection”). 90   Although this chapter rejects the Spence test and its particularized message requirement as the primary method for defining expressive conduct—see supra section 3.3—it nevertheless analyzes whether sexual sensation contains a particularized message, independent of the Spence test, for the purpose of establishing communicative value. 91   As a parallel example, Felix Mendelssohn argues that instrumental music, although most commonly regarded as nonpropositional, can more accurately describe certain feelings and emotions than can language. Chen, supra note 53, at 424. 92   Miller v. Civil City of South Bend, 904 F.2d 1081, 1086−87 (1990), rev’d sub nom. Barnes v. Glen Theater, Inc., 501 U.S. 560 (1991) (explaining that “[t]he dominant theme communicated here by the dancers is an emotional one; it is one of eroticism and sensuality”). Although South Bend was overturned by the Supreme Court in Barnes v. Glen Theatre, the Court in Barnes never denied that a sensual message was a particularized message that implicates the First Amendment.

BARFIELD PRINT.indd 290

13/11/2018 11:52

VR, haptics, First Amendment protection for sexual sensation  291 Again, in City of Erie v. Pap’s A.M.,93 the Seventh Circuit viewed sensuality as containing a distinct message. In this case, the court upheld an ordinance aimed at nude dancing because “the ordinance [did] not attempt to regulate the primary effects of the expression, i.e., the effect on the audience of watching nude erotic dancing.”94 In this way, the Seventh Circuit implied that “the effect” of nude dancing on an audience is a message entitled to First Amendment protection. Moreover, of particular relevance to the analysis of teledildonics is the court’s reference to “the effect” of nude dancing on the audience, a word choice that seems to suggest the physical state of arousal is an audience reaction that has First Amendment value when it occurs in response to an activity such as a dance. If we thus understand sexual sensation to have a particularized message, on the account above, then sexual sensation might be granted First Amendment protection under either the truth or autonomy justification. For instance, the truth justification may afford protection if sexual sensation can give individuals an experience or understanding of their physical or social reality, but does so in a way that cannot be conveyed through language and would otherwise be inaccessible to the audience. Likewise, the autonomy justification may afford protection to the extent that the experience of sexual sensation can contribute to the self-realization and understanding of the person experiencing it. However, even if we accept that sexual sensation communicates a particular message, the message seems to lack the same baseline meaning that exists in other propositional communications. For instance, although the ultimate meaning of a poem may be ambiguous, the propositional meaning of the individual words can most often be universally understood. Therefore, despite the case law suggesting that sexual sensation may have a particularized message, it is difficult to understand how each theoretical First Amendment justification fits without any common understanding of the propositional message.95 5.2  Sexual Sensation Is an Expression of Culture and Identity Even if sexual sensation does not carry a distinct or particularized message, it may still be entitled to First Amendment protection because of its ability to express information about an individual’s identity or culture. For instance, even if a nonverbal, nonpropositional form of expression is not “tied to a specific idea or message [it] can nonetheless be closely associated with cultural, ethnic, religious, and social values.”96 In this way, societal views regarding sexual relations may give the experience of sexual sensation To the contrary, the Court affirmed that the First Amendment did in fact apply. See Barnes v. Glen Theatre, 501 U.S. 560, 564–6 (1991) (stating “that nude dancing of the kind sought to be performed here is expressive conduct within the outer perimeters of the First Amendment, though we view it marginally so”). 93   529 U.S. 277 (2000). 94   Id. at 291. 95   See Chen, supra note 53, at 425 (arguing that “without a common understanding of such thoughts, it is impossible to determine either why the expression is valuable or how the state would know what needs to be suppressed (or for that matter, how a court would know that it deserves First Amendment protection). Thus, it is difficult to see how this type of cognitive understanding could be justified under a democratic self-governance theory”). 96   Id.

BARFIELD PRINT.indd 291

13/11/2018 11:52

292  Research handbook on the law of virtual and augmented reality communicative value. Therefore, this section will explore the ways the experience of sexual sensation might serve as an expression of identity and culture. For example, sexual sensation might serve as an expression of identity because the law’s history of regulating sexuality has imbedded political and moral viewpoints in the types of sexual sensation one enjoys and chooses to engage in. In other words, by regulating sexuality—based in part on morality—the law has categorized and stigmatized people in a way that has contributed to cultural and individual identity. In fact, the Supreme Court expressly acknowledged the categorizing and stigmatizing effects of regulating sexuality in Lawrence v. Texas.97 In Lawrence, the Court found that “[t]he Texas statute [banning homosexual sodomy] undeniably seeks to further the belief that certain forms of sexual behavior are immoral and unacceptable—the same interest furthered by criminal laws against fornication, bigamy, adult incest, bestiality, and obscenity.”98 After explaining the lower moral standing that the Texas statute bestowed upon the gay community,99 the Court struck down the law.100 Ultimately, societal viewpoints regarding sexuality have contributed to the development of sexual subcultures and have made an individual’s sexual practices a notable part of personal identity. Significantly, where a certain group is socially condemned based on the shared enjoyment of a particular sexual practice, the communication of the sexual sensation associated with that practice may certainly express the identity or culture that results from the group’s common experience of stigmatization. And while it is true that all conduct, and its accompanying experiences, can carry messages of identity and culture, the ability of teledildonics to manufacture and distribute sexual sensation to audiences separate and apart from sexual acts gives the culture and identity aspects of sexual sensation new relevance to the First Amendment. If we understand sexual sensation to be an expression of culture and identity, the truth and autonomy justifications would likely provide First Amendment protection because the experience and communication of cultural diversity furthers the quest for truth and enhances an individual’s sense of self.101 Moreover, sexual sensation as an expression of culture and identity could also trigger the democracy justification because it has the ability to influence sexual politics. For example, the mass distribution of certain types of sexual sensation can influence society’s sexual mores and allow certain ­marginalized

  539 U.S. 558, 599 (2003).   Id. More recently, in Obergefell v. Hodges, where the Supreme Court declared state laws banning same-sex marriage unconstitutional, the Court declared that “liberties extend to certain personal choices central to individual dignity and autonomy, including intimate choices that define personal identity and beliefs.” 135 S. Ct. 2584, 2597 (2015).  99   Lawrence v. Texas, 539 U.S. 558, 581–2 (2003) (“Texas’ sodomy law brands all homosexuals as criminals, thereby making it more difficult for homosexuals to be treated in the same manner as everyone else. Indeed, Texas itself has previously acknowledged the collateral effects of the law, stipulating in a prior challenge to this action that the law ‘legally sanctions discrimination against [homosexuals] in a variety of ways unrelated to the criminal law,’ including in the areas of ‘employment, family issues, and housing.’”). 100   Id. at 585. 101   Cf. Chen, supra note 53, at 429–30 (finding that instrumental music as an expression of culture and religion would likely receive First Amendment protection under the truth and autonomy theories because it “advances the ability to experience cultural diversity”).  97  98

BARFIELD PRINT.indd 292

13/11/2018 11:52

VR, haptics, First Amendment protection for sexual sensation  293 sexual  ­ communities to gain mainstream acceptance.102 In this way, insofar as the experience of sexual sensation can develop political knowledge and understanding, the democracy justification could also support First Amendment protection. 5.3  Sexual Sensation Is an Expression of Emotion The third and perhaps most intuitive argument for the communicative value of sexual sensation is that sexual sensation is an expression of emotion.103 Some forms of nonpropositional expressions do “not express this or that particular and definite pleasure, this or that affliction, pain, sorrow, horror, gaiety, merriment, or peace of mind, but joy, pain, sorrow, horror, gaiety, merriment, peace of mind themselves.”104 The argument is that nonverbal, nonpropositional expressions are communicative not because they communicate an articulable emotion, but because they themselves induce emotions and feelings in the listener that cannot be reduced to words. Sexual sensation certainly induces inexpressible feelings and emotions and may, in some sense, be regarded as an emotion or feeling in and of itself. As previously mentioned, the Seventh Circuit expressly recognized “eroticism and sensuality” as an emotional message, noting that “[t]he dominant theme communicated [] by the dancers is an emotional one; it is one of eroticism and sensuality.”105 By definition, the term “sensuality” refers to anything “relating to, devoted to, or producing physical or sexual pleasure.”106 Significantly, with this definition one may reason that if sensuality is defined by its goal of producing physical or sexual pleasure, then in the Seventh Circuit, sexual sensation and its accompanying physical responses must also be a component of that emotional message. Along with the Seventh Circuit, the Supreme Court has also vaguely acknowledged the emotive component of experiencing sexual sensation. For instance, in Lawrence, the Court stated that, “[w]hen sexuality finds overt expression in intimate conduct with another person, the conduct can be but one element in a personal bond that is more enduring.”107 Furthermore, in Stanley, the Court invalidated the regulation of obscene materials inside the home on the grounds that the defendant had “the right to satisfy his 102   Similarly, Thomas Scanlon argued that even hardcore pornography should be protected by the First Amendment because it has the ability to “influence the sexual mores of society.” James Weinstein, Democracy, Sex and the First Amendment, 37 N.Y.U. Rev. L. & Soc. Change 865, 866 (quoting Thomas M. Scanlon, Jr, Freedom of Expression, 40 U. Pitt. L. Rev. 519, 545 (1979)). 103   Sexual sensation might induce emotions in the viewer that do not exactly reflect the emotions of the performer. In this sense, sexual sensation would not be “communicative.” Nonetheless, protection would still be justified because the First Amendment’s autonomy justification values communications that induce solitary experiences of emotion. See Stanley, 394 U.S. at 565; see also Marc Jonathan Blitz, Freedom of 3D Thought: The First Amendment in Virtual Reality, 30 Cardozo L. Rev. 1141, 1149 (2008) (noting that individuals engaged in virtual reality activities can “benefit immensely . . . from the feedback loop that is formed when they are both speaker and listener”). 104  Chen, supra note 53, at 432 (quoting Arthur Schopenhauer, 1 The World As Will And Representation 261 (1958)). 105   South Bend, 904 F.2d at 1086−7. 106  Merriam-Webster, www.merriam-webster.com/dictionary/sensual (last visited Feb. 24, 2016). 107   Lawrence, 539 U.S. at 567.

BARFIELD PRINT.indd 293

13/11/2018 11:52

294  Research handbook on the law of virtual and augmented reality intellectual and emotional needs in the privacy of his home.”108 With this language, the court suggests that obscene materials are capable of communicating or inducing emotional messages or experiences.109 However, even if we take a different stance on how to define what acts should count as First Amendment speech—even if we reject courts’ insistence that nonverbal activity must have a “particularized message” to have First Amendment value—a social practice (such as use of haptic technology) that generates emotional states may merit First Amendment protection even in the absence of any such message. If sexual sensation is understood as an expression of emotion, it may struggle to find protection under the democracy justification. Perhaps under some of the most expansive interpretations of the democracy justification, which afford protection when an expression merely provides “knowledge, intelligence, [and] sensitivity to human values,”110 sexual sensation could be protected. Yet, generally, it would be difficult to understand how emotional expression contributes to public discourse in a way that would warrant constitutional protection under the democracy justification.111 Sexual sensation as a communication of emotion would, however, certainly be justified by both the truth and autonomy justifications. In fact, the Supreme Court has been clear in finding that the protection of emotional communications furthers both the autonomy and truth-seeking purposes of the First Amendment. For example, in Cohen v. California, the Court explained: much linguistic expression serves a dual communicative function: it conveys not only ideas capable of relatively precise, detached explanation, but otherwise inexpressible emotions as well. In fact, words are often chosen as much for their emotive as their cognitive force. We cannot sanction the view that the Constitution, while solicitous of the cognitive content of individual speech has little or no regard for that emotive function which practically speaking, may often be the more important element of the overall.112

Furthermore, prior to Cohen, the Court in Stanley stated that “[t]he makers of our Constitution . . . recognized the significance of man’s spiritual nature, of his feelings and of his intellect . . . They sought to protect Americans in their beliefs, their thoughts, their emotions, and their sensations.”113 Accordingly, sexual sensation as an expression of emotion would almost certainly receive First Amendment protection under the truth and autonomy justifications.

6. COUNTERARGUMENTS Although there are several arguments that sexual sensation is communicative and should receive First Amendment protection, there are also viable counterarguments that would   Stanley, 394 U.S. at 565.   Bear in mind that obscenity refers to sexual expressions “so offensive to contemporary moral standards.” FCC v. Pacifica Found, 438 U.S. 726, 745 (1978). 110  Chen, supra note 53, at 433 (quoting Alexander Meiklejohn, The First Amendment Is an Absolute, 1961 Sup. Ct. Rev. 245, 256). 111   Id. 112   Cohen v. California, 403 U.S. 15, 26 (1971). 113   Stanley, 394 U.S. at 564. 108 109

BARFIELD PRINT.indd 294

13/11/2018 11:52

VR, haptics, First Amendment protection for sexual sensation  295 advocate for denying protection. One such argument is that despite its communicative value, sexual sensation is not used to engage our intellect but is used only as a tool to achieve the purely physical responses of sexual arousal and orgasm. Others may argue that affording First Amendment protection to sexual sensation in the context of teledildonics will have the undesirable, and unavoidable, consequence of affording First Amendment protection to real-life sexual encounters. Furthermore, one might argue that the realistic qualities of simulated sensations of touch lead to constitutionally significant differences between the use of teledildonics and other traditional forms of media. These arguments, in one way or another, contend that sexual sensation operates less like protected speech and more like unprotected conduct. However, as is set forth more fully below, all should be rejected. 6.1  Functionality Doctrine One argument for finding that sexual sensation should be denied First Amendment protection is that, in spite of any communicative value it may have, sexual sensation is not being used to communicate. Instead, it is being used as a tool to achieve the purely physical responses of sexual arousal and orgasm. Doctrinally, this argument would find its roots in the “functionality doctrine”114 which denies First Amendment protection when “speech” is being used as a tool to “perform[] some task other than the communication of ideas.”115 For example, the First Amendment does not protect saying “kill” to an attack dog.116 Insofar as the physical responses of sexual arousal and orgasm are regarded as

114   The “functionality doctrine” originated in the context of intellectual property law, but a de facto functionality doctrine has developed in First Amendment law as well. Wu, supra note 19, at 1520, 1524 (“[W]e can safely say that, like other areas of the law, the First Amendment usually denies protection to carrier/conduits and mere communicative tools. The reasons, while rarely given, seem similar to those found in intellectual property cases: judges fear the consequences of allowing a law intended to protect expression to venture into areas where other motivations are paramount. They may also fear opportunism on the part of the lawyers—an effort to use the Constitution for goals quite unrelated to speech”); see, e.g., Winter v. G.P. Putnam’s Sons, 938 F.2d 1033, 1035−36 (9th Cir. 1991) (finding that aeronautical charts, although “graphical depictions,” are “highly technical tools” denied First Amendment protection). 115  Wu, supra note 19, at 1521. The functionality doctrine will also deny First Amendment protection to a “speaker” when the alleged speaker is acting only as a carrier of information, making its involvement with the information too mechanical or too distant to afford protection. Id. at 1520; see Rumsfeld v. Forum for Academic & Institutional Rights, 547 U.S. 47 (2006) (finding that law schools were not “speaking” nor endorsing the hiring policies of the military when hosting military recruiters for on-campus interviews); but see Turner Broadcasting System v. FCC, 512 U.S. 622, 636 (1994) (finding that cable operators received First Amendment protection because they had editorial discretion over which program to include). In analyzing the First Amendment protection afforded to the use of teledildonics, the relevant speaker is the pornographic performer and the carrier of information would be teledildonics. This chapter does not argue that teledildonics, as an independent entity, should receive First Amendment protection as a speaker, but only that the performer using teledildonics should. Therefore, this prong of the functionality doctrine is not relevant. 116  Wu, supra note 19, at 1523 (citing Catharine A. Mackinnon, Only words 12 (1993)). As Mackinnon explains, “social life is full of words that are legally treated as the acts they constitute without so much as a whimper from the First Amendment.” Id. (internal quotation marks

BARFIELD PRINT.indd 295

13/11/2018 11:52

296  Research handbook on the law of virtual and augmented reality noncommunicative or noncognitive “tasks”, sexual sensation can be categorized as a “tool” outside the scope of First Amendment protection. Understanding orgasm as a “task” and sexual sensation as a “tool” would naturally follow from many of the arguments opposing the First Amendment’s protection of pornography. These arguments contend that the physical effects of viewing pornography make the exchange noncognitive and purely physical. For example, Frederick Schauer argues that pornography should receive less First Amendment protection because it is “essentially a physical activity” that is “no more cognitive than any other experience with a prostitute.”117 Similarly, Cass Sunstein argues that, “[t]hough comprised of words and pictures, pornography does not have the special properties that single out speech for special protection; it is more akin to a sexual aid than a communicative expression.”118 In concluding that the observation of pornography is a purely physical engagement outside the scope of First Amendment protection, both Schauer and Sunstein presume that acts with a primary function of generating physical responses, such as sexual arousal and orgasm, cannot be communicative. Insofar as Schauer and Sunstein are correct, haptics or other mediums that generate sexual sensation should be denied First Amendment protection because they are “tools” used to achieve the noncommunicative, noncognitive “tasks” of arousal and orgasm. However, creating a dichotomy between the communicative value of speech or art and the state of arousal it generates, as Schauer and Sunstein would urge, is unsound. For instance, no one would argue that a painting’s depiction of sadness is precluded from First Amendment protection simply because the painting is being used as a tool to make the viewer cry.119 To the contrary, it is primarily through the physical response of crying that the viewer is able to access and understand the feeling of sadness being communicated. Likewise, the physical states of arousal and orgasm are likely necessary to understand the communicative value of sexual sensation. Therefore, it would not make sense to argue that the physical responses of arousal and orgasm in any way undermine the communicative value of sexual sensation, let alone preclude First Amendment protection. Yet it must be conceded that there is a critical difference between tears induced by a painting and sexual arousal induced by experiencing sexual sensation. Before crying, a

omitted). In this sense, the concept of a First Amendment “functionality doctrine” substantially overlaps with the concept of what speech philosophers call “speech acts.” Id. at 1523; see also What is a Speech Act? Center For Advanced Research On Language Acquisition, University of Minnesota, http://carla.umn.edu/speechacts/definition.html (last visited Mar. 11, 2016). 117  Weinstein, supra note 102, at 868−9 (citing Frederick Schauer, Speech and “Speech”— Obscenity and “Obscenity”: An Exercise in the Interpretation of Constitutional Language, 67 Geo. L.J. 899, 922 (1979)). Schauer has argued that hardcore pornography is no different than “rubber, plastic, or leather sex aides,” for “if every other aspect of the experience is the same,” the “mere fact that in pornography the stimulating experience is initiated by visual rather than tactile means” does not matter. Id. at 869. 118   Cass R. Sunstein, Pornography and the First Amendment, 1986 Duke L.J. 589, 606 (1986). 119   As James Weinstein notes: “Appeals to emotions, whether through words, pictures or music, can make us laugh, cry, clench our fists, jump in our seats or even feel sick to our stomachs. None of this would disqualify First Amendment protection in a medium of mass communication used by a speaker as a means of trying to change people’s attitudes on matters of public concern.” Weinstein, supra note 102, at 886.

BARFIELD PRINT.indd 296

13/11/2018 11:52

VR, haptics, First Amendment protection for sexual sensation  297 viewer must first cognitively interpret the painting as sad. There is, however, likely little cognitive interpretation needed to arrive at sexual arousal and orgasm from the experience of sexual sensation.120 Nonetheless, this chapter argues that even if the communication of sexual sensation circumvents the cognitive interpretation that has traditionally been required to induce physical responses, those responses are not precluded from holding cognitive or communicative value. For even when cognition is not required to induce physical responses, the mere presence of physical response is crucial to cognitive interpretation and experience of emotional messages. Notably, understanding physical response as a necessary component of experiencing emotion is consistent with many scientific theories of emotion.121 While there are varying schools of thought, several theories suggest that a person can only arrive at a conscious state of emotion by first identifying their state of physical arousal.122 Consider the following example: You are walking down a dark alley late at night. You hear footsteps behind you and begin to tremble; your heart beats faster and your breathing deepens. You notice these physiological changes and interpret them as your body’s preparation for a fearful situation. You then experience fear.123

This example specifically illustrates the James–Lange theory of emotion, which “argues that an event causes a physiological response first and then we interpret that arousal. Only after the interpretation of the arousal can we experience emotion.”124 Significantly, the James–Lange theory, along with several others, contends that a person cannot experience emotion in the absence of physical arousal.125 120   A viewer’s cognitive engagement with a form of communication appears to be a significant reason for affording First Amendment protection. The primary response to the arguments of Schauer and Sunstein is that the viewer of pornography only arrives at sexual arousal after having interpreted the pornography as arousing. This cognitive interpretation, it is argued, is a reason for affording First Amendment protection. Weinstein, supra note 102, at 869 (citing Martin H. Redish, Freedom of Expression: A Critical Analysis 74–5 (1984) (defending the First Amendment protection afforded to pornography by arguing that the physical response of sexual arousal requires the viewer to intellectually interpret pornography); Margo Kaplan, Sex Positive Law, 89 N.Y.U. L. Rev. 89, 113 (2014) (criticizing the premises for denying obscenity protection by noting that “[t]he physical reaction of arousal occurs not because obscenity acts on a purely physical level, but because of the way the consumer interprets what he views or reads as arousing”). 121   See Psychology 101. Chapter 7: Motivation and Emotion. Section 3, Emotion. All Psych Online. All Psych and Hefner Media Group, Inc., http://allpsych.com/psychology101/emotion. html. 122   Id. 123   Id. 124   Id. 125   Id. In addition to the James–Lange theory, the Schacter–Singer and Facial Feedback theories similarly suggest that physical response is necessary to experience emotion. Id. For instance, according the Schachter–Singer theory, “an event causes physiological arousal first. You must then identify a reason for this arousal and then you are able to experience and label the emotion.” Id. Likewise, the Facial Feedback theory provides that the acts of smiling and frowning trigger emotions. Id. In other words, “[i]t is the changes in our facial muscles that cue our brains and provide the basis of our emotions. Just as there are an unlimited number of muscle configurations in our face, so too are there a seemingly unlimited number of emotions.” Id.

BARFIELD PRINT.indd 297

13/11/2018 11:52

298  Research handbook on the law of virtual and augmented reality For these reasons, the physical responses of sexual arousal and orgasm should naturally be protected by the First Amendment as necessary components of understanding the communicative value of sexual sensation. More importantly, never should physical responses that are pivotal to interpreting emotional messages be used as a basis to characterize otherwise communicative expressions as “tools” falling outside the scope of First Amendment protection. Because arousal and orgasm are necessary to understand the communicative value of sexual sensation, neither can, nor should, be a reason to deny First Amendment protection. This has significant implications for how we understand teledildonics: Where haptics in a virtual reality environment generates such sexual sensations, they should not lose their First Amendment value because the emotions generated by the haptics necessarily have a physical component. If physical arousal is—as the James–Lange theory says—a necessary precondition of emotional experience, then there will be no way to convey or generate emotional experiences without it. In video games or movies that depict a fastpaced sporting event, such as a race, or a heart-stopping action scene where a character confronts danger (such as an armed enemy, or a risky climb up a mountain), the vicarious thrill experienced by the viewer—or player—is possible only because it comes with physical arousal. This physically laden emotional experience is likely to be even more intense in a virtual reality video game or other simulation where one feels one is racing a car, climbing a mountain, or fleeing from an attacker. And the use of haptics to enhance both the physical arousal and emotional reaction is functionally equivalent to the use of three-dimensional sound and imagery. One might argue that haptics are different—that they do not have the First Amendment status of visual and auditory elements that have long been a part of movies, and more recently of video games. In Brown v. Entertainment Merchants Association, the Court said that video games deserve First Amendment protection in part because “[l]ike the protected books, plays, and movies that preceded them, video games communicate ideas—and even social messages—through many familiar literary devices (such as characters, dialogue, plot, and music) and through features distinctive to the medium (such as the player’s interaction with the virtual world).”126 One could argue that use of haptics to generate emotional experiences breaks the link between virtual experiences and “protected books, plays and movies”—that where virtual reality uses imagery or plots like the latter, it is protected, but where it uses haptics it wanders outside of the First Amendment’s realm. But this is an unwarranted conclusion. Like the simulated sights and sounds of such a VR experience, the haptics components are giving individuals a sense of being a part of the story and experiencing it from a character’s vantage point. Moreover, the Court said that First Amendment protection applies not only to familiar devices for letting a viewer enter into a character’s experience, but also to “features distinctive to the medium (such as the player’s interaction with the virtual world).”127 For reasons explained in section 2 of this chapter, incorporating haptics has long been a part of the virtual reality design. As explained in that section, haptics has been used in entertainment for decades and is already a standard feature in

126 127

  Brown, 131 S. Ct. at 2731.   Id.

BARFIELD PRINT.indd 298

13/11/2018 11:52

VR, haptics, First Amendment protection for sexual sensation  299 both video games and cellphones, and it is not surprising that Oculus Rift, HTC Vive, and other contemporary virtual reality companies are including haptic gloves, foot cuffs, or harnesses in their game consoles; that other companies (such as AxonVR) are developing full body suits for haptics; or that Kiiro is adding to individuals’ watching of sexual films. In all these cases, haptics is being used to enhance the emotional experience of participating in an immersive virtual experience, or the less immersive experience of watching a film. It should be noted, however, that this argument may be less persuasive in other forms of virtual reality where the physical sensations induced are not typically associated with interpreting one’s own emotions. For example, while the physical reactions following a virtual hug are easily associated with feelings of love and connection, the physical reactions, such as pain and bruising, following a virtual punch to the face might be less useful in interpreting emotions. Accordingly, determining whether or not the physical responses induced through haptics will define a virtual experience as noncognitive is a fact-sensitive inquiry. 6.2  Real-Life Sexual Encounters Another possible argument for denying First Amendment protection to the use of teledildonics is the practical implications that affording such protection might have on the government’s ability to regulate real-life sexual encounters. If sexual sensation distributed through the use of teledildonics can carry messages that are communicated, in part, through sexual arousal and orgasm, then one might argue that the First Amendment must also afford protection to actual sex. However, affording First Amendment protection to real-life sexual acts would complicate, if not prohibit, many government regulations, such as those on prostitution. Therefore, it must be determined whether, for the purposes of the First Amendment, sexual sensation communicated through the use of teledildonics can be distinguished from the sexual sensation communicated through real-life sexual acts. As explained in section 6.1, many arguments in favor of denying First Amendment protection to pornography rely on the premise that pornography and sex are essentially the same—that both primarily seek to induce the noncognitive responses of arousal and orgasm. However, James Weinstein counters this argument by noting that “[u]nlike a live encounter with a prostitute, the usual channels of hardcore pornography—magazines, films and the Internet—are media of mass communication.”128 Weinstein further explains: [D]espite its alleged minimal intellectual content, sexual activity captured on film or a DVD and widely distributed in these media has far more potential to influence public attitudes towards sex, morals, the role of women and other matters of public concern than does the same activity taking place in the flesh and observed by a single person.129

In this way, Weinstein argues that because pornography employs methods of mass communication in a way that a private sexual encounter does not, there are constitutionally sound reasons for distinguishing pornography from real-life sexual encounters under the First Amendment.130  Weinstein, supra note 102, at 873.   Id. 130   Id. 128 129

BARFIELD PRINT.indd 299

13/11/2018 11:52

300  Research handbook on the law of virtual and augmented reality Notably, Weinstein’s argument can also be used to distinguish the use of teledildonics from real-life sexual encounters because the use of teledildonics in pornographic films also utilizes mass communication in a way that private sexual encounters do not. When pornography uses teledildonics, it allows a single pornographic performer to communicate sexual sensation to each member of the audience. Therefore, the use of teledildonics can communicate whatever message may be contained within sexual sensation to large audiences in a way that real-life sexual encounters cannot. For this reason, the First Amendment can protect pornography’s use of teledildonics while continuing to deny protection to real-life sexual encounters. What the Supreme Court observed to be true of video games—that, “[l]ike the protected books, plays, and movies that preceded them,” they “communicate ideas” and immerse people in stories with literary and artistic ­elements—is also true of virtual reality. And where an expressive medium immerses someone in a fictional experience, it is protected by the First Amendment even if it is not First Amendment activity to engage in the physical version of such activity outside of that expressive medium. Notably, the same argument could be made for other forms of virtual reality that use haptics to simulate real-life encounters, such as boxing or fighting, insofar as the sensations communicate a message protected by the First Amendment and are distributed through a means of mass communication, or perhaps some other medium of expression. 6.3 Harmfulness One might also argue that sexual sensation, and the sensation of touch generally, provide a more realistic experience than traditional visual and audio communications and should, therefore, be treated differently under the First Amendment. Marc Jonathan Blitz explores this argument in the context of providing First Amendment protection to virtual reality (VR). Blitz believes it is “possible that there are constitutionally-significant differences between some forms of VR and other, less technologically-sophisticated ways of representing our fantasies.”131 As Blitz notes, “even when we know that an armed VR attacker is an illusion . . . we might find that the fears we experience—or other visceral and emotional impacts—are more comparable to those we experience in a real-life attack than when we watch a film depicting violence.”132 Similar to other forms of VR, the ability of teledildonics to engage the viewer’s sense of touch may create an experience that is of an entirely different character than the experience provided by traditional pornography. For instance, one could argue that the experience of sexual sensation will provide a more immersive experience to viewers and have the effect of exacerbating some of the harmful effects associated

131  Blitz, supra note 103, at 1175. For example, Blitz notes that “[a] virtual car ride will give us a sense of genuine movement, not an experience where we merely watch (or use video game controls to generate) a representation of such movement.” Id. at 1174. 132   Id. A similar argument was made by Justice Breyer’s dissent in Brown v. Entertainment Merchants Ass’n. Justice Breyer argued that “[a] typical video game involves a significant amount of physical activity. And pushing buttons that achieve an interactive, virtual form of target practice (using images of human beings as targets), while containing an expressive component, is not just like watching a typical movie.” Brown, 131 S. Ct. at 2766 (2011) (Breyer J dissenting).

BARFIELD PRINT.indd 300

13/11/2018 11:52

VR, haptics, First Amendment protection for sexual sensation  301 with pornography, such as sexualizing the humiliation and subordination of women.133 Or even more fundamentally, one could argue that the lifelike experience provided by the sensation of touch, in combination with visual and audio communication, should lead us to treat this experience as a real-life sexual encounter.134 For these reasons, it could be argued that the sensation of touch, and sexual sensation in particular, can be distinguished from the traditional constitutionally protected appeals to our senses of sight and hearing. However, to the extent that sexual sensation may have uniquely harmful effects on its audiences, or otherwise creates an exceptionally realistic experience, such considerations do not speak to whether First Amendment protection applies in the first place. Determining whether the First Amendment applies deals only with whether a given expression has the requisite communicative value to implicate the First Amendment; it does not involve interest balancing.135 After the First Amendment is deemed applicable, the analysis would then most likely move to the factors outlined in U.S. v. O’Brien, a case that used a balancing test to weigh the government’s interest in regulating the expressive conduct against the communicative value of the expression.136 Accordingly, any harms that may result from experiencing sexual sensation should not be considered when determining if the First Amendment applies. Instead, these potential harms, and any other characteristics that are unique to sexual sensation, should be accounted for under the O’Brien factors when balancing the government’s interest in regulating the use of teledildonics against the communicative value of sexual sensation.

133   Again, a parallel point was made in Brown’s dissenting opinion, where Justice Breyer noted that “video games can cause more harm in this respect than can typically more passive media, such as books or films or television programs.” Id. at 2768. However, in the context of teledildonics, the immersive characteristics of experiencing sexual sensation while watching pornography can just as easily have exceptionally positive effects. “For sexual minorities, such as gay people, pornography may provide rare positive images that affirm their own sexuality. It can be an important weapon in the fight against AIDS and other sexually transmitted diseases, by eroticizing safer sex practices.” Bret Boyce, Article: Obscenity and Community Standards, 33 Yale J. Int’l L., 344 (2008). If teledildonics can exacerbate the troubling effects of viewing pornography, it can exacerbate the positive effects as well. 134   Even if the use of teledildonics is equivalent to a real-life sexual encounter from the perspective of the end user, the use of teledildonics can still be distinguished from reality. For instance, unlike an actual sexual encounter, during the use of teledildonics the sensation of sexual sensation is being unilaterally communicated by the performer and distributed to the masses. It is this property that not only implicates the First Amendment, but also distinguishes teledildonics from real-life encounters. See section 4.2. 135   See Burstyn, 343 U.S. at 502–3 (acknowledging that the applicability of the First Amendment is separate from the degree to which the government can regulate). 136   U.S. v. O’Brien, 391 U.S. 367, 377 (1968). In O’Brien, the Court ruled that the government’s regulation of expressive conduct is justified when “[1] it is within the constitutional power of the Government; [2] it furthers an important or substantial governmental interest; [3] if the governmental interest is unrelated to the suppression of free expression; and [4] if the incidental ­restriction of the First Amendment freedoms is no greater than necessary to further the ­government interest.”

BARFIELD PRINT.indd 301

13/11/2018 11:52

302  Research handbook on the law of virtual and augmented reality

7. CONCLUSION Ultimately, pornography’s use of teledildonics should most likely be considered expressive conduct because the sexual sensation being distributed to audiences contains communicative value that implicates the First Amendment. Although sexual sensation induces physical responses by circumventing cognitive interpretation, these physical responses do not undermine, nor preclude, sexual sensation from having communicative value. To the contrary, physical responses such as sexual arousal and orgasm are likely necessary for the viewer to interpret and understand the emotional messages that can be communicated through sexual sensation. Significantly, these conclusions also bring clarity to the larger issues raised by haptic technology and the First Amendment’s distinction between protected speech and unprotected conduct. Generally, although the sensation of touch does not likely require cognitive interpretation to induce physical responses, those physical responses still play a critical role in helping the viewer understand what is being communicated. As an integral part of understanding emotional messages, noncognitive physical responses should be regarded as a component of protected speech and not unprotected conduct. Moreover, to the extent that the sensation of touch creates an exceptionally realistic experience that can effectuate harms more like those resulting from conduct, those potential harms are not relevant to the applicability of the First Amendment. They justify allowing government to regulate haptics that generate sexual sensations or other experiences, but do not justify the conclusion that haptics lack First Amendment value altogether.

SOURCES Cases Abrams v. United States, 250 U.S. 616, 617 (1919) Am.’s Best Family Showplace Corp. v. City of New York, 536 F. Supp. 170 (E.D.N.Y. 1982) Ashcroft v. Free Speech Coalition, 535 U.S. 234, 248 (2002) Barnes v. Glen Theatre, Inc., 501 U.S. 560 (1991) Brown v. Entertainment Merchants Ass’n, 131 S. Ct. 2729 (2011) City of Erie v. Pap’s A.M., 529 U.S. 277 (2000) Cohen v. California, 403 U.S. 15 (1971) Edelman v. Croonquist, No 09–1938, 2010 WL 1816180 (D.N.J. May 4, 2010) FCC v. Pacifica Found, 438 U.S. 726 (1978) Joseph Burstyn, Inc., 343 U.S. 495 (1952) Lawrence v. Texas, 539 U.S. 558 (2003) Miller v. California, 93 S. Ct. 2607, 2615 (1991) Miller v. Civil City of South Bend, 904 F.2d 1081 (1990), rev’d sub nom. People v. Freeman, 46 Cal.3d 419 (1988) Red Lion Broad. Co. v. FCC, 395 U.S. 367 (1969) Reno v. Am. Civil Liberties Union, 521 U.S. 844 (1997) Rumsfeld v. Forum for Academic & Institutional Rights, 547 U.S. 47 (2006) Spence v. State of Wash., 418 U.S. 405 (1974) Stanley v. Georgia, 394 U.S. 557 (1969) Texas v. Johnson, 491 U.S. 397 (1989) Troster v. Pennsylvania State Department of Corrections 65 F.3d 1086 (3d Cir. 1995) Turner v. Broadcasting System V. FCC, 512 U.S. 622 (1994)

BARFIELD PRINT.indd 302

13/11/2018 11:52

VR, haptics, First Amendment protection for sexual sensation  303 U.S. v. O’Brien, 391 U.S. 367 (1968) Ward v. Rock Against Racism, 491 U.S. 781 (1989) Winter v. G.P. Putnam’s Sons, 938 F.2d 1033 (9th Cir. 1991)

Statutes U.S. Const. amend. I

BARFIELD PRINT.indd 303

13/11/2018 11:52

10.  Augmented and virtual reality, freedom of expression, and the personalization of public space Marc Jonathan Blitz

1. INTRODUCTION Drive along any major thoroughfare in America, and you will see numerous billboards and other displays: Advertisements for nearby restaurants and other businesses, announcements of new television shows, even inspirational or religious messages (such as signs stating “Happiness is Contagious,”1 “Enjoy the Moment,”2 or “There Is Evidence for God”3). Such messages are typically protected by the First Amendment of the United States Constitution.4 So too are the images that accompany them or that sometimes constitute the entire content of a display.5 Still, courts have long found that such signs may be constitutionally subject to government regulations aimed at assuring the “safe and convenient flow of traffic,”6 or at controlling “visual blight” or otherwise safeguarding the community’s “aesthetic interests” in an attractive public space.7 There is, however, a duplicate version of public space where this regulatory power largely vanishes—where government has far less power to place limits on the signs plastered over streets and highways: a version of public space where free speech protection is

1   Alena Hall, The Can’t-Miss Billboards In These Cities Have An Important Announcement: Smile, HuffPost, Feb. 12, 2015, at www.huffingtonpost.com/2015/02/12/the-joy-team-billboardhappiness_n_6631450.html (accessed on Nov. 10, 2017). 2   See The Joy Team, http://thejoyteam.org/about/ (accessed on Nov. 10, 2017). 3   See Ann-Derrick Gaillot and Adam Pape, Signs From God, Fader, Dec. 5. 2016, at http:// www.thefader.com/2016/12/15/religious-billboards-america (accessed on Nov. 10, 2017); Christian Aid Ministries, Billboard Evangelism, Oct. 24, 2016, at https://christianaidministries.org/billboardevangelism/ (accessed on Nov. 10, 2017). 4   Metromedia, Inc. v. City of San Diego, 453 U.S. 490, 502 (1981). In this case, the Supreme Court found that government may not, consistent with the First Amendment, control “the communicative aspects” of billboards. 5   See Madsen v. Women Health Center, Inc., 512 U.S. 753, 773. This case struck down a portion of state court injunction against abortion clinic protest signs that displayed “images observable to . . . the patients inside the clinic,” which assumed that such images are protected by the First Amendment. 6   See Lone Star Security and Video, Inc. v. City of Los Angeles, 827 F.3d 1192, 1201 (9th Cir. 2016). 7   Id; Taxpayers for Vincent, 466 U.S. 789, 808 (1984). As Ashutosh Bhagwat notes, when the government creates rules for speech in public space (including display of signs) it faces a problem that First Amendment law has to give it room to solve: that of scarcity. “[T]he amount of speech, and number of speakers, on any particular piece of property will have to be restricted.” In Defense of Content Regulation, 102 Iowa L. Rev. 1427, 1456 (2017).

304

BARFIELD PRINT.indd 304

13/11/2018 11:52

Freedom of expression, and the personalization of public space  305 stronger, and government’s regulatory interests substantially weaker. This is the version of public space one finds not in the physical world, but in a custom-designed augmented or virtual reality experience. Imagine, for example, that I am viewing signage not as I walk on a concrete street or drive on an asphalt road, but rather as I stare at a computer or mobile device that simulates such movement over a street or road. Google Earth is one of the most familiar technologies for offering such personalization of public space. When I look at a satellite image of a location in Google Earth, I see images of real streets and buildings (captured at a recent point in time by Google’s army of vehicle-mounted cameras). But superimposed on that image of physical reality are words or images that are not part of the physical landscape captured by the picture:8 street names and neighborhood names, red Hs signifying hospitals, small yellow icons signifying churches or other religious centers, cartoonish purple drinks signifying bars, and other symbols I can choose from a list of “layers.” If I select tens of layers at once, then the city underneath may hardly be visible: Google Earth’s satellite image of New York City, for example, becomes hard to discern under the dense cluster of icons representing restaurants, museums, subway stations, and links to Wikipedia articles. It is hard to see why the government should typically have any legitimate interest here in preventing me from clicking an option that allows me to view, in my private virtual experience, any kind of informational display that I wish to see—even the in-app equivalent of the monstrous and distracting billboards that might be banned or restricted by town ordinances in the real world. No car drivers are distracted by such in-app displays. I am the only one who sees them, and I do so as I sit safely in front of my computer or look upon the screen of a personal mobile device. These virtual signs don’t fill any physical space—only pixels and computer memory. And the electronic space they do fill is not space that I share with the public (under the government’s watchful eye). It is “space” that I control on my own computer or mobile device, and experience in my own home or in my own private interaction with a digital device. In short, this version of public space occurs in a private experience, and one that I tailor to my own preferences regarding how I want the street to appear to me.9 This is true of a flat virtual world I watch on a computer screen. It also true of a simulation of public space in an immersive “virtual

8   Frank Taylor, Google Earth Layers, Google Earth Blog, Mar. 22, 2009, www.gearthblog. com/blog/archives/2009/03/google_earth_layers_1.html (accessed on Nov. 10, 2017). 9   The power of VR and AR to work such a transformation—from public shared experience to private customized experience—has been explored in recent work by other legal scholars. In an article on regulating virtual violence, Jaclyn Seelagy observes that VR can “creat[e] a space” where aspects of life that are normally “outside of our control,” such as the effects our actions have on others, are “products of human choices, from the developers to the users.” Virtual Violence, 64 UCLA L. Rev. Discourse 412, 416 (2017). Mark Lemley and Eugene Volokh similarly explain how VR and AR can allow “a broader diversity of experience within the same environment”: rather than have a shared experience of public space and how we interact with people in it, we can each custom design our own “sensescape.” Mark A. Lemley and Eugene Volokh, Law, Virtual Reality, and Augmented Reality, 166 U. Pa. L. Rev. 1051 (2018). Joshua Fairfield has also examined the legal questions raised by “virtual worlds that mirror the real world,” such as “[t]he full 3-D version of Google Earth”: Mixed Reality: How the Law of Virtual Worlds Govern Everyday Life, 27 Berkeley Tech. L.J. 55, 99 (2012).

BARFIELD PRINT.indd 305

13/11/2018 11:52

306  Research handbook on the law of virtual and augmented reality reality” (VR) environment that seems to surround me. My personalized version of public space may include alterations more substantial than informational signs or commercial billboards overlaid over an otherwise accurate image of a neighborhood I wish to explore in New York or another setting: I could obtain (or perhaps program) a futuristic virtual version of New York, where movable buildings shift their positions throughout the day or individuals hover over streets in flying cars. Whether the visual alterations to a true to life image of New York simply provide factual information, in the form of billboards or signs, or fictional transformations, the First Amendment most likely makes it my decision—not that of the government—what I choose to see in a private virtual reality (or VR) experience. For the same reason, the freedom of speech protected by the First Amendment will likely shut government regulators not only out of my use of virtual reality technology to reproduce (and perhaps alter) my experience of public space, but also out of augmented reality technology that allows me to overlay information or images onto my direct perception of public space itself. When I look at a building with augmented reality (or “AR”) glasses, for example, I might see projected over it information about its history: When it was built, who the architect was, and an account of what kind of businesses I might find inside of it. Augmented reality apps on a smartphone can overlay similar information on the image of the world I see through the smartphone’s camera lens. And, as with VR images, AR images projected onto a street scene may be visible only to me, not to others standing near me and looking at the same street scene. As such, they too will be largely insulated from government control in a way that physical changes to property are not. Spraypainting graffiti onto a building’s surface can count as “defacing property” and would be subject to criminal punishment. But this is not true where the graffiti is part of an AR image that is superimposed on a building’s surface, as is the case in augmented reality apps such as Stiktu,10 or World Brush.11 Such AR graffiti leaves no physical trace on building walls or physical surfaces, and remains invisible to those not viewing the world through the lens of the AR app that projects it. This transformation of public real space to private virtual space in VR or AR thus has an important legal consequence: It largely eliminates the government interests that typically justify government control over the information we perceive in signs or artworks affecting buildings or other structures in our public environment. The community’s aesthetic interests no longer provide a basis for government interference in individual choice, because a visual scenario chosen by, and visible only to, that individual is not generally for the community to control. Just as individuals have First Amendment freedom to choose the books they read or the films they watch, so they almost certainly will usually have a First Amendment right to select what kind of information, or virtual images, to overlay on the surrounding environment using augmented reality technology. These observations tell us something about the First Amendment status of augmented and virtual reality. But they also tell us something about the First Amendment itself: The protection it offers for our aesthetic or information-seeking choices is typically strongest

  Andrew Webster, www.theverge.com/2012/6/13/3082734/stiktu-social-augmented-reality.   Lewis Painter, Best AR Apps for iPhone and iPad, Macworld, at www.macworld.co.uk/ feature/iosapps/best-ar-apps-for-iphone-ipad-3664321/. 10 11

BARFIELD PRINT.indd 306

13/11/2018 11:52

Freedom of expression, and the personalization of public space  307 when the object of choice is a world of private sensations or experiences, one that is limited to the individuals or to groups of people who have consented to share these sensations or experiences among themselves (for example, by using an AR app to view graffiti or other art they can render simply by using the app). By contrast, First Amendment protection for our art, expression, or information seeking is weaker when it requires altering the physical or social aspects of a brick and mortar landscape that we share with the rest of society: This is clear in the cases permitting content-neutral regulation of the traffic or other interactions that occur in shared physical space.12 VR and AR can thus enlarge that portion of our experience that is strongly shielded by First Amendment protection, by moving many experiences from a public shared version of public space to a customized and personalized version of it. VR typically does so to a greater degree by substituting all of public space with a virtual reproduction of it. AR, by contrast, moves some features of our public experience into a realm of private perception—like the graffiti on a bank’s exterior wall that we see through AR glasses but which remains invisible to bystanders who look at the same wall with only their natural vision. But some features of this experience—like the wall itself—are not private in this sense: They are part of a brick and mortar reality that continues to exist beneath the AR augmentation and that we share with the rest of the public. But it would be wrong to conclude that VR and AR additions or transformations to public space are therefore wholly beyond the government’s reach—and the gaps or uncertainties in First Amendment protection of VR and AR also tell us something about First Amendment law. In short, AR and VR never entirely remove us from the realm of shared social interaction. This is particularly true of AR, which may affect the way we move through a shared world, and our capacity to do so safely. Government may concern itself with and regulate these real-world effects of AR. For example, government is expected to protect drivers and those around them from the risk of automobile accidents. And such accidents may become more likely when drivers are distracted by AR images superimposed on the road – on smartphone screens, AR glasses, car dashboards, or perhaps a technologically enhanced car window. Similarly, where AR directions superimposed on a screen lead individuals to create a nuisance for a private property owner, the property owner might argue that it may target not only the game players, but also the AR creator. This in fact was the theory behind a class action lawsuit filed against the company Niantic for a nuisance allegedly created by its AR game Pokémon Go.13 Apart from causing individuals to take action that is harmful to people or property, AR or VR may also come within government’s regulatory sphere when it provides information that misleads or confuses people about individuals or businesses. False and insulting information about an individual may give that person a basis to bring a defamation claim, whether that information takes the form of a statement on a blog, or appears in a paragraph that floats (within a VR or AR environment) in front of the person’s face as onlookers view him. An AR advertisement or display might likewise violate another

  See text accompanying notes 73–86 infra.   See Andrew L. Ross, Gotta Catch. A Lawsuit? A Legal Insight into the Intellectual, Civil and Criminal Battlefield “Pokemon Go” Has Downloaded onto SmartPhones and Properties Around the World, 16 J. Marshall Rev. Intell. Prop. L. 329, 345–9 (2017). 12 13

BARFIELD PRINT.indd 307

13/11/2018 11:52

308  Research handbook on the law of virtual and augmented reality company’s trademark or trade dress. Imagine, for example, that a new, little known burger chain pays the developers of an AR app to project the distinctive yellow McDonald’s “M” logo over or near its locations, drawing customers who are seeking McDonald’s restaurants but will perhaps be willing to settle for the lesser known burger chain once they’ve arrived there. It would likely be an unauthorized use of McDonald’s trade dress if the unknown burger chain displayed a physical yellow “M” above its stores, and courts may find that the AR display of the “M” is a trade dress violation as well. In fact, in all of the above cases, there is little novel in the way AR or VR interact with First Amendment law: Government can create civil liability for defamation and act against deceptive commercial speech, and it can do so whether that defamatory or deceptive speech appears in AR or VR, in an email, or in a pamphlet or letter. Still, considering AR and VR examples helps to underscore that the leeway the First Amendment provides for government to regulate such speech does not always depend on whether the speech is or is not isolated from the brick and mortar world: Damage might be done by the falsity or deception in this speech even when it occurs in the virtual realm, isolated from the physical world outside. Moreover, it is also useful to understand how, in other circumstances, AR and VR might well make a difference to whether, and how much, free speech law protects certain expression. It might be more difficult for the state to show it is engaging in permissible “content-neutral” regulation of speech in an AR or VR version of public space than in the more familiar public space that is shared with others. In a public square in the physical world, the government may claim, for example, that it is justified in placing limits on the time, place, and manner of a protest march in order to ensure that such a march does not block traffic. In a purely virtual world, however, there need be no real traffic—and whatever virtual cars there are might simply fly over the protest, or teleport around it. The “mixed reality” one finds in AR is less fully removed from a shared public space where government must regulate traffic, noise, and otherwise assure that we respect each other’s interest in free movement, shared access to public space, and community aesthetics. But by personalizing at least some features of this public space, AR might at least reduce the potential for social friction that much content-neutral regulation by government is intended to address. Justifying speech restriction as “content-neutral”—as steering clear of censoring ideas and instead focusing on speech’s physical effects—might be a harder sell in virtual environments that are largely built from ideas or perceptions, with less ­territory occupied by physical buildings and roads. AR and VR can also perhaps, in some cases, enhance rather than weaken the government’s case for intervention in speech. Government has greater leeway under existing First Amendment doctrine to regulate certain kinds of speech linked to violence or sexual excitement, or speech that might constitute a tort that undermines a person’s reputation or privacy or freedom from economic harm. It is conceivable that immersive and threedimensional versions of such speech may sometimes intensify the purported harms that justify giving government greater power to censor such speech: A misleading portrayal of a commercial product in a real-seeming virtual image might be more misleading than a misleading description in words. Where 3D games allow one to beat and injure an avatar that is the spitting image of a real person, this may strike courts as likely to cause greater emotional distress, or to constitute a greater offense to privacy, than the protected First Amendment acts of writing a fictional story about that person or drawing a picture of

BARFIELD PRINT.indd 308

13/11/2018 11:52

Freedom of expression, and the personalization of public space  309 him. As Mark Lemley and Eugene Volokh write, “[w]e base many rules on the distinction between the mental and the visceral, between things we perceive and things we experience,” and VR and AR can change the way we understand and apply this distinction.14 Finally, there is one other potential intersection this chapter will briefly examine between First Amendment law and the way that AR and VR transform public space. I have been assuming, in the above discussion, that AR and VR provide information about, or superimpose artistic images (like graffiti) onto, a shared public space that continues to exist in a real, physical form outside of our virtual experience. The graffiti may be virtual, but the building upon which it is superimposed physically exists. The street signs we look at may be electronic images, and, on a computer running Google Earth, the street itself may be an image on a computer screen rather than a physical space through which we can move our bodies. But there continues to exist a real street outside of Google Earth, and our motivation for viewing it on Google Earth or Google Street View is often to preview, and plan for, a visit to the physical version of the street—to make a purchase, tour a neighborhood, or visit a friend. The First Amendment ground rules may be different, however, if virtual space does not continue to exist alongside a physical space it simulates (and allows us to plan on visiting) but instead fully replaces it. In the future it may be that the only space in which we can visit a certain store or congregate with certain friends or likeminded citizens is a virtual public square, because there simply is no physical space that can serve as a platform for these particular activities. Even where we continue to walk on a physical street, it may be the case, in the future, that the only image most people will see of a particular building or a house is the one they see through a universally adopted AR application. This is already true of certain commercial or social interactions that cannot easily occur outside of Web pages, internet chatroom, or other virtual spaces: Some purchases or communications take place only in these electronic spaces. In such cases, First Amendment questions may arise about how the First Amendment rules that govern physical spaces translate to virtual spaces that fully displace them, and take over their functions. The final part of this chapter will thus consider whether the First Amendment “public forum doctrine” that has evolved to constrain, but also in some ways empower, government managers of public space might work when that public space is made of virtual or augmented reality rather than the stuff of the physical world. By looking at how virtual and augmented reality can transform public space—and the First Amendment difference this transformation makes—we can thus learn something about the legal implications of these technologies. We can also learn something about the extent to which the strength of First Amendment protection does and does not depend on the degree to which our activities are drawn into a realm of pure perception and in turn removed from the physical realm where we can more easily injure and interfere in each other’s lives. Under the First Amendment, it is up to each of us what aesthetic content we wish to view and enjoy—for example, in photographs, films, or video games, and also in virtual and augmented reality environments. But it is, to some extent, up to lawmakers and government officials what we can safely do while driving, how we can go about marketing products or services (and competing with others who do so), and how we engage in certain speech which, although it takes the form of words, can nevertheless   Lemley and Volokh, Law, supra note 9 at 4, 11.

14

BARFIELD PRINT.indd 309

13/11/2018 11:52

310  Research handbook on the law of virtual and augmented reality do physical or financial harm to others, or cause other harms that might justify limits on speech. Some uses of VR, or more likely AR, can spill over into the latter realm of speech that has worrisome physical, financial, or other real-world effects beyond simply altering people’s beliefs and feelings. This chapter will thus explore what uses of AR or VR images are insulated against government control under the First Amendment, which are fair game for government regulation, and what lessons this exploration of AR and VR’s status has for First Amendment law. As should be clear from the above introduction, my focus here is not on free speech protection generally, but rather on the specific form that free speech protection takes in American constitutional law.

2.  VIRTUAL AND AUGMENTED TECHNOLOGY It is useful to begin by briefly summarizing virtual reality and augmented reality and the differences between them. “Virtual reality” refers to the use of technology—generally, computer technology—to substitute a person’s experience of the physical world with the experience of another computer-generated world that does not physically surround her but only appears to do so. When I walk through New York City in physical space, I see buildings rising high on either side of me and I can be sure that the images are coming from brick and mortar structures. I might see these same buildings in virtual reality, thanks to a head-mounted display that covers my eyes and projects images from a computer. But while a powerful VR reproduction might leave me with the feeling that the buildings are just as much in front of me as they are when I’m physically standing beside them, in this case the images of the buildings do not result from the presence of the buildings themselves, but are merely a computer simulation. I might, as I look upon this New York architecture, be physically located thousands of miles from New York City in a rural location without a skyscraper nearby. By contrast, augmented reality or “AR” does not shut out the physical world around me and replace it with an entirely new artificial experience. Rather, as noted previously, it superimposes information over the light that continues to come through from the surrounding environment—through hightech glasses, a smartphone, or some other electronically supplemented window I can use to view that environment. Both VR and AR have been in existence for decades. Indeed, more primitive equivalents of these technologies have existed for centuries: Anticipating computer recreations of faraway sites in VR, designers have created small replicas of cities or towns.15 Anticipating the way people overlay information on the world with AR, maps and astrolabes have respectively allowed people to link a mountain they see in front of them with facts described on a map, or a star they view through an astrolabe with the name and celestial coordinates of the star. In the nineteenth century, stereoscope makers found ways to add depth to two-dimensional pictures, making images of buildings or mountains seem to appear to rise in front of a person rather than remain confined within the corners

15   See, e.g., Madurodam, Atlas Obscura, describing a miniature city built in 1952 in Scheveningen, The Hague, that is a “model of a Dutch town on a 1:25 scale, composed of typical Dutch buildings and landmarks.” www.atlasobscura.com/places/madurodam.

BARFIELD PRINT.indd 310

13/11/2018 11:52

Freedom of expression, and the personalization of public space  311 of a photograph. And in the mid-twentieth century, film technologists such as Waller and Morton Heilig gave movies a three-dimensional appearance by using screens that surrounded the user, similar to the ways in which Omnivision films and certain kinds of Cave Automatic Virtual Environment (“CAVE”) rooms do now.16 The most significant development in AR and VR technology, however, came when engineers turned to computers to generate VR simulations, or AR supplements to realworld perceptions. In the late 1960s, Ivan Sutherland developed the first head-mounted display.17 Thomas Furness developed similar technology as part of a flight simulator he was creating.18 And in the subsequent decades, other engineers have refined the technology so that it can better track individuals’ head movements and otherwise react to what a person does. They have also added auditory stimuli, and in some cases gloves, body suits, or other “haptic” technology that can give individuals the sense of touching the simulated environment and feeling resistance when they push against it.19 In 2018, as I write this chapter, this technology is not only much better than it was a decade ago but is also part of a huge and widely reported push by technology companies to make AR and VR a familiar part of consumers’ everyday existence. While VR was once confined to laboratories, Oculus Rift and HTC Vive, and more recently Sony, have developed virtual reality gaming consoles that are affordable for some consumers.20 In addition to the option to play immersive 3D versions of video games where the player shoots zombies or aliens, or pilots a spaceship, other 3D applications allow individuals allow users to create three-dimensional paintings in the surrounding space,21 or to view a dark, starfilled (virtual) sky that would normally be invisible in an urban location due to cloud cover, smog, or light pollution.22 AR is also now being touted as a technology that will soon become much more common—and much more useful—in individuals’ lives. Indeed, in 2016 AR was at the center of a cultural phenomenon: the massively popular smartphone game Pokémon Go, by Niantic. Players of those games had to hunt and capture distinctive cartoon monsters, called Pokémon, that could be found superimposed over locations in the surrounding environment. Pokémon were programmed to appear over certain locations as a person viewed them on a smartphone screen. While thousands of players became happily engrossed in playing this AR game, complaints arose from homeowners who saw crowds of peoples descend upon or near their property in search of Pokémon monsters—and law enforcement warned that distracted Pokémon players were endangering themselves, 16   Howard Rheingold, Virtual Reality: The Revolutionary Technology of Computer-Generated Artificial Worlds—and How it Promises to Transform Society, 355 (First Touchstone ed. 1992) (1991). 17   Tobias Hollerer and Dieter Schmalstieg, Introduction to Augmented Reality (2016). 18   Fred Moody, The Visionary Position: The Inside Story of the Digital Dreamers Who Are Making Virtual Reality a Reality xxiii (1999). 19   Ken Pimental and Kevin Texeira, Virtual Reality: Through the New Looking Glass 42–5 (2d ed. 1994). 20   Stuart Dredge, The Complete Guide to Virtual Reality—Everything You Need to Get Started, The Guardian, Nov. 10, 2016. 21   Frank Rose, The Making of Virtual Reality Art with Google’s Tilt Brush, NY Times, Jan. 4, 2017; James Bricknell, PaintVR Brings 3D Paintings to Life, VRHeards, Sep. 19, 2017. 22   Team develops virtual reality edition of Stellarium for Oculus Rift, Phys.Org, Jul. 7, 2015.

BARFIELD PRINT.indd 311

13/11/2018 11:52

312  Research handbook on the law of virtual and augmented reality and possibly others, by focusing only on the small portion of the surrounding environment seen on their smartphone screen instead of attending fully to their situation.23 Two distracted Pokémon players reportedly fell over a cliff in California (fortunately, they survived).24 Other AR apps have likewise seen some adoption by smartphone users and in 2017, the largest and most visible makers of mobile devices—Apple, Google, and Microsoft—each made determined efforts to make AR much easier to incorporate into smartphone and tablet applications, or other computer technology. In its late 2017 release of the iOS 11 operating system, Apple included a program called ARKit that allows developers to easily incorporate AR into iPhone and iPad applications.25 As a result, there are now many more iOS games that, like Pokémon Go, incorporate AR images; many more tourism and mapping applications that overlay information about sites of interest; and also new commercial applications. Furniture companies, for example, now offer AR technology allowing one to insert the image of a couch in a room in order to see how the physical equivalent would look if it were purchased and placed there.26 Google has similarly enabled Android developers to create AR applications with its ARKCore platform.27 Microsoft is enhancing AR applications, with technology such as the Hololens—a headset that can be worn over one’s eyes and project images that blend into the scene one is looking at.28 And the company Magic Leap announced that in 2018 it would release a widely anticipated AR headset that uses what it calls “lightfield technology” to create “virtual animations, objects and characters [with] a sense of depth.”29

3. FIRST AMENDMENT COVERAGE AND VR AND AR TECHNOLOGY It is likely that—just as films, photographs, video games, and other media for disseminating information or aesthetic experience have already done—AR and VR will give rise to First Amendment battles over whether, and how much, the content within them is protected from government censorship or restriction. In fact, a battle over how the First

23   See Ross, supra note 13; Eriq Gardner, Will a Class Action over “Pokemon Go” Redefine What It Means to “Trespass”?, Hollywood Reporter, Mar. 27, 2017, www.hollywoodreporter.com/ thr-esq/will-a-class-action-pok-mon-go-redefine-what-means-trespass-989042. 24   Veronica Rocha, 2 California Men Fall Off Edge of Ocean Bluff while Playing “Pokemon Go,” Los Angeles Times, Jul. 14, 2016. 25   Nick Statt, Apple Shows Off Breathtaking New Augmented Reality Demos on iPhone 8, The Verge, Sep. 12, 2017. 26   Ariella Pardes, Ikea’s New App Flaunts What You’ll Love Most About AR, Wired, Sep. 20, 2017; Lauren Goode, Houzz Has a New AR Mode that Lets You Try Furniture before You Buy, The Verge, May 3, 2017. 27   Ron Amadeo, Google’s ARCore Brings Augmented Reality to Millions of Android Devices, ArsTechnica, Aug. 29, 2017. 28   David Meyer, Microsoft Just Announced a Major Expansion of Its Augmented Reality, Gear, Fortune, Nov. 2, 2017. 29   Hugh Langley, Magic Leap One: All the Details on the Mysterious Mixed Reality Glasses, Wareable, Dec. 20, 2017.

BARFIELD PRINT.indd 312

13/11/2018 11:52

Freedom of expression, and the personalization of public space  313 Amendment applies to AR is already taking place: As I write this, the developer of the AR game Texas Rope’Em is pursuing a First Amendment challenge against a Milwaukee ordinance that requires a permit for anyone to use “augmented reality” within a municipal park.30 This and other challenges will require courts to confront one or both of two questions about First Amendment speech protection: the question of free speech “­coverage,” and the question of free speech “protection.” 3.1  Coverage and Protection in Free Speech Law As Frederick Schauer explains in his seminal formulation and elucidation of that distinction in free speech law, we might conceive of the difference between coverage and protection by imagining the First Amendment’s free speech protection as equivalent to a knight’s armor.31 Such armor covers the knight, and thus provides some protection for his body from external threats; but, as Schauer notes, it “does not protect against everything.”32 A rock will likely fail to penetrate it, but gunfire will easily pass through and injure or kill the knight. A legal right, says Schauer, works in the same way: Each right will cover certain human conduct, but not all of it. The right to freedom of speech, for example, covers much of the expression we engage in with words, with art, or possibly through other means. But it does not cover many other human activities, such as driving a car or selling or buying an item. But even if we know a kind of conduct is covered by the right of free speech, that does not tell us how much protection it receives. Freedom of speech may cover certain kinds of verbal expression, but the protection affords it may be strong, weak, or somewhere inbetween. Moreover, its strength is not only a function of the legal armor itself but also a function of the nature of the threat with which it is faced: Just as a knight’s armor can protect against rocks but not bullets, the legal armor of free speech may protect against government regulation with paternalistic aims (where the state purports to determine what it is good for us to think or hear), but it may fail to protect against government regulation for other purposes, such as assuring a vital national security interest. In short, then, coverage is a matter of whether the constitutional armor is there and providing any protection at all—something it does, in the free speech context, if the activity in which a person is engaging is speech (or possibly other First Amendment activity, such as receiving speech or forming the beliefs that might later find expression in it). Once it is established that an activity is covered by First Amendment “armor,” the next question is what level of protection it receives from that armor. In modern First Amendment law in the United States, such protection generally takes the form of a specific tier of judicial “scrutiny.” When courts make free speech armor around expression so strong that almost no government regulation can pass through, such protection is something courts refer to as “strict scrutiny.” It is only a government interest of the most extraordinary weight, a socalled “compelling government” interest, that can justify speech regulation here. And even then, courts demand that government do no more damage to speech than is necessary. At

  See Candy Lab, Inc. v. Milwaukee County, 2017 WL 3088380 (E.D Wis. Jul. 20, 2017).   Frederick Schauer, Free Speech: A Philosophical Inquiry 89 (1982). 32   Id. 30 31

BARFIELD PRINT.indd 313

13/11/2018 11:52

314  Research handbook on the law of virtual and augmented reality times, however, government has more need to regulate how speakers express themselves: Where a protest might block traffic, for example, or interfere with others’ use of a street, government is generally allowed by courts to regulate the “time, place and manner” of protestors’ speech. Courts still scrutinize such limits on when, where, and how individuals express themselves, but their scrutiny is only “intermediate,” not “strict.” Government oversight and management of shared public spaces is necessary, so courts weaken the free speech “armor” so that government is not shut out entirely: Any “substantial” or “significant” government interest can justify government intervention and government is permitted to pursue those interests as long as it speech limits are not substantially more excessive than is necessary to accomplish the government’s objectives. Applied to VR or AR, the coverage question asks whether a particular kind of VR or AR content, or interaction with such content, counts as “speech” under the First Amendment—and thus qualifies for any protection at all. Are we engaging in speech protected by the First Amendment when, in a VR environment, we shoot at imaginary zombies with a simulated gun? Are we engaged in such speech when we use an AR application on a smartphone to “shoot,” from within the app, at people in front of us with a cartoon “gun” that is visible only on our smartphone screen? If the answer is “no”—if courts rule that such simulated shooting is not verbal communication, artistic expression, or any other kind of activity covered by the Constitution’s free speech protection—then lawmakers will be able to restrict it without worrying about any First Amendment limits. If, on the other hand, such simulated shooting is First Amendment expression, then lawmakers may still be able to restrict or regulate it, but only if they can overcome the constitutional hurdles which First Amendment law places in their way to ensure they do not unnecessarily censor speakers without sufficient justification. 3.2  Straightforward Coverage Cases in VR and AR: Speech and Nonspeech Conduct For many VR and AR activities, the coverage question is likely to be easy for courts—and likely to receive a positive answer (that the First Amendment does protect that AR or VR activity). First, many VR and AR activities already entail communicating ideas in ways that are already protected by free speech law, no matter what the setting. Consider, for example, a person’s statement—in a particular conversation or to the public at large—that “the President of the United States should be impeached,” or that “the Republican Party deserves our support.” Both of these statements are protected speech. They are not only protected when a person voices them in the open air while standing at a street corner; they are just as staunchly protected when they are stated in ink on a letter sent by oldfashioned mail service, in an email, or in a Web-based blog. And it is clear that they would be just as protected if they were voiced or written in a virtual or augmented reality environment. The same is true of many examples of speech that lacks a political character. If I provide somebody directions to a church service,33 and the government intervenes because it wants to stop individuals from receiving this information, it is violating the First Amendment—and it will do so whether my directions are placed on a cardboard sign erected on my lawn, posted on a website, emailed to prospective churchgoers, or placed 33

  See Reed v. Town of Gilbert, 135 S. Ct. 2218 (2015).

BARFIELD PRINT.indd 314

13/11/2018 11:52

Freedom of expression, and the personalization of public space  315 upon a sign one sees in a virtual reproduction of public space, or in an augmented reality layered over public space as people look at it (through smartphones or smart glasses). For this reason, a digital map would likely receive First Amendment protection against a law or regulation aimed at suppressing information on it—and this should be true in VR and AR as well as on an app or a Web-based mapping application that simply displays mapping information on a flat screen. Similarly, where VR or AR serves as a medium for displaying photographs, paintings, or animations to audiences interested in them, such sharing of art would be just as covered by the First Amendment as sharing of photography or art on canvases or on the World Wide Web, or sharing of animations on film reels or on computer video software, such as YouTube or RealPlayer. I will here refer to the two types of expression I have just described—words that convey meaning and instances of a recognized form of art—as “elements of expression,” because such words and images generally take on an expressive role when they are part of some larger social practice, such as people’s practice of having verbal conversations, reading books to learn about the world, or viewing films, photographs, paintings, or other arts.34 The same words and images, however, can also serve nonexpressive functions in other circumstances. Consider, first, an example outside of AR or VR. The phrase “AstrosFan” could very well be expressive in a blog, for example, if I used it as a signature to a sportsthemed blog post, or comment to such a post, to indicate that I was a fan of the Houston Astros baseball team. However, if someone uses the same phrase as a (not very good) confidential password for an app or Web service, the phrase is not in this case serving as an element of an expressive practice (namely, an exchange of opinions about sports). Rather, it is an element of a social practice generally considered to be nonexpressive, wherein it serves as an electronic key of sorts to generate access to a protected digital space. Some words or images in a VR and AR environment may well operate in the same way. They may be elements of expression when they occur in certain social practices involving VR or AR—such as that of posting augmented reality graffiti or other artwork for other AR users to admire, or creating maps to provide tourists or curious locals with information about their surroundings. But in other circumstances, the same word or image projected onto a smartscreen in an AR software application is not functioning as a word, phrase, or image that conveys information, nor as a picture that an AR user might find aesthetically pleasing. It is rather an AR image that one can interact with, and, in doing so, trigger some action that is at least partly nonexpressive in character. For example, imagine that when a person starts a particular app on their smartphone, and then points it at a fancy restaurant, a seating map of the restaurant emerges that will allow the user to reserve a particular table for dinner. A set of photos of daily specials then emerges, allowing the user to enter a food order. Here, the maps and photographs arising in such an AR experience are not merely conveying information. Nor are they there simply to be admired. They are tools for engaging in a nonspeech act – namely, reserving a space in a restaurant and ordering food as part of a commercial exchange.35 Or imagine that a set of AR images or animations arises when one points a phone toward a certain locked door, or perhaps a closed turnstile at a

  Robert Post, Recuperating First Amendment Doctrine, 47 Stan. L. Rev. 1249, 1279 (1995).   It is likely that some of the communication that occurs immediately preceding, or during,

34 35

BARFIELD PRINT.indd 315

13/11/2018 11:52

316  Research handbook on the law of virtual and augmented reality transportation station, and it allows one to then react to the image in a way that can open the door or the turnstile. In such a situation, the AR image is part of a social practice that is generally not considered primarily expressive: As in my “AstrosFan” password example above, an image that can serve as an expressive element is instead an element in the process of creating access to an otherwise locked off physical space. In short, then, sometimes the question “is a use of AR or VR covered by the First Amendment” can be answered by pointing to characteristics of the activity that have little to do with the AR or VR technology being used to make it possible. Certain uses of VR or AR—for example, to convey words or images critical of the President—clearly count as First Amendment expression. Such uses are just as much First Amendment speech as is political expression in a book, film, or blog. Other uses of AR or VR technology—for example, to provide a means of exchanging virtual currency with real value—clearly would not count as First Amendment expression: government can regulate what kind of physical items (dollar bills, coins) can count as legal tender and how that tender may be used to pay for items, and in doing so it is not generally subject to First Amendment limits. There is no reason why the constitutional ground rules should be any different for virtual currency. 3.3  Difficult Coverage Questions in VR and AR: The Example of AR Gameplay But in other circumstances, it is not as easy to say whether a particular use of words or images in AR or VR would count as expression. In some cases, perhaps, this is simply because individuals may seek to express themselves without words or recognized artistic practices, and instead try to communicate with means that are normally noncommunicative. Consider an example from outside VR and AR: people might set fire to an action figure portraying a government official or presidential candidate—and even if setting fire to a toy is normally not a means of sending a message, it may be one here. Of course, a person who is charged with destruction of property, or other illegal action, might well seek the cover of the First Amendment by portraying anything they do as containing some kind of message. But the Court has made it clear that it “cannot accept the view that an apparently limitless variety of conduct can be labeled ‘speech’ whenever the person engaging in the conduct intends thereby to express an idea.”36 Rather, not only (1) must that person intend to express a particularized message in taking certain conduct, but also (2) that message must be likely to be understood by an audience given the context.37 Burning an action figure, in other words, only counts as First Amendment expression when (1) it is intended to convey a message of protest or some other “particularized” message and (2) the burning will be understood by onlookers to convey that message. This two-pronged test for when nonspeech conduct can function as speech is called the “Spence test” because the Court proposed it in the case of Spence v. Washington. The Spence test, however, is useful only for answering a very limited question—namely, when does a practice that is typically noncommunicative (such as burning a toy) come

a commercial exchange will receive the First Amendment protection accorded to “commercial speech.” But the selling or purchase itself generally counts as a non-speech act. 36   United States v. O’Brien, 391 U.S. 367 (1968). 37   Spence v. Washington, 418 U.S. 405, 409–11 (1974).

BARFIELD PRINT.indd 316

13/11/2018 11:52

Freedom of expression, and the personalization of public space  317 to be communicative because of features that make it function in that way for both the speaker and her audience? That question may well arise in AR or VR. For example, imagine that a computer hacker finds a way to reconfigure Google Earth, or some other virtual globe, and programs it to replace the Trump Tower’s image with one of the Kremlin (while putting the virtual Trump Tower in Moscow, where the Kremlin actually stands). Replacing one building with another is not typically a communicative act, and we should not, perhaps, rush to assume it would be in even when the buildings, and the cities they are in, are virtual. But if, given the context, users of the virtual world understand this to be a protest against President Trump, by an activist (and hacktivist) who wishes to implicitly accuse him of colluding with Russia, then such a building switch may count as speech under the Spence test. The hacking may still well receive criminal punishment under the Computer Fraud and Abuse Act. But it will not be because the act is free of expression, but rather because, in this case, it is subject to content-neutral laws criminalizing computer hacking. In any event, the most significant First Amendment coverage complication in analyzing AR and VR is not the kind that can be answered with the Spence test. It arises not when AR or VR users employ a nonspeech practice (like swapping one virtual building with another) to communicate a message, but rather when AR or VR users or designers claim that a certain use of those communications is inherently expressive, like art or photography, and should receive First Amendment protection even where there is no particularized message. This question is especially likely to arise when an individual uses AR technology to play games, or otherwise explore real-world settings. Consider, for example, an AR game where individuals seek to “shoot”—with video game-like virtual guns, or missiles, or other shooting devices—at certain targets. In some AR games, these targets may be images of monsters or zombies superimposed on their view of the real world. In Pokémon Go, for example, players aim “Pokéballs” at the cartoon creatures who populate the world of that game.38 In the smartphone game The Walking Dead: Our World, players will shoot at zombies called “walkers” that appear “in the street, in the park, on your sofa, whenever and wherever you feel like it.”39 Companies are increasingly using AR to develop scavenger hunts, or to enhance sightseeing experiences. In other, sometimes more troubling, scenarios, the targets at which one “shoots” in an AR game are not cartoon monsters or zombies or any other fictional entities generated by the AR itself—they are people who are actually appearing in the field of view of one’s smartphone screen or other AR viewing device. What augmented reality generates here is not a virtual target, but rather a virtual bullet or other virtual weapon to simulate an attack against a person who is really there. In 2013 a Louisiana teenager was arrested after he used a mobile phone application called “RealStrike” to simulate shooting at his high school classmates, and then posted a YouTube video showing his simulated shooting.40 Police and school authorities were understandably worried that such a video could be   John Davison, WTF Is “Pokémon Go,” Explained, Rolling Stone, Jul. 11, 2016.   James O’Connor, The Walking Dead: Our World AR Game Will Have Us Shuffling Around Staring at Our Phones like Its Eponymous Zombies, VG24/7, Aug. 30, 2017. 40   Caroline Moss, A Teenager Was Arrested after Using an iPhone App to Go on a Virtual Shooting Spree, Business Insider, Sep. 16, 2013. 38 39

BARFIELD PRINT.indd 317

13/11/2018 11:52

318  Research handbook on the law of virtual and augmented reality a prelude to actual violence—and, for reasons explained later in the chapter, it is quite possible that government could make it illegal, or at the very least a violation of school rules, to use AR in this way, on the ground that the AR video could disrupt a school’s functioning or count as a threat of violence. But the first question it makes sense to ask is whether the use of the AR technology to engage in simulated shooting in a real-life setting counts as First Amendment speech at all. The same question could be raised by similar games where the people depicted as targets are not oblivious to the AR user’s actions, but rather willing players (who may also be doing augmented reality “shooting” themselves). For example, the LyteShot augmented reality gaming platform allows individuals to use smartphones in an AR version of the game Assassin.41 As Wikipedia explains, Assassin is a game in which “players try to eliminate one another using mock weapons, in an effort to become the last surviving player.”42 The game has long been played in physical real-world settings, such as college campuses, and players typically eliminate another player either by touching him with the mock weapon or pointing that weapon at the target in a way that is understood by all players to cause elimination.43 In the AR version of the game, the players eliminate each other by pressing a button on an AR app on their smartphone. First Amendment coverage questions might also arise about uses of AR to supplement scavenger or treasure hunt games. An Austin, Texas tour company, Austin Cluetivity, offers “an interactive GPS scavenger hunt and tour,” where participants use iPads equipped with augmented reality applications to meet challenges that allow them to learn about parts of Austin or play games similar to “room escape games.”44 Google has created a game called Ingress, wherein teams of players compete to capture energy-emitting “portals” by seeking these augmented reality items “in a variety of creativity-oriented public places—libraries, museums, and public parks.”45 There is also a variety of Web-based blogs and courses instructing software creators (or app users) on how to add augmented reality to scavenger hunts or to “geocaching” challenges, wherein individuals search for hidden objects using lists of global position system (GPS) coordinates.46 Is the playing of these AR games—which, in a sense, extend video game play or puzzle solving into a real-world setting, with real-world movement—First Amendment expression? If the game as a whole is not First Amendment expression, might any components of it nonetheless count as First Amendment expression—so long as they are confined to looking at objects or words on a smartphone screen or on the surface of one’s glasses, and do not involve physical movements or other physical action? These questions have only just begun to receive authoritative answers from courts. In July 2017 the United States District Court for the Eastern District of Wisconsin heard a claim, brought by AR game creator Candy Lab, asking for a preliminary injunction to stop Milwaukee from enforcing a new ordinance requiring AR and VR developers to gain a permit before they “introduce a location-based augmented reality game into the   See Lyteshot, at www.lyteshot.com/gamers/.   Assassin (game), Wikipedia, at https://en.wikipedia.org/wiki/Assassin_(game). 43   Id. 44   See Austin Cluetivity, at www.austincluetivity.com. 45   See Google’s Ingress Mobile Game Combines Geocaching, Augmented Reality, and Puzzle Games, MapDash, n.d. 46   Dmitry Konig, Geocaching in Augmented Reality, Medium, Dec. 6, 2016. 41 42

BARFIELD PRINT.indd 318

13/11/2018 11:52

Freedom of expression, and the personalization of public space  319 Parks.”47 More specifically, the Milwaukee ordinance states that “[v]irtual and locationbased augmented reality games are not permitted in Milwaukee County Parks except in those areas designated with a permit for such use by the Director of the Department of Parks, Recreation, and Culture (the “DPRC”).”48 To obtain a permit, AR and VR developers wishing to “introduce” any such games into the park must submit an application providing specific information about the game to the DPRC, which will then conduct “an internal review . . . to determine the appropriateness of the application based on site selection, protection of rare flora and fauna, personal safety, and the intensity of game activities on park lands.”49 As the federal district court noted, the application itself is a ten-page “special events” application used not only for VR and AR games, but for other group events requiring permits from the city. It requires an applicant to provide significant information including “estimated attendance, location within the park, event dates and times, a site map, and whether and how the event will be advertised.”50 It also “requires detailed plans for garbage collection, on-site security, and medical services, and warns that applicants will be responsible for these services,” as well as an affirmation that organizers of events have, and can demonstrate to the DPRC that they have, liability insurance able to cover damage to the park should it occur. In order to determine whether to grant the preliminary injunction that Candy Lab requested to stop this permit requirement from going into effect, the federal district court had to decide, among other things, whether Candy Lab was likely to succeed in its claim that Milwaukee’s permit requirement violates the First Amendment’s freedom of speech guarantee. The court was not, at this stage, making a final determination of whether such a violation occurred. Rather, it was determining whether—while proceedings were under way to answer that First Amendment question—Milwaukee should be enjoined from enforcing its ordinance, or allowed to do so. Still, Candy Lab’s likelihood of ultimately succeeding in its First Amendment claim was one of the factors the court looked at to determine whether to issue a preliminary injunction against the ordinance, and the court found that the likelihood of success was high. It relied heavily on the Supreme Court’s opinion in Brown v. Entertainment Merchants Association.51 There, the Supreme Court held that a California law violated the First Amendment when it barred the sale of violent video games to minors. Video games, it said, were protected speech, not only because they communicate ideas with “many familiar literary devices (such as characters, dialogue, plot, and music),” but also because they allow communication of ideas through “the player’s interaction with the virtual world.”52 This description, said the federal district court here, also applies to augmented reality games such as Texas Rope’Em. Like a video game, noted the court, this AR game immerses users in a fictional world, and presents challenges they are to complete in that world: It places them in “a Western-themed virtual environment, complete with a Texas-themed game title, color scheme, and graphics, 47   Candy Lab, Inc. v. Milwaukee County, 2017 WL 3088380 (E.D Wis. Jul. 20, 2017); Milwaukee County Code of General Ordinances, §47.03(3). 48   Id. 49   Id. 50   Id. 51   Brown v. Entertainment Merchants Association, 564 U.S. 786, 790 (2011). 52   Id at 2741–2.

BARFIELD PRINT.indd 319

13/11/2018 11:52

320  Research handbook on the law of virtual and augmented reality allowing the player to corral favorable playing cards using an animated lasso.”53 It also “conveys ideas related to the Wild West and scavenger hunting to lend an air of excitement and novelty to a traditional card game.”54 This, said the Court, appears to give this AR game at least some expressive content: The game “conveys ideas related to the Wild West and scavenger hunting to lend an air of excitement and novelty to a traditional card game.”55 This position—that free speech protection should be extended to AR games on the ground that they are very like video games—is likely to be reached by other courts as well. But it is useful to consider more carefully why courts are likely to reach this conclusion, and why it is at least possible some courts will resist it (with respect to at least some elements of AR games). Note that this decision did not systematically consider—as at least one prior court, and one legal scholar, had suggested—the possibility that playing a video game may be outside of the First Amendment’s coverage even if designing a video game is inside of it.56 Like the Supreme Court in Brown, it seemed to assume that, just as the First Amendment gives individuals not only a right to create books and movies but also a right to read or watch the books and movies created by others, so it gives individuals a right to seek out and play videogames and not only to design them. But it is useful to look at this distinction a little more closely. The act of designing and publishing most AR games would almost certainly count as First Amendment speech: AR games such as Pokémon Go and Texas Rope’Em generally take the form of software that generates words and images on the screen of a mobile device when a person carries that mobile device to a particular location in the physical world, or uses it to view a particular scene in the environment. It is unquestionably First Amendment expression when a tour book author and publisher provide an individual with a catalog of paragraphs or images that the reader can relate to real-world sites, which that reader can view as she carries the book from place to place within a city or other setting. Imagine, for example, that a person buys a book with pictures or descriptions of the sites in Chicago’s history, and directions to get to each of these sites. A book of this sort, like those in Arcadia Publishing’s “Then and Now” series, might allow users to wander the streets of Chicago to see what certain present-day sites looked like in an earlier historical

  Candy Lab, Inc. v. Milwaukee County, 2017 WL 3088380 (E.D Wis. Jul. 20, 2017).   Id. 55   Id. 56   In a 2006 opinion, the United States Court of Appeals for the Sixth Circuit noted that “there are features of video games which are not terribly communicative, such as the manner in which the player controls the game.” James v. Meow Media, Inc., 300 F.3d 683, 696 (6th Cir. 2002). In his dissent in Brown v. Entertainment Merchants Association, Justice Breyer similarly noted that “[v]ideo games combine physical action with expression” and that “were physical activity to predominate in a game, government could appropriately intervene, say by requiring parents to accompany children when playing a game involving actual target practice.” Brown v. Entertainment Merchants Association, 564 U.S. 786, 847 (2011) (Breyer J dissenting). Professor Kevin Saunders argues that although the First Amendment protects the creation and design of the game, it shouldn’t protect playing because the player is not communicating with anyone when she plays­—she is simply manipulating a computer. Virtual Worlds—Real Courts, 52 Vill. L. Rev. 187, 196 (2007). 53 54

BARFIELD PRINT.indd 320

13/11/2018 11:52

Freedom of expression, and the personalization of public space  321 era.57 Or it might superimpose drawings or descriptions of fictional characters or events onto real-life photographs of Chicago neighborhoods. All of these depictions of real-life neighborhoods in a book would count as First Amendment expression. And they would also count as First Amendment expression in an electronic book one carries on one’s smartphone (whether as part of a library in a book-reading app, like Amazon Kindle, or as a standalone app designed solely for that ebook). Such an ebook would be very similar to an AR smartphone app that automatically displays old pictures of a neighborhood in Chicago whenever the smartphone is pointed at a location in that neighborhood, instead of leaving it to the smartphone user herself to make the effort to scan the ebook for that old picture. Similarly, the fact that AR imposes a fictional character over a certain Chicago scene when the software detects a certain “data tag” (based, for example, on the images in the scene or the phone’s GPS coordinates) does not make the drawing or display of that character any less expressive than it is when it has to be located by a user in a print book or ebook. If it is First Amendment expression to write and publish tour books or other materials designed for users who wish to link certain information to real-life settings, it is likely also First Amendment expression to create AR software that automatically displays that information when a mobile device user encounters these locations. However, this does not mean that those who play AR games are necessarily also involved in expression when they not only read words or view pictures but also enter certain physical spaces, such as parks or streets, and travel through them in ways that correspond to certain “moves” within the AR game—or, for that matter, a location-based game that involves only printed information. To take a simple example, imagine that a person is playing a scavenger hunt game that involves entering and taking pictures of certain listed locations in Chicago. A more hightech version of this scavenger hunt may dispense with the picturetaking and simply have a user carry a GPS-capable smartphone that automatically registers when someone has gone to a certain location on the list. An augmented reality version of this game may have the screen display a new puzzle piece (of several they have to collect and then fit together) only when the user points their smartphone at a certain architectural feature at each location on the list. Even if the words and images in these games count as First Amendment expression, that does not mean it is First Amendment expression to physically enter each listed site (especially where doing so would involve trespassing, for example, into a museum or other location after it is closed). As noted above, the Supreme Court held in 2011 that the playing of videogames, and not simply their creation, constitutes expression that the First Amendment protects. Video games, however, are generally played and experienced in a visual environment separate from the public space we find in streets and other shared social spaces. In all likelihood, VR games that one experiences with a head-mounted display would receive the same First Amendment protection as the Court has already granted to their less immersive, two-dimensional equivalents. AR gameplay, however, is arguably different because it entails moving through “real-world” physical spaces where government has the regulatory powers it lacks in the purely virtual spaces in one’s computer. Cities can, after all, require permits to engage in certain expressive activities, such as holding parades or

57   See Arcadia Publishing, Then and Now, at www.arcadiapublishing.com/series/then-and-nowboo​ks.

BARFIELD PRINT.indd 321

13/11/2018 11:52

322  Research handbook on the law of virtual and augmented reality music concerts, in a certain location. They can place limits on how many individuals can congregate in a public meeting room at a certain time. They might likewise impose certain limits on AR gameplay, or for that matter on more primitive scavenger hunt games, at least to the extent that these games involve physical action (such as movement across streets) that might cause accidents, lead to large crowds gathering, or otherwise affect safety and others’ ability to use the public space that is the setting for the AR or other game. While videogames are now clearly inside the First Amendment’s scope, thanks to the Supreme Court’s 2011 decision in Brown, the same is not true of the football, baseball, or ultimate Frisbee that people play on a grassy field rather than a videogame console. Although some scholars have argued that real-world athletic competition should sometimes count as First Amendment speech,58 when courts consider restrictions on athletic competitions—such as boxing or mixed martial arts—they generally do not appear to see such restrictions as speech restrictions (and don’t consider the possibility that such restrictions raise First Amendment problems).59 One might thus argue that where an AR game involves physical movement in a real-world setting it should, like a football, baseball, or Frisbee game, count as nonspeech conduct rather than expression. It is also important to point out that courts need not necessarily categorize AR games, or other uses of AR (such as use of it in navigation), as purely speech or purely nonspeech conduct. They might instead find that a particular use of AR involves a First Amendment hybrid of sorts—a mix of expressive and nonexpressive conduct. The most familiar example of such a combination of expressive and nonexpressive elements is expressive or symbolic conduct. When a person burns the flag or a draft card, they don’t use any words. They don’t write or say anything. However, when their burning of an item is meant as a protest, it often communicates a message—such as a message of political protest against America’s wartime action. That does not mean it is purely expressive: Burning something may send a message, but it also has effects in the physical world. It often destroys the item that is burned; it creates smoke that might affect individuals’ health or comfort; and there may be a risk that the fire will spread from the flag or draft card to surrounding structures. The Supreme Court thus created a First Amendment test—in United States v. O’Brien—that allows government to restrict the nonexpressive conduct side of this expressive conduct, so as to prevent property destruction or protect public safety, as long as it is clear that government is (1) not doing so as a pretext to suppress the message of protest entailed by the conduct and (2) is not restricting far more speech than is necessary to further (3) a “substantial” government interest—which is the type of interest it must show it has when it bars burning of a draft card or flag.60

58   See Genevieve Lakier, Sport as Speech, 16 U. Pa. J. Const. L. 1109 (2014); Howard M. Wasserman, Symbolic Counter-Speech, 12 Wm. & Mary Bill Rts. J. 367 (2004); Michael T. Morley, “Exceedingly Vexed and Difficult”: Games and the First Amendment, 112 Yale L.J. 361, 366 (2002). 59   As Lakier notes, “courts have largely rejected the possibility that spectator sports are expressive acts and therefore entitled to First Amendment protection.” She cites several cases where courts did do so, including Justice v. NCAA, 577 F. Supp. 356 (D. Ariz. 1983), where a court found football to be conduct rather than speech. Lakier, supra note 58, at 1114. 60   United States v. O’Brien, 391 US. 367, 377 (1968). The Court also added a fourth requirement, which is that the government restriction of an expressive conduct must not only accord with these three requirements but also otherwise be within the constitutional power of government (for

BARFIELD PRINT.indd 322

13/11/2018 11:52

Freedom of expression, and the personalization of public space  323 Courts could likewise treat the playing of certain AR games, or performance of other AR activity, as having both expressive and nonexpressive dimensions. In fact, a Second Circuit case has already done precisely this with respect to a different kind of software—software for decrypting copyright protections. In University City Studies v. Corley, the US government prosecuted Corley after he posted on a website “a copy of the decryption computer program ‘DeCSS’,” a program designed to circumvent an “encryption technology that motion picture studios place on DVDs to prevent the unauthorized viewing and copying of motion pictures.”61 He also published the object and source code of DeCSS at the end of an article about this program in a print magazine for “the hacker community.”62 The government claimed that because he had disseminated a copy of DeCSS, Corley was subject to punishment under the Digital Millennium Copyright Act, which imposed penalties not only on those who circumvented digital copyright protections in movies or other copyrighted works, but also on those who “traffick[ed] in a technology primarily designed to circumvent a digital wall.” The Second Circuit concluded that computer code was speech for First Amendment purposes, but it also said that sometimes, such computer code does not function as speech: “Unlike a blueprint or a recipe, which cannot yield any functional result without human comprehension of its content, human decision-making, and human action, computer code . . . can instantly cause a computer to accomplish tasks and instantly render the results of those tasks available throughout the world via the Internet.”63 In this respect, it said, computer code had a double character: It was not only a form of communication, but was also “like a skeleton key that can open a locked door, a combination that can open a safe, or a device that can neutralize the security device, attached to a store’s products.”64 AR applications might similarly have a double character. The images or words one finds in an AR mapping program, for example, may sometimes communicate interesting information to an interested smartphone user. They can tell her about the history of a park she is wandering through, or a statue she is viewing within that park. On the other hand, the same AR images or words might, in other circumstances, not simply provide information, but—like the DeCCS program in Corley—serve as a trigger for actions in the physical world, or in digital space. We can thus recast, and in some respects qualify, the observation I made at the beginning of this chapter: that when a street sign or graffiti appears, thanks to AR or VR, in a private and personalized version of public space, the First Amendment places it beyond government control. The problem with that claim is that some AR or VR may not be a matter of First Amendment law at all: If certain elements of AR gameplay do not count as First Amendment expression, we cannot successfully raise a First Amendment objection to government restriction of it. It may be, perhaps, that we could raise some other constitutional objection. Even when the First Amendment does not shield some activity, for example, courts generally hold that, under the Equal Protection Clause of example, that it not violate a person’s due process rights or impose federal power over activity that Constitution does not empower the federal government to regulate). Id. 61   Universal City Studios v. Corley, 273 F.3d 429, 435 (2nd Cir. 2001). 62   Id. 63   Id. at 451. 64   Id. at 453.

BARFIELD PRINT.indd 323

13/11/2018 11:52

324  Research handbook on the law of virtual and augmented reality the Fourteenth Amendment, government still needs a “rational basis” to be allowed to regulate it consistent with the Constitution: Government needs to show that its regulation has a legitimate government interest and that the means legislators choose (in designing the law) to achieve that interest have a rational relationship to it.65 This is a very hard test for the government to flunk: To the extent that an AR game creates safety risks (for example, where AR games might lead smartphone users to wander into traffic while staring at their phone), it seems clear that government would have a legitimate safety interest in concerning itself with the activity in question. Still, it is at least conceivable that if government wishes to regulate purely virtual graffiti simply because it did not want an AR user to experience visual stimuli the government viewed as unattractive, this might flunk the “legitimate government interest” requirement even in the unlikely event that the Court found the AR user lacked a First Amendment right to see whatever visual scene she wished to see superimposed over public space.

4. AR, VR, AND CIRCUMSTANCES WHERE SPEECH RECEIVES LESS PROTECTION Even if and when a particular use of AR or VR technology counts as First Amendment speech, government may still restrict it. It will generally not be permitted to target certain ideas or speech content in AR or VR settings simply because it believes that content to be objectionable. Under the First Amendment, even deeply unpopular ideas—in fact, even ideas that can be described as “hate speech”—are generally protected against government censorship. Government, however, has more leeway to restrict speech under certain circumstances. First, it may do so when it is not regulating the speech’s content, but rather some “noncontent” feature of the speech—such as its decibel level, the location in which it is occurring, or, if the speech is expressed through nonverbal conduct such as burning a flag or draft card, the physical impact that the burning has on property.66 Second, government may regulate certain categories of content that have traditionally be considered unprotected, or less protected, than speech content usually is. For example, if speech incites others to commit an unlawful act, it is not protected by the First Amendment.67 Government can likewise restrict “fighting words” or threats of violence, obscenity, child pornography, and other speech integral to criminal activity; defamatory or fraudulent speech; and a number of other categories of content free from First Amendment limits.68 Where speech is “commercial,” that is, where it advertises a product or otherwise has a commercial character under the Supreme Court’s somewhat unclear definition of what counts as “commercial speech,” it is still protected by the First Amendment but does 65   See, e.g., Dunagin v. City of Oxford, Miss., 718 F.2d 738, 753 (5th Cir. 1983) (holding that under “minimal scrutiny . . . the classification challenged need only be rationally related to a legitimate state interest”). 66   See, e.g., March v. Mills, 867 F.3d 46 (1st Cir. 2017). 67   See United States v. Stevens, 559 U.S. 460, 470 (2010) (finding government may regulate these categories of low value speech). 68   See id.

BARFIELD PRINT.indd 324

13/11/2018 11:52

Freedom of expression, and the personalization of public space  325 not receive the same level of protection as political discussion, art, or other expression in which people engage to communicate ideas or give form to their thoughts.69 Third and finally, government has more leeway to regulate all manner of content where the role it is playing is different from the role it normally plays when it enacts, or enforces, laws or rules restricting individuals’ behavior. For example, government plays the role of educator when it operates public schools, and can in that capacity bar students from engaging in speech that legislators cannot ban in public.70 It plays the role of employer in public employment, and, for example, can fire a government worker for taking action that is inconsistent with the requirements of his job.71 It can likewise limit what people say on some of the property government manages—as it does when it bars solicitation of money in government-owned airports—or what people do with government money (that is provided only to be use for speech the government wishes to support).72 There is insufficient space to explore all of the ways in which these aspects of First Amendment doctrine could allow government to place limits on AR content superimposed over public space, or VR content set in a virtual recreation of public space. Moreover, in many cases, there is no reason to think the First Amendment grounds rules for regulation of certain speech content in AR or VR would be any different from those which apply to the same content in other settings. For example, imagine that a restaurant business has a plausible case that a person has defamed it by falsely printing, in an independent newspaper, that the restaurant’s kitchen is infested with rats. A person has no First Amendment right to make such a defamatory statement. The law could subject such defamation to civil liability whether it occurs in a newspaper, on a website, or in a box that is superimposed over the restaurant in an AR environment. Similarly, a sexual image that counts as obscene, and thus is unprotected by the First Amendment, in a film shown on a flat screen would likewise count as obscene if it appeared in a more immersive virtual reality equivalent of the two-dimensional movie. There is no reason to think AR or VR technology would make a difference to the First Amendment analysis in these cases. However, there are a few questions about First Amendment doctrine where certain aspects of AR or VR may well make a difference. Here, I am most interested in types of AR or VR which seek not only to replicate public space, but also to add customized features to it—by superimposing—over people’s view of the physical environment—­ historical or commercially useful information, graffiti images, or other artwork or images or words central to certain kinds of gameplay. 4.1  Content-Neutral Regulation of Speech One reason why government may be able to regulate such AR or VR modifications of how we see public space is that it can regulate them in a “content-neutral” manner. Instead 69   See Cent. Hudson Gas & Elec. Corp. v. Pub. Serv. Comm’n of N.Y., 447 U.S. 557, 563–4 (1980); Marc Jonathan Blitz, The Pandora’s Box of 21st Century Commercial Speech Doctrine: Sorrell, R.A.V., and Purpose-Constrained Scrutiny, 19 Nexus: Chapman JL & Pol’y 19, 45–8 (2013–14). 70   See Tinker v. Des Moines Indep. Cmty. Sch. Dist., 393 U.S. 503, 513 (1969). 71   See Connick v. Myers, 461 U.S. 138 (1983). 72   See Int’l Soc’y for Krishna Consciousness v. Lee, 505 U.S. 672, 698 (1992).

BARFIELD PRINT.indd 325

13/11/2018 11:52

326  Research handbook on the law of virtual and augmented reality of targeting only AR or VR images that communicate certain ideas, it can target all AR or VR images, or a subset of them, based on the noncommunicative impact they have on those who use the technology. For example, a state legislature may believe that any use of AR by a driver will cause distractions and increase the risks of accident. Drivers, the legislature might reason, should be focused entirely on how to navigate the actual physical environment—they should focus on the road, and on the behavior of other drivers or pedestrians nearby, not on signs or floating messages in an AR world. Pedestrians, legislators might argue, should similarly be barred from using AR or VR in a way that is distracting. Multiple newspaper stories in 2015 carried reports of distracted Pokémon Go players falling into pot holes, or—as mentioned earlier—in one remarkable case, off a cliff.73 An AR restriction is also arguably content-neutral when it prevents AR technology from being used in a certain location, or under circumstances that predictably attract large crowds to that location. So long as government avoids discriminating for or against certain kinds of speech content, it may require individuals to obtain a permit before using public space in a way that would interfere with others’ use of that space, or might present threats to the condition of that space. For example, the National Parks Service requires groups to seek a permit before holding any “demonstration” that will involve more than 25 participants on the National Mall or any other park space under its jurisdiction. A “demonstration,” under its guidelines, is any “speechmaking, picketing, vigils, marching or religious services etc., and all other like forms of conduct that are reasonably likely to draw a crowd or onlookers.”74 Certain parts of the Milwaukee augmented reality permit requirement being challenged by Candy Lab are arguably content-neutral in this sense. Indeed, the United States District Court for the Eastern District of Wisconsin found them to be so. It found that, while the impetus for the ordinance was Pokémon Go, the Milwaukee ordinance was not targeting this or any other game based on the particular content of the game. It did not apply to any game on the basis of the “topic discussed or the idea or message expressed.” It rather targeted AR or “virtual games” primarily based on their use of such immersive technology and on their being “location-based,” and did so “regardless of their content, be it poker, zombie-killing, or Pokémon-catching.”75 Under existing First Amendment doctrine, courts may look closely at a content-neutral statute to see if it is a pretext for targeting certain ideas. But the court noted that the ordinance’s aim was to “prevent litter, vandalism, traffic, and other problems attendant upon AR games like Pokémon Go” and there was no evidence it was “premised on a dislike for Japanese art or culture.”76 However, simply because a measure targeting AR (or VR, for that matter) is contentneutral does not mean it is automatically constitutional. In the first place, courts do not simply take government’s word that the restriction is content-neutral. They scrutinize   See text accompanying note 24.   National Park Service, Instructions for Public Gathering Permit, www.nps.gov/nama/­plany​ ourvisit/upload/public-gathering-application-2014-final-with-doi-seal-fillable-and-dark-­linesfinalse​ -2.pdf. 75   Candy Lab, Inc. v. Milwaukee Cty., No. 17-CV-569-JPS, 2017 WL 3088380, at *8 (E.D. Wis. July 20, 2017). 76   Id. 73 74

BARFIELD PRINT.indd 326

13/11/2018 11:52

Freedom of expression, and the personalization of public space  327 the statute to assure the government’s interest is not in suppressing certain ideas. After that is established, a content-neutral speech rule has a greater chance of being found constitutional by a court. It is not, like a content-based speech restriction, treated as unconstitutional—except in the very rare and unlikely circumstance that government can show its speech restriction is absolutely “necessary” to fulfill a “compelling” government interest, and is the least speech-restrictive means to do so.77 That “strict scrutiny” standard is almost impossible for the government to meet. Content-neutral restrictions on speech must meet only the lower standard of “intermediate scrutiny.” Government only requires a “substantial” interest, not a “compelling” one,78 and preventing litter, controlling traffic, or protecting the aesthetic features of a park might well qualify as such an interest. Moreover, in pursuing that interest, it is not required to choose the one measure that restricts expression the least;79 it only has to avoid imposing “substantially more” restriction on speech than is necessary. These features of intermediate scrutiny apply to content-neutral regulation of speech whether the avoidance of regulating content is because it focuses on the physical conduct side of expressive conduct (like the burning of a draft card, rather than the protest message) or on the “time, place, and manner” of speech, rather than what it says.80 In this case, the Milwaukee ordinance regulated the time, place, and manner of speech. Its principal limit was on the location of expression entailed in AR games, and perhaps also its manner and time to the extent that it placed conditions on how and when such games were played in Milwaukee parks. It is worth pausing here to note that, given these features of First Amendment doctrine regarding content-neutral speech regulation, AR activity seems far more likely to be a fair target for content-neutral speech restriction than VR. This is because, where speech regulation is content-neutral, it is typically because its regulatory focus is not on the ideas in the speech but rather on the noncommunicative effects they have on the surrounding environment: The pollution or property damage they cause, the traffic tie-ups they generate, or the high-decibel level of the noise they generate. AR activities can have these effects when they occur in real space: Large crowds of Pokémon Go players might trample grass in a park, knock over fencing, or create noise that awakens those living near a game site. When someone plays a VR game at home, by contrast, the virtual-world activity need not leave any trace in the physical world outside. Gameplay or expressive activity in a shared VR environment might, to be sure, damage virtual property or pollute virtual land in that VR environment (many Second Life users have experienced vandalism of their virtual property by others in the Second Life). But such damaging VR activity is typically controlled (or punished) by the gamemaster—who works at the relevant gaming company—whereas AR damage to physical streets and parks is the kind of harm addressed by parks departments or other officials. AR activity, in other words, remains anchored in a shared physical space, and it is that continuing connection that leaves it a likely target for regulation of when, where, and how speakers and listeners exercise First 77   See United States v. Windsor, 133 S. Ct. 2675, 2717 (2013) (holding that laws subject to strict scrutiny must be “‘narrowly tailored’ to achieve a ‘compelling’ government interest” (quoting Parents Involved in Cmty. Schs. v. Seattle Sch. Dist. No. 1, 551 U.S. 701, 720 (2007))). 78   Ward v. Rock Against Racism, 491 U.S. 781, 791 (1989). 79   Id. 80   Id.

BARFIELD PRINT.indd 327

13/11/2018 11:52

328  Research handbook on the law of virtual and augmented reality Amendment rights in physical space. It is this persistent connection to physical space that allows government officials to plausibly claim they have a substantial government interest in regulating location-based AR, and to try to show that their restriction is proportional to that interest. Time, place, and manner restrictions are also subject to an additional limit: They must leave “ample alternative channels” open for those engaging in the expression in question, even if they close one or more channels in a content-neutral way. If Milwaukee typically approved permit applications, it would be providing a channel for the expression entailed in AR games. When rejecting permit applications, Milwaukee and other municipalities barring use of an AR game in parks (after permit consideration) would need to convince courts either that there are other opportunities for game designers to introduce, or individuals to play, the same location-based AR or virtual game elsewhere—or they would have to make what seems to be a less plausible case that the game designers or game players could effectively express the same speech content without the augmented or virtual reality aspects that subject the game to rejection. Moreover, where the form taken by a time, place, and manner restriction is a permit requirement, then there is still another First Amendment test that must be met: the decision about whether to grant the application must not be left to an official’s “boundless discretion.”81 It must instead “contain adequate standards to guide the official’s decision and render it subject to effective judicial review.”82 Without such standards, an administrator could use a content-neutral permit requirement, meant simply to ensure against harm to park conditions, and instead make it a basis for censorship of content the administrator happens to oppose. It was this First Amendment test that led the federal district court to conclude that Candy Lab would likely succeed in its challenge to Milwaukee’s permit requirement for AR games: The Milwaukee ordinance, noted the court, states that “Milwaukee County Parks in its sole discretion may grant, deny, revoke, or suspend any permit, at any time and for any reason.” This, said the court, leaves an official with unconstrained and unguided discretion to censor expression in AR or VR applications.83 One might also argue that, even if a local permit decision of this sort is constrained by “adequate standards,” imposing such a limit on all “[v]irtual and location-based augmented reality games” restricts far more speech than is necessary to secure a locality’s interest in protecting the condition of, and public use of, the parks. Not all location-based AR games, after all, will generate the following of Pokémon Go. Even in certain popular, widely used AR games, sites that generate AR images may be so widely distributed across a city that it is unlikely that users will congregate in one particular park at one particular time. Courts thus might ask, as they often do in applying intermediate scrutiny, whether there is a substantially less speech-restrictive alternative that would achieve the same goals achieved by an ordinance such as this. It is also worth noting that, under the Supreme Court’s secondary effects doctrine, officials may sometimes be able to constitutionally place limits on location-based—or

  See Candy Lab, Inc. v. Milwaukee County, 2017 WL 3088380, 9 (E.D Wis. Jul. 20, 2017).   Id at 9–10. 83   Id at 3, 10–11. 81 82

BARFIELD PRINT.indd 328

13/11/2018 11:53

Freedom of expression, and the personalization of public space  329 other—AR or VR games, even where the behavior targeted by the law is a predicted effect of these games rather than an inherent part of them. In City of Renton v. Playtime Theatres, Inc., the Supreme Court held that a city could impose a zoning ordinance preventing “adult motion picture theaters from locating within 1,000 feet of any residential zone, single- or multiple-family dwelling, church, park, or school.”84 The city’s stated justification for doing so was not that it objected to the films screened in such theaters, but rather that that kind of adult business had “secondary effects” on life in the immediate vicinity: It was a magnet for crime, hurt retail trade, depressed property values, and “damaged the quality of urban life.”85 The Court held that because the zoning ordinance targeted these predictable “secondary effects” of the adult film business, and not the films themselves, it was content-neutral, and also survived the intermediate scrutiny that applies to content-neutral regulation.86 It is conceivable that ordinances requiring permits for (or otherwise limiting) use of AR games could likewise be based on justifications about the predicted effects of these AR games. For example, even if an AR game does not require that crowds congregate in a certain area, or that they engage in rowdy behavior when they do so, cities might argue that where an AR game (like Pokémon Go) has tended to generate unruly crowd behavior, then the city should have a right to assume that such behavior will often occur with the playing of the game. It is conceivable that a city might—based on this “secondary effects” doctrine—target not only AR games generally but rather certain types of AR game content, and argue that it is doing so not in order to target that content (or because it has any objection to it), but rather because that AR content (unlike other kinds of content) tends to produce the rowdy behavior it is concerned with. An AR game akin to Assassin, for example, may—a city could argue—attract certain kinds of people who engage in behavior that may be incompatible with a park open to all ages, whereas a more family-oriented AR game (such as a kind of augmented reality Easter egg hunt) might not be likely to generate such behavior. These secondary effects-based arguments, however, seem fraught with difficulties. First, the secondary effect doctrine has been subject to strong criticism from scholars who believe it gives government an easy way to target disfavored content (such as explicit adult material) under the pretext of a content-neutral law that targets its nonspeech effects (such as lowering of property values). When the effects (such as lower property values) arise because the speech content strikes city residents as objectionable, then it is hard to argue that those effects—or a law responding to them—are content-neutral. As David L. Hudson, Jr puts this point: “the secondary-effects doctrine allows government officials to claim that patently content-discriminatory regulations—often those that restrict only businesses featuring adult-oriented expression”—should be “treated as content-neutral,” and the doctrine “is easily manipulated, as government officials often advance various secondary-effect justifications.”87 Second, the Supreme Court recently held, in Reed v. Gilbert, that where a speech restriction, on its face, treats certain speech content differently than other speech   City of Renton v. Playtime Theatres, Inc., 475 U.S. 41 (1986).   Id. 86   Id. 87   David L. Hudson, The Secondary Effects Doctrine: Stripping Away First Amendment Freedoms, 23 Stan. L. & Pol’y Rev. 19, 19–20 (2012). 84 85

BARFIELD PRINT.indd 329

13/11/2018 11:53

330  Research handbook on the law of virtual and augmented reality c­ ontent, courts should treat such restrictions as content-based regardless of government’s purpose in drawing a distinction. It is hard to square this holding in Reed with a doctrine that suggests that a law can target particular content (such as the screening of sexual content in “adult theaters”) and still be content-neutral.88 To be sure, officials may still draw on a variant of secondary effects doctrine where they impose a restriction on AR or VR games that is neutral on its face—for example, because it targets certain AR or VR games on the basis of number of registered users, and not on the basis of its theme, characters, or storyline. They may argue that even if an AR game with a large number of users is not inherently problematic, it may become so if the technology tends to generate crowd behavior that is problematic. Even here, however, courts are warranted in looking closely at the city’s predictions of problematic behavior, since—where such predictions are unjustified—the city may lack the kind of “substantial government interest” required to restrict expression through AR or VR. Also, if crowd behavior which the city deems problematic is itself expressive—if, for example, it involves the boisterous celebration people engage in as they battle augmented reality zombies or catch Pokémon—then the city’s targeting of it is not targeting only the “secondary effect” of speech, but targeting First Amendment speech. Finally, even where courts apply secondary effects doctrine to use of location-based AR games, they will still ask whether there are alternative regulations that would restrict substantially less speech but fulfill the city’s interest in controlling secondary effects. Rather than restricting AR games directly, for example, cities might sometimes restrict some of the behavior they believe arises from such games, but then continue to let players play the game and express themselves with such behavior minus its proscribable aspects. Such First Amendment questions about content-neutral regulation are not unique to regulation of AR or VR games. They might arise when cities regulate other behaviors in parks—from ultimate Frisbee, to oldfashioned games of Assassin (with plastic guns rather than smartphones), to playing radios or electronic guitars. These examples, of course, apply to AR technology which projects images over a smartphone screen, Google Glass or other AR-enabled glasses, or, perhaps, a futuristic car window. They do not concern VR technology where virtual imagery, like a sign, is being superimposed not over a window onto the surrounding physical environment, but rather over a simulation of that environment which is itself virtual. 4.2  “Low Value” Speech Thus far, this chapter has explained how AR and VR can weaken the government’s case for regulating expression that takes place in those environments, by transforming activities that once had a physical impact on our affairs into private activities that take place to a greater extent in a realm of pure perception and imagination. 88   See Reed v. Town of Gilbert, 135 S. Ct. 2218 (2015). See also Leslie Gieslow Jacobs, Making Sense of Secondary Effects Doctrine after Reed v. Gilbert, 57 Santa Clara L. Rev. 385 (2017). Jacobs notes that Reed appears to be in tension with the reasoning that underlies secondary effects doctrine but considers the possibility that the Court, in Reed, did not mean to reject secondary effects doctrine and that “content-correlated” ordinances do not count as facially content-based simply because of their content correlation.

BARFIELD PRINT.indd 330

13/11/2018 11:53

Freedom of expression, and the personalization of public space  331 However, there are also areas of First Amendment doctrine in which AR and VR could conceivably have the opposite effect: They might strengthen the case for government speech restriction rather than undermine it. The case for greater speech restriction power in AR and VR may be particularly true for what courts sometimes call “low value” or less or unprotected speech categories—speech such as obscenity, true threats, defamation, intentional infliction of emotional distress, and other speech torts (such as invasion of privacy). Here, draining the interaction of its physical qualities does not typically undercut the government’s case for censorship—and that is because the harm the government addressing is not in those physical qualities, but in the nature of the words themselves. Defamation, for example, causes damage to a person’s reputation even when it is voiced in a completely virtual environment. In some cases, AR or VR might even exacerbate the speech harm that justifies treating certain speech content as dangerous and deserving of greater government restriction. Consider, for example, one of the reasons the court has often given for treating commercial speech as meriting less protection than artistic or political speech. Commercial speech receives some First Amendment protection: Regulation of commercial speech must survive intermediate scrutiny. Regulation of nonmisleading commercial speech is generally permitted only when the regulation: (1) is supported by substantial government interest, (2) directly advances that interest, and (3) avoids restricting substantially more speech than necessary to achieve that interest.89 This is the so-called “Central Hudson” test for regulation of commercial speech. The court has been inconsistent in explaining why commercial speech merits less protection than does other speech, but the explanation it has sometimes given is that commercial speech often carries with it a “risk of fraud.”90 Indeed, where government can show that this risk is not merely a risk but already a ­reality—where it can show the commercial speech is misleading—it can regulate it without even having to satisfy intermediate scrutiny. How might VR and AR make a difference here? Commercial speech in these settings may well take on a different more immersive form: Advertising with such technology, for example, might allow people to actually hold in their hands, or otherwise have a realseeming experience of, a product they are considering for a purchase. VR users might test drive a virtual Ford or Toyota car before they buy a physical car to drive on real-world roads. They might sample a virtual hotel on the Greek Islands before they reserve a space in it, and travel there. To the extent that VR simulations are more likely than verbal descriptions to be taken as accurate, they may increase the risk of fraud above that which is likely to arise in verbal presentation of similar commercial offerings. Virtual reality may likewise intensify some of the harms that arise in certain speech torts. Consider defamation again. In some cases of defamation, VR or AR is unlikely to make any difference at all. A defamatory statement, put in the form of words, won’t be any less harmful in VR or AR, but nor will it be more harmful. The harm lies in the false claim within the defamation (and how many people hear it), not in whether the medium is print or digital, an immersive setting or a two-dimensional screen. But consider a case where, instead of taking the form of words, defamation takes the form of a scene in a VR

  Cent. Hudson Gas & Electric Corp. v. Pub. Serv. Comm’n of N.Y., 447 U.S. 557, 563 (1980).   Sorrell v. IMS Health Inc., 131 S. Ct. 2653, 2672 (2011).

89 90

BARFIELD PRINT.indd 331

13/11/2018 11:53

332  Research handbook on the law of virtual and augmented reality environment that shows a particular person engaging in behavior that she never actually engaged in. It may be that, in a world where artificial VR simulations are a familiar part of day-to-day life, Americans will increasingly learn that—at least in VR—they cannot trust their own eyes; that even when they see an immersive image of well-known politician engaged in scandalous behavior, they will realize this could be a very real-seeming fiction that is no more real than a vivid dream. However, it is possible that a VR scene could accompany, and powerfully reinforce, a lie about a particular figure. If, thanks to VR, the effects of defamation become harder to counter, the First Amendment may have to allow more room for courts or other officials to ferret out untruths in such speech, and impose costs (or demands for retraction) on those doing the defaming. In both of the above cases, AR and VR might challenge our current doctrine’s robust protection for free speech—and create a sense that greater exceptions to it are required—by making it easier for perpetrators of fraud or other falsity to simulate real events and give them the feel of real events. VR and AR could also increase the need to provide government greater power to fight another feature of some low value speech. Apart from protecting us against inaccurate factual claims (in commercial speech or defamation), government is permitted under First Amendment law to protect us against certain kinds of attacks on our emotional wellbeing or dignity. It can enact laws letting us sue others for intentional infliction of emotional distress or invasion of privacy. Under the Restatement (Second) of Torts, intentional infliction of emotional distress (or IIED) occurs when there is “outrageous conduct by the defendant, with the intention of causing, or reckless disregard of the probability of causing, emotional distress, actual suffering of severe or extreme emotional distress, and actual and proximate causation of the emotional distress by the defendant’s outrageous conduct.”91 Verbal statements can certainly count as IIED—as they may, for example, if a person falsely tells another that a close relative or friend has died. But IIED would arguably occur in VR or AR if, for example, one creates a vivid virtual scene where a particular individual targeted by the VR or AR content creator, or a relative of that person, undergoes torture or other virtual harm. It is conceivable that such virtual torture would count as “outrageous conduct” of a kind that can and does cause severe emotional distress. As one article notes, VR encounters with simulated danger or tragedy can generate experiences far more intense than those in frightening or disturbing movies: “[a]lthough horror games are intended to scare you, when experienced in virtual reality they can provoke intense fear—and potentially, even trauma.”92 As Jaclyn Seelagy writes, a vivid attack, one that “feels like real touch,” could have a powerful, and damaging, impact: “[e]ven without physical pain or injury, [virtually simulated] violence in this context could do serious psychological harm.”93 Indeed, Justice Alito underscored this capacity of evolving virtual technologies in his concurring opinion in Brown v. Entertainment Merchants Association: He worried that video games of the “near future” would “allow troubled teens to experience in an extraordinarily personal and vivid way what it would be like to carry out unspeakable acts of violence.”94 He also

  Restatement (Second) of Torts, Section 46.   Laura Hudson, “It’s Not Real. It’s Not Real. It’s Not Real,” Slate, May 2, 2016. 93  Seelagy, supra note 9, at 424. 94   Brown v. Entertainment Merchants Association, 564 U.S. 786, 819 (2011) (Alito J concurring). 91 92

BARFIELD PRINT.indd 332

13/11/2018 11:53

Freedom of expression, and the personalization of public space  333 noted that “it will not be long before video-game images will be seen in three dimensions” and “provide sensory feedback” that will allow a player to experience physical sensations supposedly felt by a character.”95 I have noted, in a previous article on First Amendment implications of virtual reality, that it is conceivable that the intuitions of judges and other legal actors may well differ when, instead of watching torture scenes in a movie such as Hostel, Reservoir Dogs, or A Clockwork Orange, VR allows a person to vividly experience “what it is like to be in the midst of such cruelty, and perhaps be its perpetrator.” 96 To be sure, where people immerse themselves in such a harrowing virtual experience rather than imposing it upon others, government justifications for restricting such use of VR would likely be paternalistic rather than based on protecting one person from another’s intentional imposition of a distressing experience. It is possible that future First Amendment doctrine will give government more room to limit such violent virtual experience than it has thus far—at least where the users of VR are minors. However, for courts to reach this conclusion, they would need to shift away from First Amendment doctrine that has, in recent years, been characterized by great reluctance to recognize new categories of “low value” speech where content-based speech limits are permissible,97 as well as a deep skepticism toward government regulations rooted in paternalism (apart from those which protect minors against obscene content). 4.3  Public Forum Doctrine The above discussion questioned how First Amendment law might deal with situations where AR or VR technology supplement public space that continues to exist outside of it—for example, in location-based AR games that superimpose images onto, and allow certain activities in, the physical world, or in VR or AR environments that provide historical or other descriptive information about a building that isn’t otherwise visible. But virtual reality, in particular, can do more than supplement physical space. In some ways, it can replace it. How, one might ask, would the First Amendment apply to government action that shapes such virtual environments for speaking? Public forum doctrine is the part of First Amendment law that generally answers questions about how government can set rules for the use of the shared public space it controls: a city’s streets and parks; meeting rooms in convention centers run by the government; airports and bus terminals; military bases and prisons. The key principle in this law is that most of this space is classified as being in one of two categories: (1) a public forum, such as a street or a park, where the space is set aside for robust debate and other First Amendment speech and where government must thus avoid placing limits on what people can talk about or advocate;98 (2) a nonpublic forum, where the principal use of the space is for a nonspeech public purpose (such as allowing for safe and efficient air travel or making it possible for the U.S. armed forces to engage in training). In the view of the courts, it   Id.   Marc Jonathan Blitz, The Freedom of 3D Thought: The First Amendment in Virtual Reality, 30 Cardozo L. Rev. 1141, 1146–7 (2008). 97   United States v. Stevens, 559 U.S. 460, 460 (2010); Brown, 564 U.S. at 791. 98   Perry Educ. Ass’n v. Perry Local Educators’ Ass’n, 460 U.S. 37 (1983). 95 96

BARFIELD PRINT.indd 333

13/11/2018 11:53

334  Research handbook on the law of virtual and augmented reality is clear that both such spaces are necessary in a modern democratic society. On the one hand, in order for free speech to occur and reach a public audience, people need spaces they can speak from, and in which they can speak to other citizens—including strangers. This would not be possible if people could speak freely only on their own property, or on the private property of a friend, but were subject to government censorship anywhere in the public space that the government controls. So courts have held that people’s freedom to advance even deeply unpopular or heretical ideas extends outside of their homes.99 However, there are at least two respects in which the absorption of public space by virtual reality may make a First Amendment difference. First, for public forums to provide the platform they are supposed to provide for public discourse, people have to be able to express themselves not just to interested audiences, but to other citizens in that space. Speakers have traditionally proselytized about religion and politics in parks and streets not only to connect with audiences who are already seeking them out, but also to attract the attention of pedestrians who might otherwise ignore them. Indeed, even nonpublic forums are supposed to allow room for such broadcasting of views to the world, to the extent that such room is compatible with the primary purposes of the public space (like those in an airport, a military base, or some other function-specific government-operated space or structure). But VR and AR pose a possible problem for this aspect of public forums. We have seen earlier that there are characteristics of these technologies that allow people greater freedom to speak without causing disturbance for others (thus giving the government less reason to intervene). But these same characteristics may also allow people to modify public space in a way that shuts out expression of the kind that is supposed to reach them. If, on a virtual reality street that we share with others, people can select a setting that automatically screens out any verbal statements that would distract them from their single-minded mission (of getting to a particular store and buying a particular product), then they will not be open to hearing the messages people have traditionally heard in streets and parks. They will always be able to “mute” the kind of speech that the First Amendment has traditionally allowed people to confront them with. This is a problem that has already been discussed by legal scholars and journalists analyzing the internet. In Republic.com, Cass Sunstein argues that when individuals surf the World Wide Web, they often limit their exploration to sources of information that already align with their own views: Conservatives will seek out conservative news sources and read conservative blogs; liberals will seek out liberal news sources and read liberal blogs.100 This kind of “confirmation bias” may extend even into public spaces: We may be more likely to listen to a speaker, or read a pamphlet we obtain, while walking in the street if the words we hear in the speech or see in the pamphlet title indicate it will support

 99   There are also settings where government is not generally subject to public forum doctrine’s constraints because the setting is exclusively for the government’s own speech, which it is free to shape as it wishes: see, for example, Pleasant Grove City v. Summum, 555 U.S. 460 (2009); or because it has to make judgments about the accuracy or quality of certain views, as public libraries do, for example, if they order only science books with accurate and up to date science, or make available on library Internet-connected computers only Web pages with such accurate information. See United States v. Am. Library Ass’n, Inc., 539 U.S. 194 (2003) (plurality opinion). 100   Cass Sunstein, Republic.com 2.0, 1–23 (2007).

BARFIELD PRINT.indd 334

13/11/2018 11:53

Freedom of expression, and the personalization of public space  335 our preexisting views, and more likely to ignore a speaker who seems to be disagreeing with our views. But at least, in the streets and other physical public forums, speakers with unfamiliar positions are not entirely blocked off. These considerations raise the question of whether, if public forums such as streets and parks which have physical form come to be experienced as virtual reality environments instead, government should be able to mandate, consistent with the Constitution, that the design of such VR environments retain the features of public space that allow individuals to broadcast messages to other citizens (even when these other citizens might prefer to block them out). It seems at odds with First Amendment values, if not with First Amendment law, for government to require individuals who are moving through public space to listen to words they wish to ignore. The First Amendment generally leaves it up to us what we choose to say, and also leaves it up to us what other private citizens we wish to listen to (except where we have some duty to attend to certain features of the environment, for example, when we are driving a car). But the First Amendment is also designed to make possible robust debate and discussion between citizens.101 Thus, one might argue that this value should leave room for government to mandate that other citizens’ messages at least be audible or visible to us as we move through public space—that virtual reality environments replacing public streets not be so subject to customization that we can shut other voices out entirely. There is also a second First Amendment question and challenge raised by the transformation of public space into a virtual equivalent: Most of the virtual worlds and social networking sites that have increasingly absorbed individuals’ time, and served as platforms for individuals to debate and discuss their political and other views, are not owned or operated by government. They are run by private companies, such as Facebook, Twitter, Instagram, Yelp, or Amazon. These electronic environments are, in a sense, substitutes for public forums (and to some extent nonpublic forums), but they are not legally equivalent to these forums. In a public forum, government must refrain from restricting content. Private companies, by contrast, generally have power to formulate any rules they like regarding who can say what. The First Amendment does not constrain them at all. Facebook and Twitter, for example, have terms of service, and users of these social media sites are constrained by them. Consequently, the move from physical to virtual public space may involve privatization of such space in more than one sense: Not only can individuals personalize and customize such space to a greater degree than physical space, but they will also be bound by rules for using that virtual space that are set by a private company and not by a public official. How then can First Amendment law protect speech in privatized versions of public space where the First Amendment does not apply? Imagine, for example, that individuals not only go to virtual stores, but increasingly congregate and spend spare time interacting in VR streets or parks near those stores (by taking the form of “avatars” that visually represent individuals in the VR world). Operators of that VR platform might discriminate against certain views; they might allow signs promoting their own business interests but not signs they view as adverse to those interests. Or they could even ban verbal comments

101   See N.Y. Times Co. v. Sullivan, 376 U.S. 254, 270 (1964); Red Lion Broadcasting Co. v. F.C.C., 395 U.S. 367, 391–2 (1969).

BARFIELD PRINT.indd 335

13/11/2018 11:53

336  Research handbook on the law of virtual and augmented reality in the VR world that contain positions they find objectionable. Whereas oldfashioned streets and parks are public forums where First Amendment protections are at their strongest, these virtual streets and parks would be First Amendment-free zones. One solution is to find that, when private companies take over functions previously performed by the state—such as managing shared public space—they should be subject to the same First Amendment constraints that bind the state. This is what the Supreme Court held in the 1946 case of Marsh v. Alabama.102 In that case, a Jehovah’s Witness wished to distribute religious literature on the streets of a privately owned town in Alabama owned and run by the Gulf Shipping Corporation for the benefit of its employees. The corporation argued that, since the town was privately owned, it could control speech on its streets free from constitutional limits. The Court disagreed: Because the company was performing “public functions” of the kind that are the hallmark of state power—such as controlling and assuring safety on streets—the company was bound by state responsibilities that come with such state powers, namely the power of respecting First Amendment and other constitutional limits on the exercise of such power.103 Peter Jenkins has argued that virtual worlds are like company towns: They control electronic streets which people increasingly use as substitutes for physical streets. However, there are a number of difficulties with this position.104 In the first place, the Court has steadfastly refused to extend Marsh’s holdings to other situations. It has made clear, for example, that even where people congregate in a privately owned mall or shopping center, much in the way they might once have congregated in the public streets of a town’s central commercial district, their speech is unprotected against speech restrictions by the owner of that mall or shopping center.105 In the second place, as Greg Lastowka and Dan Hunter point out, placing constitutional constraints on virtual-world operators—and, in the future, on designers of immersive virtual reality—would likely constrain these virtual world operators’ own First Amendment freedom to design interesting virtual environments.106 I have previously endorsed such arguments against treating virtual worlds as company towns, but this assumed that people could find alternative options for expressing themselves, that when their views or favored topics for discussion are shut out of one virtual environment, they can find another.107 If this is not true of future VR substitutes for public space, then the company town analogy may be more compelling. In any event, a more promising method for extending First Amendment-style protections to VR replication of public space is to do so by federal or state statutes, rather than through First Amendment law itself. This is along the lines that Marvin Ammori recommends for the internet: Although the First Amendment itself does not require that the internet be designed in a certain way, he argues, it should at least allow room

  Marsh v. Alabama, 326 U.S. 501 (1945).   Id. 104   Peter S. Jenkins, The Virtual World as a Company Town: Freedom of Speech in Massively Multiple Online Role Playing Games, 8 J. Internet L. 1, 8 (2004). 105   Hudgens v. NLRB, 424 U.S. 507 (1976). 106   F. Gregory Lastowka and Dan Hunter, The Laws of the Virtual Worlds, 92 Cal. L. Rev. 1, 72 (2004). 107   Marc Jonathan Blitz, A First Amendment for Second Life: What Virtual Worlds Mean for the Law of Video Games, 11 Vand. J. Ent. & Tech. L. 779 (2009). 102 103

BARFIELD PRINT.indd 336

13/11/2018 11:53

Freedom of expression, and the personalization of public space  337 for government to assure that a speech-supporting architecture flourishes in internet communications, for example, by enacting net neutrality rules. The First Amendment might likewise allow government to enact rules requiring that VR companies who recreate public space in certain ways will refrain from discriminating against expression of certain views or discussion of certain topics.108 While the Supreme Court found that the First Amendment does not itself apply to the operations of a mall,109 it has also held that it does not prevent California or other states from adopting of state constitutional law provisions, or enacting of legislation, that requires mall owners to respect speech rights.110 It is unclear whether this analogy would apply to VR environments: Those who design and operate VR environments could argue that, like video game designers, their decisions about how to build their VR world (and what to allow in it) are expressive in a way that operating a mall is not. But it is possible that courts would reject that argument, holding that those who design virtual streets cannot be any more free from government regulation of such environment shaping than are the architects and builders who create physical streets and parks. It is also possible that courts would apply the First Amendment differently to different kinds of virtual environments. Finally, it is worth noting one other implication of a scenario in which VR or AR experiences of public space replace, rather than supplement, our current, less technologically filtered perception of it. I wrote in the introduction to this chapter that, in a personalized and customized version of public space, we might have First Amendment protection where we overlay the images of our choice over a building, monument, or street sign—even though we would be defacing property and violating the law if we did so with spraypaint or marker in the physical world. Where physical graffiti would be a crime, AR graffiti that we write only for those AR users who wish to see it is likely speech protected by the First Amendment.111 But matters may be different if a particular AR overlay was not an option that could be easily be avoided by those who wanted to leave it out of their world, but rather a default image that was impossible (or almost impossible) to avoid. For example, imagine a world where everyone uses one authoritative, and government-sanctioned, AR app to navigate public streets. In that case, a view of particular building overlaid with AR graffiti may be the only view of the building most people have. The graffiti would thus be part of citizens’ shared experience of public space whether the owner wanted it there or not. In such a case, it would cause damage similar to that currently caused by graffiti spraypainted onto a physical building wall.112

  Marvin Ammori, First Amendment Architecture, 2012 Wis. L. Rev. 1 (2012).   Hudgens v. NLRB, 424 U.S. 507 (1976). 110   Pruneyard v. Robins, 447 U.S. 74 (1980). 111   See text accompanying notes 10–11. 112   Joshua Fairfield discusses an example of this in his analysis of “mixed reality.” He considers a hypothetical situation where someone “defaces your home with an obscene word that appears when your house is viewed using a Mixed Reality application.” As Fairfield notes, this kind of situation is most problematic for the home owner where a particular AR “channel or application . . . become[s] dominant.” See Fairfield, supra note 9, at 59, 97, 101. 108 109

BARFIELD PRINT.indd 337

13/11/2018 11:53

338  Research handbook on the law of virtual and augmented reality

5. CONCLUSION A key Supreme Court case about billboards, Metromedia, Inc. v. San Diego, found the following when confronted with the question of whether the First Amendment’s free speech clause allows cities to place restrictions on billboards. Billboards have a double nature under the First Amendment. On the one hand, said the Court, billboards are huge permanent structures that “creat[e] a unique set of problems for land-use planning and development,” problems that government must be left with power to address. On the other, even if city governments must have power to address the land use problems generated by billboards, and the space they occupy, this does not mean that such governments can constitutionally regulate the messages or images on billboards.113 AR and VR, however, have the potential to make one side of this double character disappear. Giant billboards are still visible to individuals wearing AR glasses, or “driving” on the virtual reality highway. They still confront people with visible, attention-catching messages. But the permanence and space-filling character that accompanied this communication in solid billboards is gone, as are the land use problems it generates. All that is left of a billboard is its communicative or expressive dimension, the part that the state is foreclosed from restricting. Seen in this way, it may appear that AR and VR experiences are regulation-free zones, thanks to free speech protection. A closer look, however, reveals a more nuanced picture of AR and VR’s First Amendment status, and of what they tell us about First Amendment theory: Some elements of AR and VR environments may not be protected speech at all; others are protected speech but stubbornly continue to link to, and shape, physical world activity (such as physical movement) that government is duty-bound to monitor and regulate. Finally, some of the harms that free speech law gives government room to regulate flow not from the physical manifestations of speech, but from the meaning of the words or pictures. For better or worse, these harms follow expression into VR or AR.

SOURCES Cases Brown v. Entertainment Merchants Association, 131 S. Ct. 2729 (2011). Candy Lab, Inc. v. Milwaukee County, 2017 WL 3088380 (E.D. Wis. Jul. 20, 2017). Cent. Hudson Gas & Elec. Corp. v. Pub. Serv. Comm’n of N.Y., 447 U.S. 557 (1980). City of Renton v. Playtime Theatres, Inc., 475 U.S. 41 (1986). Connick v. Myers, 461 U.S. 138 (1983). Denver Area Educ. Telecomm. Consortium, Inc. v. FCC, 518 U.S. 727 (1996). Dunagin v. City of Oxford, Miss., 718 F.2d 738 (5th Cir. 1983). Heffron v. Int’l Soc’y for Krishna Consciousness, Inc., 452 U.S. 640 (1981). Hudgens v. NLRB, 424 U.S. 507 (1976). Int’l Soc’y for Krishna Consciousness v. Lee, 505 U.S. 672 (1992). James v. Meow Media, Inc., 300 F.3d 683 (6th Cir. 2002). Justice v. NCAA, 577 F. Supp. 356 (D. Ariz. 1983). Lone Star Security and Video, Inc. v. City of Los Angeles, 827 F.3d 1192 (9th Cir. 2016).

113

  Metromedia, Inc. v. City of San Diego, 453 U.S. 490, 502 (1981).

BARFIELD PRINT.indd 338

13/11/2018 11:53

Freedom of expression, and the personalization of public space  339 Madsen v. Women Health Center, Inc., 512 U.S. 753, 773. March v. Mills, 867 F.3d 46 (1st Cir. 2017). Marsh v. Alabama, 326 U.S. 501 (1945). Metromedia, Inc. v. City of San Diego, 453 U.S. 490 (1981). N.Y. Times Co. v. Sullivan, 376 U.S. 254, 270 (1964). Packingham v. North Carolina, 137 S. Ct. 1730 (2017). Perry Educ. Ass’n v. Perry Local Educators’ Ass’n, 460 U.S. 37 (1983). Pleasant Grove City v. Summum, 555 U.S. 460 (2009). Pruneyard v. Robins, 447 U.S. 74 (1980). Red Lion Broadcasting Co. v. F.C.C., 395 U.S. 367, 391–2 (1969). Reed v. Town of Gilbert, 135 S. Ct. 2218 (2015). Sorrell v. IMS Health, Inc., 131 S. Ct. 2653 (2011). Taxpayers for Vincent, 466 U.S. 789 (1984). Tinker v. Des Moines Indep. Cmty. Sch. Dist., 393 U.S. 503 (1969). United States v. Am. Library Ass’n, Inc., 539 U.S. 194, 205 (2003). Universal City Studios v. Corley, 273 F.3d 429 (2nd Cir. 2001). United States v. O’Brien, 391 US. 367 (1968). United States v. Stevens, 559 U.S. 460 (2010). United States v. Windsor, 133 S. Ct. 2675 (2013). Ward v. Rock Against Racism, 491 U.S. 781 (1989).

Statutes Milwaukee County Code of General Ordinances, Section 47.03(3). Restatement (Second) of Torts, Section 46.

BARFIELD PRINT.indd 339

13/11/2018 11:53

11.  Beyond unauthorized access: laws of virtual reality hacking Gilad Yadin*

1. INTRODUCTION In Ray Bradbury’s 1951 story The Veldt, two children develop an unhealthy attachment to a home virtual reality nursery. When their parents threaten to shut the nursery down, the children hack the virtual reality system and use it to trap and murder them.1 Bradbury’s dark, futuristic tale cautions against losing human values in a technological world, while reminding that any technology can be subverted. In the second decade of the twenty-first century, virtual reality technology is no longer fictional or futuristic. The $2 billion acquisition of virtual reality startup Oculus VR by social networking giant Facebook in 2014 sparked a research and development race that some are calling a virtual reality renaissance. Facebook founder Mark Zuckerberg said this about virtual reality: The incredible thing about the technology is that you feel like you’re actually present in another place with other people. People who try it say it’s different from anything they’ve ever experienced in their lives . . . Imagine enjoying a court side seat at a game, studying in a classroom of students and teachers all over the world or consulting with a doctor face-to-face­—just by putting on goggles in your home. This is really a new communication platform. By feeling truly present, you can share unbounded spaces and experiences with the people in your life.2

But virtual reality has a dark side. Malicious hacking poses a real threat to virtual reality platforms and users. Virtual worlds—limited online virtual reality environments—have already suffered damaging cyberattacks.3 This chapter introduces and characterizes *  This chapter is based in part on Gilad Yadin, Virtual Reality Intrusion, 53 Willamette L. Rev. 63 (2016), and other work by the author. The author would like to thank Tal Zarsky, Marc J. Blitz and Woodrow Barfield for their helpful comments. 1   See Ray Bradbury, The Veldt, in The Illustrated Man (1951). 2   Mark Zuckerberg, Facebook (Mar. 25, 2014), www.facebook.com/zuck/posts/10101319050​ 523971. 3   On August 2012, a cyberattack on World of Warcraft, a popular virtual game world, left significant damage within the virtual environment. See World of Warcraft Hack: Attack Kills Thousands of Players, Destroys Several Major Cities (VIDEO), Huffington Post, Oct. 10, 2012, www.huffingtonpost.com/2012/10/08/world-of-warcraft-hack-hacker-thousands-killedcities-destroyed_n_1948843.html; see also Michael Thumann, Address at the Black Hat Briefing: Hacking Second Life (Feb. 2, 2008), www.blackhat.com/presentations/bh-europe-08/Thumann/ Presentation/bh-eu-08-thumann.pdf (Second Life is yet another popular virtual world); F. Gregory Lastowka & Dan Hunter, Virtual Crime, in The State of Play: Law, Games, and Virtual Worlds 121 (Jack M. Balkin & Beth Simone Novack eds, 2006); Orin S. Kerr, Criminal Law in Virtual Worlds, 2008 U. Chi. Legal F. 415 (2008).

340

BARFIELD PRINT.indd 340

13/11/2018 11:53

Beyond unauthorized access  341 virtual reality hacking, a crime which deserves special attention by legal scholars as it may significantly impact existing cybercrime law. During the last two decades of the twentieth century, lawmakers in countries all over the world enacted new and specialized computer misuse legislation, imposing broad civil and criminal liability for unauthorized access of computer systems and data. This normative move was advocated by legal scholars and practitioners based on the notion that the considerably more constrained common law doctrines of criminal trespass and burglary are insufficient to the task of prosecuting and resolving computer hacking cases. Specialized computer misuse laws and their accompanying case law have become tenets of modern cyberlaw. Unfortunately, the unauthorized access regime central to these laws has created significant problems. One is the potential criminalization of everyday technological behavior, brought about by an overly extensive normative scope; another is chronic underenforcement; yet another is a wider chilling effect on creativity and digital freedoms. Debated solutions to these problems range from fixing the unauthorized access regime to abandoning it altogether. Virtual reality technology presents an intriguing challenge to legal computer misuse doctrine. Unlike other information technologies, virtual reality is built to deliver a psychological effect believably simulating the physical world; it possesses three-dimensional spatial characteristics, infuses users with real legal expectations, and mirrors human institutions and values. Many actions within virtual reality, lawful and criminal, are subjectively and conceptually closer to physical acts than to user actions in cyberspace—after all, virtual reality is specifically designed to convincingly imitate reality. Subsequently, considering some forms of virtual reality hacking may warrant a new look at the ancient common law doctrines of burglary and trespass as an alternative to the flawed modern computer misuse laws governing unauthorized access to computers and data. Section 2 of the chapter describes the appearance and evolution of broad unauthorized access provisions in specialized computer misuse legislation and the significant problems inherent to these provisions. It then presents the two competing approaches to solving these problems—reforming computer misuse laws or abandoning them altogether. Section 3 focuses on the legally meaningful characteristics of virtual reality environments and how they serve to distinguish the crime of virtual reality hacking from other forms of computer hacking. Section 4 examines, through different prisms, the applicability of the existing computer misuse legal framework—a framework based on broad concepts of authorization and access to computers and data—to virtual reality environments, suggesting that the appropriate normative framework for virtual reality hacking should be based on the ancient, time-tested common law doctrines governing physical intrusion. Section 5 discusses the implications of this move for virtual reality security and general computer misuse law.

2.  CRIMINAL HACKING LAW AND ITS SHORTCOMINGS When considering cybercrime, we commonly distinguish between a situation in which a computer serves as a tool in the commission of a crime and a situation in which a computer serves as the target of a crime; the former is typically handled by traditional criminal

BARFIELD PRINT.indd 341

13/11/2018 11:53

342  Research handbook on the law of virtual and augmented reality laws and the latter by specialized computer misuse statutes.4 Criminal hacking, where an offender targets and invades a computer system or network by unlawfully exceeding his or her privileges on the system, is a classic case of computer misuse. When such offenses first emerged, before specialized computer misuse legislation, legal academics and practitioners contemplated using traditional common law doctrine, particularly criminal trespass and burglary, to prosecute hacking crimes; this approach met with difficulty and was eventually abandoned.5 The enactment of wide-reaching legislation criminalizing unauthorized access to computers and networks in countries all over the world followed and a new legal regime emerged—an inherently flawed regime, which brings with it, as we will see, significant problems. Possible solutions range from fixing the law to abandoning it in favor of a return to traditional common law doctrine, a normative move with significant implications. This normative move may prove warranted when considering virtual reality hacking, a unique crime committed within a unique environment. 2.1  Hacking, Trespass, and Burglary Computer crime cases began appearing before the courts in the 1970s, sparking a lively discussion among academics, practitioners, and lawmakers regarding the legislation and prosecution of computer hacking and computer crimes in general.6 In the absence of specialized computer crime legislation, criminal trespass and burglary were deemed natural analogues to computer hacking.7 Intuitively, it appears that a physical intruder breaks into a building just as a computer hacker breaks into a computer system or network, and with similar intent. Criminal trespass and burglary are closely related offenses; the difference between them is in the criminal intent required to convict a perpetrator. A trespasser enters a premises knowing he or she is not legally entitled to enter said premises, while a burglar enters a premises with the additional purpose of committing an offense inside.8 The main problem with using criminal trespass and burglary provisions to prosecute computer hacking cases is that these offenses are closely tied to the physical world. There are specific spatial elements integral to both criminal trespass and burglary. As mentioned, the actus reus element of these offenses in the United States is based on the act of entering,9 interpreted by common law as crossing a threshold and the subsequent presence of a human body in a physical place.10 This is a fundamental constituent element of the crime, a substantive requirement framing a principal of right and wrong rather than a procedural one, defining what the prosecution is required to prove beyond reasonable doubt, rather than how an   See Scott Charney & Kent Alexander, Computer Crime, 45 Emory L.J. 931, 934 (1996).   Orin S. Kerr, Cybercrime’s Scope: Interpreting “Access” and “Authorization” in Computer Misuse Statutes, 78 N.Y.U. L. Rev. 1596, 1607 (2003).  6   See Jay Becker, The Trial of a Computer Crime, 2 Computer L.J. 441, 443 (1980); Susan H. Nycum, Legal Problems of Computer Abuse, 1977 Wash. U. L.Q. 527, 531 (1977); Kerr, supra note 5.  7  Kerr, supra note 5, at 1605­–6.  8   Susan W. Brenner, Is There Such a Thing as “Virtual Crime”? 4 Cal. Crim. L. Rev. 1, 20 (2001); Model Penal Code §§ 221.1(1); 221.2(1).  9   Id, Model Penal Code; Ronald J. Bacigal & Mary K. Tate, Criminal Law and Procedure: An Overview 111 (4th ed. 2014). 10  Kerr, supra note 5, at 1606–7.  4  5

BARFIELD PRINT.indd 342

13/11/2018 11:53

Beyond unauthorized access  343 unlawful act is to be prosecuted.11 The substantive social duty defined by trespass and burglary statutes is inherently spatial in nature; an act lacking spatiality cannot reasonably fall within the scope of these offenses. It is important to note that unlike burglary and criminal trespass to land, the tort of trespass to chattels was actually used extensively in computer-related civil actions, mostly involving bulk email.12 The successful, albeit heavily criticized, application of this cause of action to cyberspace can be attributed to the lessened burden of proof required in civil proceedings, as well as to the fact that trespass to chattel is not as closely linked to the physical world as its criminal cousins, focusing on interference with chattel rather than on physical entry.13 2.2  Computer Misuse Legislation From the mid-1970s to the mid-1980s, as the inadequacy of existing criminal law for the prosecution of computer crimes became evident, an increasing number of scholars, practitioners, and lawmakers pushed for specialized computer misuse legislation, addressing, among other things, hacking of computers and networks.14 In addition to that, lawmakers felt compelled to address overhyped public fears of powerful criminal hackers, partly brought about by the 1983 movie WarGames which depicted a near-nuclear war scenario as a result of hacking of a military computer system.15 The result was the legislation of computer misuse statutes in all US states and the enactment of the federal Computer Fraud and Abuse Act (CFAA) of 1986.16 The CFAA and all state statutes share a common foundation, criminalizing “unauthorized access” to computers.17 This general language is also shared by the UK Computer Misuse Act of 1990, as well as by the criminal laws of countries all over the world by way of the 2001 Council of Europe Convention on Cybercrime, which includes a model for computer misuse legislation.18 For computer misuse legislators, trespass and burglary served as a natural point of

  Bacigal & Tate, supra note 9, at 3.   Dan L. Burk, The Trouble with Trespass, 4 J. Small & Emerging Bus. L. 27 (2000) (criticizing this usage of the trespass to chattel tort). 13   Id at 28. 14   Professor Kerr estimates that 30–40 law review articles and notes published during that period advocated specialized computer crime statutes. Kerr, supra note 5, at 1613. 15   Stephanie R. Schulte, Cached: Decoding the Internet in Global Popular Culture 22 (2013); Jay P. Kesan & Carol M. Hayes, Mitigative Counterstriking: Self-Defense and Deterrence in Cyberspace, 25 Harv. J. L. & Tech. 429, 544 (2011); Matthew Kapitanyan, Beyond WarGames: How the Computer Fraud and Abuse Act Should Be Interpreted in the Employment Context, 7 I/S: J. L. & Pol’y for Info. Soc’y 405 (2012) (noting that the 1984 House Report supporting passage of specialized computer misuse legislation referenced the movie). 16   18 U.S.C. § 1030 (2016). 17   Susan W. Brenner, State Cybercrime Legislation in the United States of America: A Survey, 7 Rich. J. L. & Tech. 28 (2001). 18   C.E.T.S. No. 185, www.coe.int/en/web/conventions/full-list/-/conventions/rms/090 00016800​ 81561; see also Amalie M. Weber, The Council of Europe’s Convention on Cybercrime, 18 Berkeley Tech. L.J. 425 (2003); Susan W. Brenner, The Council of Europe’s Convention on Cybercrime, in Cybercrime: Digital Cops in a Networked Environment 207 (Jack Balkin et al. eds, 2007). 11 12

BARFIELD PRINT.indd 343

13/11/2018 11:53

344  Research handbook on the law of virtual and augmented reality departure.19 The conceptual similarities between the traditional offenses and the modern ones are still evident. One indication is that many computer misuse statutes use titles like “computer trespass” with regard to computer misuse.20 Another is that most of these laws are structured so that there is a lesser offense of knowingly accessing a computer without authorization and a greater offense of accessing a computer without authorization and with the intent of committing another offense, corresponding to trespass and burglary.21 2.3  Problems with Unauthorized Access Statutes The unauthorized access regime of the CFAA and similar legislation is considered extensive in scope. Key terms used in these statutes, such as “authority” and “access,” are deemed vague and, therefore, relatively open to interpretation by law enforcement, prosecutors, and courts; furthermore, key definitions like that of “computer” are very broad.22 Scholars believe that this poses a serious problem. For whatever reasons, many Facebook users—perhaps as much as 25 percent—supply false personal information on their profiles.23 Some of them will undoubtedly be surprised to learn that under the current US unauthorized access regime, they could be prosecuted as computer criminals. Providing false personal information on a profile violates Facebook’s terms of service,24 an action which can be construed as unauthorized access under the CFAA and actually has served as a basis for criminal prosecution in the past.25 This demonstrates the basic problem with unauthorized access statutes: Vague language has made them absurdly broad in scope, so much so that in the United States they could be considered unconstitutional.26 Under the vagueness doctrine, a statute would be held unconstitutional when citizens must necessarily guess at its meaning and differ as to its application.27 The US Supreme Court has ruled that the unconstitutionality of such statutes stems from the fact that they provide absolute and unlimited discretion to law enforcement as to their application.28

  See Kerr, supra note 5, at 1618.   Id. 21   See Brenner, supra note 8, at 22. 22   See Kerr, supra note 5, at 1619; Orin S. Kerr, Vagueness Challenges to the Computer Fraud and Abuse Act, 94 Minn. L. Rev. 1561 (2010). 23   See Doug Gross, Survey: 1 in 4 Users Lie on Facebook, CNN (May 4, 2012), edition.cnn. com/2012/05/04/tech/social-media/facebook-lies-privacy. 24   Statement of Rights and Responsibilities § 4(1), Facebook, www.facebook.com/legal/terms (last visited Oct. 18, 2016). 25   In the well-known cyberbullying case of United States v. Drew, the defendant was charged under the CFAA with violation of social network MySpace’s terms of service: see United States v. Drew, 259 F.R.D. 449 (C.D. Cal. 2009); see Kerr, supra note 22, at 1578; Kesan & Hayes, supra note 15, at 494. 26   See Kerr, supra note 22, at 1578; David J. Schmitt, The Computer Fraud and Abuse Act Should Not Apply to the Misuse of Information Accessed with Permission, 47 Creighton L. Rev. 423 (2014); David Thaw, Criminalizing Hacking Not Dating: Reconstructing the CFAA Intent Requirement, 103 J. Crim. L. & Criminology 907 (2013); Beryl A. Howell, Real World Problems of Virtual Crime, 7 Yale J. L. & Tech. 103, 113 (2005). 27   See Bacigal & Tate, supra note 9, at 8. 28   See id; Chicago v. Morales, 527 U.S. 41, 56 (1999); see also Drew, 259 F.R.D. at 463. 19 20

BARFIELD PRINT.indd 344

13/11/2018 11:53

Beyond unauthorized access  345 The vagueness problem with unauthorized access provisions is compounded by the fact that the meaning of “unauthorized access” is shifting as more of the world around us becomes computerized.29 The definition of “computer” under the CFAA is very broad, encompassing today things such as vehicles, wearable technology, home appliances, and, obviously, cellular phones.30 In the 1980s, computers were large, stationary, expensive things used mostly by financial institutions and the government. This model is no longer relevant today, as computers are everywhere and computer access is common and frequent. Going through authorization procedures has become a normal daily action, contributing to a closed digital society and adversely affecting other areas of cyberlaw.31 This is the second significant problem with the unauthorized access legal regime. Julie Cohen compellingly argues that broad unauthorized access laws have worked against an open digital culture by promoting secrecy, limiting creativity through unbalanced intellectual property rights enforcement, discouraging reverse engineering, and encouraging the use of invasive security measures.32 Professor Cohen sees the unauthorized access regime as a major factor limiting individual digital freedoms, particularly with regard to freedom of information, copyright law, trade secrets, patents, and the right to privacy.33 Criminal laws should clearly delineate socially acceptable and unacceptable behavior and allow for everyday practice while targeting wrongdoers; unauthorized access statutes fail to do so.34 For instance, it is unclear whether or not circumventing copyright protection systems constitutes unauthorized access.35 Additionally, unauthorized access laws sometimes target people who can hardly be considered wrongdoers, such as those who may have violated terms of a standard form contract.36 Another example is the use of invasive tracking software to enforce digital access control, usage which would not be considered legitimate in other contexts.37 29   See Julie E. Cohen, Configuring the Networked Self: Law, Code, and the Play of Everyday Practice ch. 8, at 13 (2012), www.juliecohen.com/page5.php; see generally Niva ElkinKoren, Affordances of Freedom: Theorizing the Rights of Users in the Digital Era, 6 Jerusalem Rev. Legal Stud. 96 (2012). 30   See 18 U.S.C. § 1030(e)(1) (2016) (“The term ‘computer’ means an electronic, magnetic, optical, electrochemical, or other high speed data processing device performing logical, arithmetic, or storage functions, and includes any data storage facility or communications facility directly related to or operating in conjunction with such device, but such term does not include an automated typewriter or typesetter, a portable hand held calculator, or other similar device”). 31   See Cohen, supra note 29 ch. 8, at 16 (arguing that the unauthorized access regime has encouraged secrecy, facilitated copyright enforcement, limited reverse engineering, and made invasive security measures an acceptable norm); Trevor A. Thompson, Terrorizing the Technological Neighborhood Watch: The Alienation and Deterrence of the White Hats under the CFAA, 36 Fla. St. U. L. Rev. 537 (2009); Randal C. Picker, Access and the Public Domain, 49 San Diego L. Rev. 1183 (2012). 32   See Cohen, supra note 29 ch. 8; Picker, supra note 31, at 1207. 33   See Cohen, supra note 29 ch. 8, at 16–24. 34   Id ch. 9, at 27–9. 35   The Digital Millennium Copyright Act of 1998 specifically uses vague language which corresponds to computer misuse legislation, such as “unauthorized access” and “control access,” to define circumvention of copyright protection systems. See 17 U.S.C. § 1201 (2016); see also Niva Elkin-Koren, Making Room for Consumers under the DMCA, 22 Berkeley Tech. L.J. 1119 (2007). 36   See Kerr, supra note 22, at 1578–9. 37   See Cohen, supra note 29 ch. 8, at 16.

BARFIELD PRINT.indd 345

13/11/2018 11:53

346  Research handbook on the law of virtual and augmented reality The third problem with unauthorized access laws is that they are poorly and ineffectively enforced. Intuitively, a broad criminal statute should lead to a large number of prosecutions; however, this is not the case with computer misuse legislation, specifically the federal CFAA. In 2014, US attorneys nationwide filed only 194 CFAA criminal cases, constituting a mere 0.334 percent of the total federal criminal cases filed by US attorneys that year.38 This certainly does not reflect existing cybercrime activity. The FBI Internet Crime Complaint Center (ICCC) logged 269,422 complaints in 2014, with more than $800 million in estimated damages.39 Internet security companies estimate, perhaps unobjectively, an economic impact for cybercrime comparable to that of the entire criminal trade in narcotics;40 according to the US Department of Justice, there are 12 new cybercrime victims every second.41 Incidentally, the 303,809 complaints received by the ICCC in 2010 resulted in merely six computer misuse convictions.42 The US Department of Justice admits that in comparison to other federal crimes, CFAA offenses are not charged frequently; prosecutors are directed to proceed with a CFAA prosecution only when it would serve a substantial federal interest.43 Additionally, the average sentence for CFAA offenses is below the minimum recommended by federal sentencing guidelines.44 It seems that computer misuse, at least under the CFAA, is de facto a low priority for law enforcement. Possible reasons include the relative difficulty in investigating and prosecuting computer misuse crime, the high cost of such investigations, and the fact that it is often not considered “real crime” by police and the public.45 The underenforcement of cybercrime should not come as a surprise: Underenforcement is a classic characteristic of white collar crime.46 A broad statute does not necessarily lead to increased or effective enforcement; financial and corporate crime, for example, is underenforced despite the incredibly broad normative scope of some of the relevant legislation.47 A broad scope statute that is sporadically enforced is doubly problematic

38   See U.S. Att’y Ann. Stat. Rep., U.S. Dept. of Justice Exec. Office for U.S. Att’ys Table 3B at 14 (2014), www.justice.gov/sites/default/files/usao/pages/attachments/ 2015/03/23/14statrpt.pdf. 39   See 2014 Internet Crime Report, Fed. Bureau of Investigation: Internet Crime Complaint Ctr. Rep. (2014), www.ic3.gov/media/annualreport/2014_IC3Report.pdf. 40   See Net Losses: Estimating the Global Cost of Cybercrime, Ctr. For Strategic & Int’l Studies (June 2014), www.mcafee.com/jp/resources/reports/rp-economic-impact-cybercrime2.pdf. 41   David Bitkower, Principal Deputy Assistant Att’y Gen. of the Crim. Div. at the Dep’t of Justice, Opening Remarks at the George Washington Law Review Symposium: Hacking into the Computer Fraud and Abuse Act: The CFAA at 30 (Nov. 6, 2015), www.gwlr.org/ observations-on-hacking-into-the-computer-fraud-and-abuse-act. 42   See 2010 Internet Crime Complaint Report, Fed. Bureau of Investigation: Internet Crime Complaint Ctr. Rep. at 5–6 (2010), www.ic3.gov/media/annualreport/2010 _IC3Report.pdf. 43   See Leonard Bailey, U.S. Dep’t of Justice Special Counsel for Nat’l Security, Address at the Black Hat Briefing: How the Dept. of Justice Uses the Fraud and Abuse Act (Aug. 5, 2015), www. blackhat.com/docs/us-15/materials/us-15-Bailey-Take-A-Hacker-To-Work%20Day-How-FederalProsecutors-Use-The-CFAA.pdf. 44   Id. 45   See Marc D. Goodman, Why the Police Don’t Care about Computer Crime, 10 Harv. J. L. & Tech. 465, 478 (1997). 46   See Edwin H. Sutherland, White-Collar Criminality, 5 Am. Soc. Rev. 7 (1940). 47   Such as the criminal provisions of the Sarbanes–Oxley Act of 2002. See Lisa H. Nicholson, The Culture of Under-Enforcement: Buried Treasure, Sarbanes–Oxley and the Corporate Pirate, 5

BARFIELD PRINT.indd 346

13/11/2018 11:53

Beyond unauthorized access  347 as it leads to lack of accountability and loss of deterrence.48 Additionally, underenforcement does not mitigate any of the problems inherent to overbroad computer misuse legislation, namely, a potential for unwarranted criminalization of everyday technological behavior at the unlimited discretion of law enforcement and a chilling effect on creativity and digital freedoms. In fact, it may serve to mask them; if computer misuse statutes were better enforced, their problematic nature would have been more evident to the public and to public officials, potentially moving them to act and remedy the situation. Much of the problematic nature of the CFAA legal regime is illustrated in the case of Aaron Swartz. Swartz, an accomplished entrepreneur and vigorous internet activist, was indicted in 2011 for downloading millions of academic journal articles from the JSTOR digital library through the MIT computer network, with the intent to publish them as part of his campaign for free and open access to information.49 Faced with multiple CFAA and other charges amounting to 50 years in prison, he committed suicide in 2013.50 The unusually heavyhanded law enforcement approach taken with Swartz was attributed to early federal law enforcement involvement,51 a desire to make an example of him,52 and bullying by prosecutors.53 Obviously, other cases with similar characteristics are handled differently, if at all.54 We may never know what truly motivated law enforcement actions in the Swartz case, but it is obvious that these actions were facilitated by a vague, sporadically enforced unauthorized access statute which unreasonably criminalizes anticopyright activism. 2.4  Two Approaches to Solving Unauthorized Access Problems The significant problems brought about by the unauthorized access legal regime had not gone unnoticed. Over the past two decades, scholars, activists, and policymakers, particularly in the United States, have raised several ideas regarding how to remedy some or all of these problems. These solutions range from encouraging a narrow interpretation DePaul Bus. & Comm. L.J. 321 (2007); John C. Coates, The Goals and Promise of the Sarbanes– Oxley Act, 21 J. of Econ. Perspectives 91 (2007). 48   See Nicholson, supra note 47, at 325–6. 49   Anne Cai, Aaron Swartz Commits Suicide, The Tech (Jan. 15, 2013), tech.mit.edu/V132/ N61/Swartz.html; Harold Abelson et al., Report to the President: MIT and the Prosecution of Aaron Swartz (July 26, 2013), Swartz-report.mit.edu/docs/report-to-the-president.pdf. 50   Id; Thaw, supra note 26, at 944; Sarah A. Constant, The Computer Fraud and Abuse Act: A Prosecutor’s Dream and a Hacker’s Worst Nightmare—The Case Against Aaron Swartz and the Need to Reform the CFAA, 16 Tul. J. Tech. & Intell. Prop. 231 (2013). 51   The MIT police contacted a Cambridge Police Department detective, who was incidentally a member of the New England Electronic Crimes Task Force and was working at one of the Task Force offices in a federal building in Boston. He responded accompanied by two other Task Force Members, one of them a special agent of the U.S. Secret Service. Abelson et al., supra note 49, at 21. 52   John Dean, Dealing with Aaron Swartz in the Nixonian Tradition: Overzealous Overcharg­ing Leads to a Tragic Result, Verdict (Jan. 25, 2013), verdict.justia.com/ 2013/01/25/ dealing-with-aaron-swartz-in-the-nixonian-tradition. 53   Lawrence Lessig, Prosecutor as Bully, Lessig (Jan. 12, 2013), www.lessig.org/2013/01/ prosecutor-as-bully. 54   See Bailey, supra note 43.

BARFIELD PRINT.indd 347

13/11/2018 11:53

348  Research handbook on the law of virtual and augmented reality of the CFAA by the courts,55 to making some changes to the language of the law,56 to a complete reform of unauthorized access norms.57 Concurrently, a bipartisan group of legislators are pushing for computer misuse reforms, so far unsuccessfully, in the wake of public outcry following the Aaron Swartz case.58 The proposed solutions are aimed at limiting the scope of unauthorized access provisions while maintaining the general normative framework of specialized computer misuse legislation;59 this, however, may not be the only viable approach. A more radical approach would be a full or partial abandonment of unauthorized access provisions in favor of a return to the laws governing criminal trespass, burglary, and similar physical world offenses. One justification for this is the notion, held by some notable scholars, that conceptual similarities between unauthorized access provisions and traditional laws, such as those criminalizing trespass and burglary, are substantial enough to make computer misuse laws redundant and unnecessary; therefore, a return to old criminal norms would be the preferred course of action.60 This justification is independent of the problematic nature of unauthorized access and corresponds with the greater theoretical notion that specialized computer legislation of any kind is unneeded and potentially harmful.61 Another justification would be that narrowing the scope of unauthorized access provisions is only a partial solution, as it does not address the underenforcement problem and may actually aggravate it by further detracting from law enforcement’s already diminished motivation to investigate and prosecute computer hacking cases. Thus, abandoning unauthorized access presents a fuller solution. The following sections of this chapter introduce virtual reality hacking and show that using unauthorized access provisions to prosecute virtual reality hacking crimes is often a poor choice of legal doctrine—criminal trespass and burglary are considerably better alternatives. Besides the value of this insight to legislators, judges, and lawyers who may someday soon find themselves considering virtual reality hacking cases, it suggests an uncertain future for the unauthorized access legal regime. This conclusion may be taken as a novel argument for the second approach to the problems of unauthorized access

 Kerr, supra note 22; Kerr, supra note 5; Schmitt, supra note 26.   See Thaw, supra note 26. 57   See Cohen, supra note 29, ch. 8 at 25–9. 58   Peter Whippy, Lofgren, Wyden, Paul Introduce Bipartisan, Bicameral Aaron’s Law to Reform Computer Fraud and Abuse Act, U.S. House of Representatives (Apr. 21, 2015), https://lofgren. house.gov/news/documentsingle.aspx?DocumentID=397911; Computer Fraud and Abuse Act Reform, Elec. Frontier Foundation, www.eff.org/issues/cfaa (last visited Oct. 23, 2016). 59   See Symposium 2015: Hacking into the CFAA, Geo. Wash. L. Rev. (Nov. 6, 2015), www.gwlr. org/symposia/symposium-2015-hacking-into-the-cfaa (video of panels 1, 2, and 4). 60   See, e.g., Brenner, supra note 8, at 28–31 (arguing against specialized cybercrime laws); see Joseph M. Olivenbaum, : Rethinking Federal Computer Crime Legislation, 27 Seton Hall L. Rev. 574 (1997) (arguing that federal unauthorized access statutes are unnecessary); see also id. 61   See Frank H. Easterbrook, Cyberspace and the Law of the Horse, 1996 U. Chi. Legal F. 207 (1996); see also Joseph H. Sommer, Against Cyberlaw, 15 Berkeley Tech. L.J. 1145 (2000); but see Lawrence Lessig, The Law of the Horse: What Cyberlaw Might Teach, 113 Harv. L. Rev. 501, 502 (1999) (“My claim is to the contrary . . . unlike Easterbrook, I believe that there is an important general point that comes from thinking in particular about how law and cyberspace connect”). 55 56

BARFIELD PRINT.indd 348

13/11/2018 11:53

Beyond unauthorized access  349 provisions: a general, probably gradual, return to traditional common law doctrine, a normative move which carries significant implications.

3.  VIRTUAL REALITY HACKING We live in an exciting time of rapid technological change; it seems like new technologies emerge every day. Virtual reality, however, should not be considered just one more new technology; it should be understood as something different and singular. Virtual reality technology is distinctly driven by insights into the workings of the human mind, affecting it directly and powerfully. Virtual reality users perceive themselves as present in an alternate environment. They perceive the presence of others and project their real-life morality and social behavior into the simulated environment. Furthermore, users project their spatial awareness into virtual reality environments, perceiving them as navigable physical places. These unique psychological characteristics make most instances of virtual reality hacking a unique form of crime with subjective spatial, social, and moral attributes. Considering virtual reality hacking may change the way the law perceives the general crime of hacking and the inherently flawed unauthorized access paradigm. 3.1  The Psychology of Virtual Reality Pragmatically, virtual reality is a psychotechnological means of manipulating perception via false sensory cues.62 It is not surprising that it has been of great interest to cognitive and social psychologists; virtual reality research is a well-established psychological field of study with a substantial body of published work based on extensive experimentation.63 It has long been understood that human perception is the result of an active mental process; the mind receives cues from various sensory sources and synthesizes them using memory and association to form a consistent internal model of the world.64 Psychologists disagree on whether the mind directly processes sensory cues or whether it forms a working hypothesis of perception and uses sensory cues to test it.65 Regardless of the exact method, it is widely accepted that the process of perception, particularly visual perception, involves a considerable amount of unconscious mental guesswork, or perceptual filling-in—the maintenance of an internal spatial model of the world based 62   Psychotechnology is the application of psychological methods to the solution of practical problems. See Psychotechnology, Merriam-Webster.com, www.merriam-webster.com/medical/ Psychotechnology (last visited Oct. 14, 2016). 63   For a detailed psychology literature review, see Jesse Fox et al., Virtual Reality: A Survival Guide for the Social Scientist, 21 J. Media Psychol. 95 (2009). See also Barbara O. Rothbaum, Using Virtual Reality to Help Our Patients in the Real World, 26 Depression & Anxiety 209 (2009). 64   See James J. Gibson, The Perception of the Visual World, Science 19 (May 4, 1951). 65   This is a discussion which has been going on for years among perception researchers. For a description of direct perception (or ecological) theory, see James J. Gibson, The Senses Considered as Perceptual Systems (1966); for the perception as hypothesis (or constructivist) theory, see Richard Gregory, The Intelligent Eye (1970). Some researchers argue that the mind actually uses both methods. See Joel Norman, Two Visual Systems and Two Theories of Perception: An Attempt to Reconcile the Constructivist and Ecological Approaches, 25 Beh. & Brain Sci. 73 (2002).

BARFIELD PRINT.indd 349

13/11/2018 11:53

350  Research handbook on the law of virtual and augmented reality on incomplete sensory stimuli.66 Manipulating this system is simply a matter of replacing real sensory cues with simulated cues, even rough ones, tricking the mind into creating and maintaining an artificial perception model and the subjective feeling of being present elsewhere; virtual reality technology does exactly that.67 Presence is the key psychotechnological element of virtual reality.68 But what does presence feel like?69 It is actually a very natural human experience. The mechanism of REM (rapid eye movement) sleep dreaming is not unlike virtual reality. The mind closes off external stimuli and provides simulated sensory cues designed to facilitate presence in a dream reality.70 It is the ability to create presence, the feeling of being present elsewhere, using technology, that makes virtual reality psychologically unique. Other experiences— browsing a website, watching a movie, reading this chapter—can be engaging, bringing about focus and concentration, but they do not create the psychological state of being present in a different place.71 Virtual reality environments strive to be as close as possible to physical environments. For the subjective mind of a user immersed in a simulated environment, virtual reality is, in that moment, the only reality.72 Researchers put virtual reality subjects on a ledge 30 floors up on the side of a skyscraper.73 They experience real physical fear, sweaty palms, and a racing heart. While they tell themselves that it is not real, it does not make the feeling go away—the primitive parts of the brain believe that there 66   See Rimona S. Weil & Geraint Rees, A New Taxonomy for Perceptual Filling-In, 67 Brain Res. Revs. 40 (2011). 67   See Martijn J. Schuemie et al., Research on Presence in Virtual Reality: A Survey, 4 Cyber. Psychol. & Behav.  183 (2001); Maria V. Sanchez-Vives & Mel Slater, From Presence to Consciousness Through Virtual Reality, 6 Nature Revs. Neurosci. 332 (2005); Julia Diemer et al., The Impact of Perception and Presence on Emotional Reactions: a Review of Research in Virtual Reality, 6 Frontiers in Psychol., Jan. 30, 2015, at 2. 68   See Jonathan Steuer, Defining Virtual Reality: Dimensions Determining Telepresence, 42 J. Comm. 73 (1992). 69   Note that the term presence is commonly used to describe a psychological state of consciousness, while the similar term immersion is used to describe the effect of a computerized virtual reality system. Users are present, systems are immersive. Social science virtual reality literature, like in the fields of psychology and communications, tends to use the term presence, while engineering and computer science literature uses immersion. In this chapter the term presence dominates, as the focus here is mostly on virtual reality users. See Mel Slater et al., Immersion, Presence, and Performance in Virtual Environments: An Experiment with Tri-Dimensional Chess, ACM Virtual Reality Software & Tech. at 163 (1996). 70   See J. Allan Hobson et al., Virtual Reality and Consciousness Inference in Dreaming, 5 Frontiers Psychol. (Oct. 9, 2014). 71   See Shamus Smith et al., Drowning in Immersion, Proceedings of UK-VRSIG at 5 (1998); Gerard Jounghyun Kim, Designing Virtual Reality Systems 5–8 (2005); Steuer, supra note 68, at 84, 93. 72   See Matthew Lombard & Theresa Ditton, At the Heart of It All: The Concept of Presence, 3 J. Computer Mediated-Comm. 1 (Sept. 1997); see also When Virtual Reality Feels Real, Science Daily (May 13, 2009), www.sciencedaily.com/releases/2009/05/ 090511091727.htm (discussing results of a series of experiments showing real physical effects of virtual reality presence). 73   See Mark Piesing, Virtual Reality: The Next Frontier, Fin. Times (Jul. 17, 2015), www. ft.com/cms/s/0/43e4aea0-2b2b-11e5-acfb-cbd2e1c81cca.html (discussing research work conducted at Oxford University). Virtual reality is considered very effective at treating phobias such as acrophobia (fear of heights). See Carlos M. Coelho et al., The Use of Virtual Reality in Acrophobia Research and Treatment, 23 J. Anxiety Disorders 563, 571 (2009).

BARFIELD PRINT.indd 350

13/11/2018 11:53

Beyond unauthorized access  351 is real danger.74 Only one in ten have the courage to jump. Subjects report that hitting the simulated concrete below creates a sensation akin to real pain.75 Virtual reality scholarship has gone further than this, exploring specific aspects of virtual reality presence. Two of the most researched aspects are spatial presence and social presence. These two concepts and the body of work which establishes them are particularly valuable to legal scholars with an interest in virtual reality. The term spatial presence means the specific perception of being physically situated within a geometrical spatial environment.76 As virtual reality users, we not only feel present in a different place, we form an awareness of the space around us as physical navigable space.77 We project our spatial awareness into the virtual reality environment and assign personal meaning to places within the environment. Communication scholars originally conceived the term social presence as a scale of effective communication as a function of a medium’s capacity to facilitate interpersonal awareness.78 In the context of virtual reality research, social presence is the awareness of others and of being in the company of others within a virtual reality environment. This awareness evolves into the perceived ability to assess others and act on that assessment, resulting in social and moral behavior analogous to real-world behavior.79 As virtual reality users, we acquire real-world awareness of other users. We believe in our ability 74  Piesing, supra note 73; see also Too Embarrassed to Ask: Virtual Reality Works for Games, But What about Real Life? Recode (July 22, 2016), www.recode.net/2016/ 8/4/12371450/jeremybailenson-stanford-university-virtual-reality-too-embarrassed-to-ask-podcast-transcript (discussing research work conducted at Stanford University). 75  Piesing, supra note 73; see also Recode, supra note 74. 76   See Thomas W. Schubert, A New Conception of Spatial Presence: Once Again, with Feeling, 19 Comm. Theory 161 (2009); Thomas Baumgartner et al., Neural Correlate of Spatial Presence in an Arousing and Noninteractive Virtual Reality: An EEG and Psychophysiology Study, 9 Cyberpsychol. & Behav. 30 (2006); David Waller et al., The Transfer of Spatial Knowledge in Virtual Environment Training, 7 Presence 129 (1998); Saniye T. Bulu, Place Presence, Social Presence, Co-Presence, and Satisfaction in Virtual Worlds, 58 Computers & Educ. 154 (2012); Sanchez-Vives & Slater, supra note 67. 77   Philip Brey, Virtual Reality and Computer Simulation, in Handbook of Information and Computer Ethics, 361, 362 (Kenneth Einar, Himma & Herman T. Tavani eds, 2008). 78   See John Short et al., The Social Psychology of Telecommunication (1976). 79   See Ralph Schroeder, Social Interaction in Virtual Environments: Key Issues, Common Themes, and a Framework for Research, in The Social Life of Avatars: Presence and Interaction in Shared Virtual Environments (Ralph Schroeder ed., 2002); Shanyang Zhao, Toward a Taxonomy of Copresence, 12 Presence 445 (2003); Kristine L. Nowak & Frank Biocca, The Effect of the Agency and Anthropomorphism on Users’ Sense of Telepresence, Copresence, and Social Presence in Virtual Environments, 12 Presence 481 (2003); Mel Slater et al., A Virtual Reprise of the Stanley Milgram Obedience Experiments, 1 PLOS ONE 1 (2006) (reconstructing the Milgram obedience experiments in virtual reality and finding similar behavior to real life); see also Indrajeet Patil et al., Affective Basis of Judgment-Behavior Discrepancy in Virtual Experiences of Moral Dilemmas, 9 Soc. Neurosci. 94 (2014) (finding virtual reality behavior similar to real life and different than textual representation, when faced with moral dilemmas); see generally Frank Biocca et al., Toward a More Robust Theory and Measure of Social Presence: Review and Suggested Criteria, 12 Presence 456 (2003); Giuseppe Riva et al., Affective Interactions Using Virtual Reality: The Link between Presence and Emotions, 10 Cyberpsychology & Behav. 45 (2007); Brian E. Mennecke et al., An Examination of a Theory of Embodied Social Presence in Virtual Worlds, 42 Decision Sci. 413 (2011).

BARFIELD PRINT.indd 351

13/11/2018 11:53

352  Research handbook on the law of virtual and augmented reality to understand other users and assess their behavior. This leads us to project real-world social and moral values into the virtual reality environment. These values then affect our behavior. Spatial presence and social presence make virtual reality a unique medium. Unlike other digital media, virtual reality can cause us to perceive, understand, feel, judge, and behave just as we would in physical reality.80 The difference between virtual reality and other digital media is not merely quantitative; studies utilizing brain imaging and other methods show that presence works on a deeper cognitive level, tricking the conscious mind into believing we are present in an alternate reality.81 3.2  Virtual Reality Hacking as a Unique Crime In criminal hacking incidents, a hacker has attempted or managed to acquire restricted access to a computer system in accordance with the privileges granted to the compromised user. Consider a scenario in which hacker Alice manages to acquire the login credentials of Facebook user Bob. Alice uses these credentials to hack Bob’s Facebook account by accessing it via her laptop. She then proceeds to browse Bob’s personal profile information, his private pictures, and information shared with Bob by his Facebook friends and groups. She ignores Facebook text chat messages from some of Bob’s friends who notice his online status. After a while, Alice logs off Facebook. Some, including lawyers and judges, would describe this incident in spatial terms: Alice broke into the site, entered Bob’s private profile, moved to other areas of the profile, and after a while exited the site. In legal academic circles, this conception of the digital medium in spatial terms, commonly termed the “cyberspace as a place” metaphor, is emphatically criticized as misleading to judges, legislators, and regulators; normatively harmful; and wrong.82 Consider a second scenario in which hacker Alice manages to acquire the login

80   This is what makes virtual reality such an effective training and learning platform; skills gained in virtual reality carry over extremely well to real life. See Neil E. Seymour et al., Virtual Reality Training Improves Operating Room Performance, 236 Annals Surgery 458 (2002); S.  Parsons & P. Mitchell, The Potential of Virtual Reality in Social Skills Training for People with Autistic Spectrum Disorders, 46 J. Intell. Disability Res. 430 (2002); Hsiu-Mei Huang et al., Investigating Learners’ Attitudes Toward Virtual Reality Learning Environments: Based on a Constructivist Approach, 55 Computers & Educ. 1171 (2010); Teresa Monahan et al., Virtual Reality for Collaborative E-Learning, 50 Computers & Educ. 1339 (2006); see also John Patrick Pullen, I Finally Tried Virtual Reality and It Brought Me to Tears, Time (Jan. 8, 2016), http://time. com/4172998/virtual-reality-oculus-rift-htc-vive-ces. 81   See Daniel S. Pine et al., Neurodevelopmental Aspects of Spatial Navigation: A Virtual Reality fMRI Study, 15 NeuroImage 396 (2002); Hunter G. Hoffman et al., Using fMRI to Study the Neural Correlates of Virtual Reality Analgesia, 11 CNS Spectrums 45 (2006); Corey J. Bohil, Virtual Reality in Neuroscience Research and Therapy, 12 Nature Revs. Neurosci. 752 (2011); Sanchez-Vives & Slater, supra note 67; Recode, supra note 74. 82   See Jonathan H. Blavin & I. Glenn Cohen, Gore, Gibson and Goldsmith: The Evolution of Internet Metaphors in Law and Commentary, 16 Harv. J.L. & Tech. 265 (2003); Mark A. Lemley, Place and Cyberspace, 91 Cal. L. Rev. 521 (2003); Dan Hunter, Cyberspace as a Place and the Tragedy of the Digital Anticommons, 91 Cal. L. Rev. 439 (2003); Timothy Wu, When Law & the Internet First Met, 3 Green Bag 2d 171 (2000); James Boyle, The Second Enclosure Movement

BARFIELD PRINT.indd 352

13/11/2018 11:53

Beyond unauthorized access  353 credentials of virtual reality environment user Charlie. She uses these credentials to hack Charlie’s virtual reality environment account by accessing it via a home virtual reality interface system. She finds herself present in Charlie’s private area within the virtual reality environment—his virtual reality home. Alice spends some time looking around and then exits the area, moving Charlie’s avatar out of his virtual reality home into a virtual reality street, a public access area. She keeps her distance from other user avatars in an attempt to avoid being noticed. Alice then enters a group access virtual reality area reserved to a group Charlie is a member of, ignoring some of his virtual reality friends who approach, wave, and verbally greet her. After a while, she exits the virtual reality environment.83 Unlike the previous scenario, it is almost impossible to describe this crime without using spatial terms. Virtual reality environments are inherently spatial and three-dimensional, with realistic physical limitations.84 The legal criticism of internet spatial metaphors should not apply here, since virtual reality environments are actually cognitively perceived as spatial by users. Spatiality is an integral part of their design. It may be a simulated environment, but to the mind of a user present in virtual reality, the environment is functional three-dimensional space; you move through it, around objects and people and along walls.85 Virtual reality hacking involves spatial action—unlawful spatial action—which mimics corresponding real-world spatial action. The action may not be real, but neither is it metaphorical. Spatial presence creates a psychological effect that is unique to virtual reality environments, distinguishing virtual reality hacking from other types of computer hacking. Both of the described hacking scenarios involve social interaction. In the first scenario, Alice ignores textual chat messages from other users. In the second scenario, she takes steps to bodily avoid other users on the street and ignores verbal and body language greetings from users who approach her in a group area. These differences between regular cyberspace social activity and virtual reality environment social activity are typical and meaningful. Virtual reality social interaction mimics real-life interaction. The interaction is rich and detailed, involving spatial movement, voice, and gestures. Social presence means that social and antisocial actions in virtual reality environments have a psychological effect similar to corresponding actions in real life. Virtual reality hacking is a crime that affects the perpetrator, the victim, bystanders, and society as a whole. The crime more closely mirrors the effects of physical crime than other types of hacking. Alice’s virtual reality hacking includes socially meaningful actions beyond her hacking into Facebook. Her interactions with both people and places have more substance because of the unique features of the medium within which they are taking place. The consequences of the crime for Charlie, whose virtual reality home was invaded, are different than the consequences

and the Construction of the Public Domain, 66 Law & Contemp. Probs. 33 (2003); Julie E. Cohen, Cyberspace as/and Space, 107 Colum. L. Rev. 210 (2007). 83   This is a purposefully simple scenario for illustration. Real-life scenarios will tend to be more complex, integrating hacking with other offenses. 84   See Grigore C. Burdea & Philippe Coiffet, Virtual Reality Technology 16 (2nd ed. 2003); Mario Gutierrez et al., Stepping into Virtual Reality 81 (2008). 85   See Stephane Redon et al., Interactive and Continuous Collision Detection for Avatars in Virtual Environments, IEEE Virtual Reality at 117 (Mar. 27–31, 2004).

BARFIELD PRINT.indd 353

13/11/2018 11:53

354  Research handbook on the law of virtual and augmented reality for Bob, whose Facebook account was compromised. The impact on Charlie’s virtual reality friends is not the same as the impact on Bob’s Facebook friends. Finally, the harmful effect of Alice’s virtual reality hacking on society is different in nature than the social effect of Alice’s hacking of Facebook. Virtual reality hacking possesses distinctive psychological attributes derived from the virtual reality medium: perceived spatiality and real-world social characteristics. These features make virtual reality hacking a substantially different crime than other types of hacking and very similar to the crimes of the physical world. The legal implications of this notion may go beyond virtual reality law, as we will see in the sections that follow.

4.  VIRTUAL REALITY HACKING LAW Criminal law prohibits behavior that represents a serious wrong against an individual or a social value.86 We have looked at a new kind of criminal behavior—virtual reality hacking. Like other types of computer hacking, virtual reality hacking cases would be prosecuted according to unauthorized access provisions within computer misuse laws such as the CFAA. But is this the best legal approach? The nature and impact of virtual reality hacking, both from the perspective of wronged individuals and from that of a wronged society, may actually allow such crimes to fall within the purview of criminal laws prohibiting physical intrusion, that is, trespass and burglary. This means that in at least some instances, criminal law should treat virtual reality environments as if they were real places. Furthermore, this conception casts a shadow on the future of existing specialized computer misuse legislation. Virtual reality hacking may be different in character than other types of hacking, but suggesting that these differences warrant a departure from computer misuse doctrine bears closer examination. This section of the chapter begins by examining the applicability, or rather inapplicability, of computer misuse doctrine to virtual reality hacking through a concrete analysis of the harmful effects of the crime. We will first look at trespass and burglary as viable alternatives in cases of virtual reality hacking by breaking these offenses down to their constituent elements. Subsequently, we will look at the choice of normative framework as a perspective problem. Finally, we will factor in reasonable expectations of virtual reality security. 4.1  Virtual Reality Hacking as Computer Misuse: A Bad Fit Every crime has an external effect: It alters preexisting conditions, constituting a socially intolerable harm.87 This principle of harm as the basis for criminalizing behavior is considered fundamental in criminal jurisprudence.88 The harmful effect, the socially intolerable harm, of computer misuse crimes can be described as interference with the 86   See Andrew Ashworth & Jeremy Horder, Principles of Criminal Law 1 (7th ed. 2013); Jerome Hall, General Principles of Criminal Law 213 (2nd ed. 2005). 87   See Hall, supra note 86, at 213, 220; Susan W. Brenner, Cybercrime Metrics: Old Wine, New Bottles?, 9 Va. J.L. & Tech. 13 (2004). 88   See Joel Feinberg, The Moral Limits of the Criminal Law: Harm to Others 31 (1984); John Stuart Mills, On Liberty 21 (1986).

BARFIELD PRINT.indd 354

13/11/2018 11:53

Beyond unauthorized access  355 proper functioning of computers and computer networks.89 This description may fit some computer hacking incidents; with regard to virtual reality hacking, however, it falls short. One of the harmful effects of virtual reality hacking is interference with the proper functioning of computers and computer networks. However, this is an incidental effect suffered by the virtual reality environment operator, a minor victim of the crime. The cardinal harm suffered by the main victims, compromised virtual reality users, is the violation of their private virtual space. The law has long recognized and strongly protected personal proprietary rights in land and chattel.90 This protection is broad enough to extend to various assets which cannot be considered classic property, such as public entitlements, or even internet domain names.91 An even more robust legal protection is accorded to possessory interests in private spaces such as homes and offices.92 There are several justifications for these legal protections. One justification is the importance accorded to the security and integrity of a home.93 Another is the fundamental importance of connections between people and their personal possessions, connections that make these possessions a part of their personality.94 Similar connections exist within virtual reality environments as a function of the spatial and social characteristics of virtual reality. The technology is designed to affect users in exactly this manner.95 Through presence, users form a connection to places in virtual reality—users spend time with other users in their virtual reality area, personalize it with virtual reality objects, and create meaningful memories. Eventually, in these users’ minds, the place becomes an extension of their physical home. Furthermore, attachment to people in virtual reality is linked to connections with places and objects, just like in the physical world. Places and objects often provide context, a foundation for interpersonal relationships. Without implying that a property rights regime is appropriate for virtual reality environments,96 protection through criminal law of the personal attachment that users   See Kerr, supra note 5, at 1603.   See William Blackstone, Commentaries on the Laws of England II 2 (1766). 91   See Charles Reich, The New Property 73 Yale L. J. 733 (1964) (public entitlements); see Anupam Chandler, The New, New Property, 81 Tex. L. Rev. 715 (2003) (Internet domain names). 92   Particularly constitutional protections from government search and seizure in the U.S. See Katherine J. Standburg, Home, Home on the Web and Other Fourth Amendment Implications of Technosocial Change, 70 Md. L. Rev. 614 (2011); Stephanie M. Stern, The Inviolate Home: Housing Exceptionalism in the Fourth Amendment, 95 Cornell L. Rev. 905 (2010); U.S. Const. amend. IV. 93   See Charles P. Nemeth, Criminal Law 361 (2nd ed. 2012); Philip E. Carlan et al., An Introduction to Criminal Law 51 (2011). 94   Based on Hegelian philosophy, Radin’s Personhood approach rationalizes the strong legal protection of private property as a function of the connection between person and property, a connection that is vital to the achievement of human personal interests. People and society possess a strong interest in protecting private property as an extension of self, an instrument of personal growth and fulfillment. See Margaret Jane Radin, Property and Personhood, 34 Stan. L. Rev. 957 (1982); Margaret Jane Radin, Reinterpreting Property (1993); Georg W.F. Hegel, Philosophy of Right (1821). See generally Milton Friedman, Capitalism and Freedom 7–32 (1962) (discussing the link between property rights and freedom). 95   See supra section 3.2; Schubert, supra note 76. 96   This is a debated point with regards to virtual worlds which is beyond the scope of this paper. See Jack M. Balkin, Law and Liberty in Virtual Worlds, in The State of Play: Law, Games and Virtual Worlds 86 (Jack M. Balkin & Beth Simone Novack eds, 2006); Yochai Benkler, There Is No Spoon, in The State of Play: Law, Games and Virtual Worlds 180 (Jack M. Balkin & Beth 89 90

BARFIELD PRINT.indd 355

13/11/2018 11:53

356  Research handbook on the law of virtual and augmented reality form with places and people in virtual reality can be justified in the same way in which real-world legal protections of such attachments can be justified. Categorizing all virtual reality hacking incidents as computer misuse crimes calls for legal institutions to focus, in such cases, on the harm caused to the virtual reality system, possibly missing the greater harm caused to virtual reality victims and, through them, to society at large. This is not surprising as computer misuse laws were created to protect interests in computer systems, their integrity, and their proper operation, rather than interests of a more personal nature.97 Traditional crime is typically personal; it is perpetrated within a community and in a specific location, facts that limit it and lend it social context. Computer misuse, however, is characterized by distance, physical and conceptual, between perpetrator and victim, as well as by the absence of a defined crime scene.98 It may seem counterintuitive to label virtual reality hacking a traditional crime, but it is perpetrated within a community of sorts, in a specific location—albeit not a physical one—and possesses a social context. This does not mean that virtual reality hacking is a personal crime in the sense that there is a likely social connection between perpetrator and victim, but it does suggest that virtual reality hacking is personal enough to, once again, fit poorly within this characterization of computer misuse. To emphasize the point: Virtual reality technology, with its spatial and social features, affects users in ways which mimic physical reality. Correspondingly, the crime of virtual reality hacking affects users in a personally harmful manner, bringing to mind the harmful effects of real-world criminal behavior on both users and society. Computer misuse doctrine falls short as its frame of reference is remote, nonspatial crime, lacking personal or social context. Virtual reality environments are designed with the express purpose of providing a realistic personal experience in a seemingly real geometrical space, facilitating rich social interactions. Protecting interests in these environments appears to be more in line with the aims of real-world criminal law. 4.2  Another Look at Trespass and Burglary As previously mentioned, while there was a time when prosecuting computer hacking cases as trespass or burglary crimes seemed like a plausible course of action, this line of thinking has been abandoned by lawmakers in favor of specialized computer misuse legislation.99 When considering virtual reality hacking, however, trespass and burglary statutes seem more relevant than ever. The characteristics that make computer misuse laws a bad fit for virtual reality hacking cases—perceived spatiality and projected social values leading to personal harmful effects—make trespass and burglary seem applicable. This is supported by examining the elements of trespass and burglary offenses against such incidents of virtual reality hacking. Simone Novack eds., 2006); F. Gregory Lastowka & Dan Hunter, The Laws of the Virtual Worlds, 92 Cal. L. Rev. 1 (2004); Andrew E. Jankowich, Property and Democracy in Virtual Worlds, 11 B.U. J. Sci. & Tech. L. 173 (2005). 97   See Olivenbaum, supra note 60, at 590. 98   See Marc D. Goodman & Susan W. Brenner, The Emerging Consensus on Criminal Conduct in Cyberspace, 10 Int’l JL & Info. Tech. 139, 151 (2002). 99   Abandoned by lawmakers but not by all scholars, as we have seen. See supra section 1.4.

BARFIELD PRINT.indd 356

13/11/2018 11:53

Beyond unauthorized access  357 In order to prove criminal trespass in the United States, the prosecution needs to show the actus reus of entering a building or occupied structure, the mens rea of knowing that there is no license or privilege to do so, and the attendant circumstances that there is indeed no license or privilege for entry.100 In order to prove burglary, the prosecution needs to show the same actus reus and attendant circumstances with the more serious mens rea of entering with the purpose of committing a crime inside.101 The term “occupied structure” is widely defined to include any structure, vehicle, or place adapted for accommodation or carrying on business, whether or not a person is present there at the time.102 It is easy to see why attempts to apply these criminal statutes to cases of computer hacking were unsuccessful. A computer is not a structure or place and a computer hacker does not enter it, except in an abstract metaphorical sense.103 Virtual reality hacking, however, is different. Criminal law theorists have long been interested in the nature of criminal action, suggesting that there is more to a criminal act than physical movement.104 Criminal action has a subjective quality. It is a function of human perception and subjective understanding.105 Neuroscientists say that donning virtual reality gear and connecting with a virtual reality environment transports a person’s consciousness to an alternate reality. Users report a feeling of physically entering and being in a different place.106 When users use terms like “entering” or “place,” they are not being metaphorical. Rather, they are describing their subjective experiences. We can expect a virtual reality hacker to describe his or her experience in similar terms, fulfilling the mens rea requirement for trespass or burglary. We can further assume that any witnesses to the hacker’s actions would describe him or her entering a place, fulfilling the actus reus and attendant circumstances requirements. Even people outside the virtual reality environment—law enforcement, prosecutors, courts—would likely put the criminal behavior in physical spatial terms. It would be almost impossible to describe the criminal behavior, and subsequently understand it, in any other way. Courts have asked certain types of questions in criminal trespass and burglary proceedings for hundreds of years. Did part of the defendant’s person pass the threshold?107 Did the defendant set foot upon the ground?108 Can the trespassed land be considered a home?109 These types of questions are inapplicable to most hacking cases but relevant to virtual reality hacking. This relevance is a function of the spatial and social features inherent to virtual reality environments. The relevance of common legal tests used in trespass   See Model Penal Code § 221.1(1).   Id § 221.2(1). 102   Id § 221.0(1). 103   See Kerr, supra note 5, at 1607. 104   See, e.g., Michael S. Moore, Act and Crime: The Philosophy of Action and Its Implications for Criminal Law 9 (Tony Honore & Joseph Raz eds., 1993). 105   See George P. Fletcher, Rethinking Criminal Law 436 (2000) (defining the perception of human acting in the criminal context as a form of intersubjective understanding). 106   Sanchez-Vives & Slater, supra note 67, at 338. 107  Kerr, supra note 5, at 1607. 108   Entick v. Carrington, 19 Howell’s State Trials 1029 (1765). 109   See Orin S. Kerr, Norms of Computer Trespass, 116 Colum. L. Rev. 1143 (2016); Standburg, supra note 92. 100 101

BARFIELD PRINT.indd 357

13/11/2018 11:53

358  Research handbook on the law of virtual and augmented reality and burglary jurisprudence to virtual reality hacking underscores the applicability of physical intrusion norms to the crime and the normative difference between virtual reality hacking and other forms of computer hacking. 4.3  Virtual Reality Hacking as a Problem of Perspective Choosing a suitable normative framework for virtual reality hacking can also be examined as a problem of legal perspective. Perspective has been studied as a means of analyzing social systems, including the law itself.110 Orin Kerr suggests that in the context of the internet, perspective is legally meaningful and often determines the way that cyberlaw norms are applied to behavior.111 For example, does sending a bomb threat via instant messaging to someone located in the same state constitute an interstate federal crime if the instant messaging service servers are located in a different state?112 From an external perspective, the message crossed state lines. Electronic impulses representing the message traveled at the speed of light from one state to another state and back to the first state. From an internal perspective, however, the message was sent and received within the same state, which would be the view of both sender and recipient. If we choose to view the message as a logical whole given the circumstances of the case, this would be their subjective legal expectation and also a reasonable legal perspective.113 When considering virtual reality environments, the notion of legally meaningful perspective becomes especially definitive. Take, for example, a virtual reality sculpture within a private virtual reality habitat. From an internal perspective, the sculpture is conceptually close to a real-life sculpture; it can be studied and artistically appreciated by its owners, has volume, is placed at a specified location within the habitat, and is secure. From an external perspective, the sculpture is intangible stored digital data, interpreted and represented via a graphical interface. Now consider a virtual reality crime where a hacker destroys the sculpture. Which perspective should the law adopt? An external perspective leads to prosecuting this illicit act as computer misuse. An internal perspective suggests prosecuting the act as burglary. The choice of normative framework can be viewed as a function of the adopted perspective. Virtual reality is a subjective medium, designed to facilitate a cognitive detachment from physical reality. This design makes virtual reality fundamentally different than other digital media. An external perspective on virtual reality takes away everything that makes it virtual reality, leaving nothing more than computer infrastructure. A virtual reality environment can be broken down into its logical parts. More than other systems, however, when a virtual reality environment is whole it creates something which is beyond the sum

110   See Douglas Litowitz, Internal Versus External Perspectives on Law: Toward Mediation, 26 Fla. St. U. L. Rev. 127 (1998); see also Scott Brewer, Scientific Expert Testimony and Intellectual Due Process, 107 Yale L. J. 1535 (1998). 111   Orin S. Kerr, The Problem of Perspective in Internet Law, 91 Geo. L.J. 357 (2003). 112   These are, roughly, the facts of a 1999 10th Circuit case. See id at 373; United States v. Kemmersell, 196 F.3d 1137 (10th Cir. 1999). 113   The court in the aforementioned case chose the external perspective and affirmed the conviction. In other, similar cases, however, courts chose the internal perspective. Kerr, supra note 111, at 374.

BARFIELD PRINT.indd 358

13/11/2018 11:53

Beyond unauthorized access  359 of its parts. Taking an external perspective on virtual reality—looking at it as a collection of bits and bytes—misses the essence of the virtual reality idea and its technological application. A possible analogy would be taking an external legal perspective on a person, seeing a collection of interconnected organs and missing the essence of a human being. Taking a legal bits-and-bytes perspective on virtual reality hacking amounts to legally viewing a case of assault and battery in terms of the damage done to a bruised arm or a broken leg instead of the damage to a living person. The law recognizes that a person is more than the sum of his or her parts. This is why there is a qualitative legal difference between damaging a person and damaging a piece of furniture. Similarly, a qualitative difference exists between criminally targeting a virtual reality environment and criminally targeting other computerized systems. A virtual reality environment adds a conceptual level of abstraction, a simulation of reality absent from websites and social networks. Taking a bits-and-bytes view of virtual reality hacking ignores the unique effects of the crime and unjustly reduces it to generic computer misuse. Justifying an internal perspective of virtual reality hacking serves to affirm the choice of trespass and burglary norms to regulate these crimes. 4.4  Protecting Expectations of Virtual Reality Security Protecting reasonable expectations is at the heart of many legal fields.114 Criminal law, however, is usually more concerned with harmful effects on society at large than with expectations of individuals.115 With that fact in mind, criminal law does recognize that rights and interests are affected by reasonable individual expectations.116 Most victims who suffer both online and offline crime feel that there are fewer ways to get help after falling victim to a cybercrime. Only one in four victims say that they have confidence in the ability of law enforcement to investigate cybercrime, and only one in five victims actually report cybercrime to police.117 Falling victim to cybercrime is one of the top overall fears of the average American.118 The general legal protection and law enforcement expectations of cybercrime victims, and potential cybercrime victims, are low, particularly relative to real-world actual and potential crime victims. Subjective expectations for legal protection of virtual reality users, however, should mirror their real-world security expectations. We have seen how spatial and social presence affects virtual reality users and facilitates an attachment to people and places similar to 114   See Bailey H. Kuklin, The Plausibility of Legally Protecting Reasonable Expectations, 32 Val. U. L. Rev. 19 (1997); Roscoe Pound, An Introduction to the Philosophy of Law 189 (1922). 115  Kuklin, supra note 114, at 24. 116   Feinberg, supra note 88, at 44; Helen Nissenbaum, Where Computer Security Meets National Security, in Cybercrime: Digital Cops in a Networked Environment 59, 65 (Jack Balkin et al. eds, 2007). 117   2011 Cybercrime Report, Symantec (2011), www.symantec.com/content/en/us/home_ homeoffice/html/cybercrimereport; Samuel Rubenfeld, Cyber Crime Emerging as Economic Threat, Wall Street Journal, Feb. 25, 2016, 12:01 AM, blogs.wsj.com/riskandcompliance/2016/02/25/ cybercrime-surging-as-economic-crime-threat. 118   According to a 2015 survey of 1,500 Americans conducted at Chapman University in California. See Elizabeth Palermo, What Do Americans Fear Most? Big Brother & Cybercrime, Live Science (Oct. 20, 2015), www.livescience.com/52535-american-fear-survey-2015.html.

BARFIELD PRINT.indd 359

13/11/2018 11:53

360  Research handbook on the law of virtual and augmented reality real-world attachments. This may lead to subjective expectations of legal protection on par with the legal protection users enjoy in the real world. At the very least, users should expect a higher level of legal protection when engaged in virtual reality environments compared to other digital media. These security expectations are legally meaningful, serving to further justify the application of real-world criminal norms to virtual reality environments.

5. IMPLICATIONS Prosecutors have used specialized computer misuse laws to prosecute computer hacking cases for 30 years. The type of system targeted by hackers was never a major concern for the law. In broad terms, if the target is a computer and it was accessed without authorization, the act can be considered computer misuse.119 Choosing the old common law trespass and burglary normative framework over the current unauthorized access legal regime for virtual reality hacking crimes carries significant implications for virtual reality security as well as for computer misuse law. 5.1  Stronger Virtual Reality Security We have seen that incidents of computer hacking are a low priority for law enforcement. This is mirrored by low user confidence in law enforcement when it comes to cybercrime. Now consider our new form of criminal behavior—virtual reality hacking. Applying the label “computer misuse” to these crimes seems, practically speaking, less than ideal. Trespass and burglary, however, are “real crimes,” familiar to the public, law enforcement, and the courts. Unlike computer misuse, these laws are vigorously enforced. Police in the US make around 160,000 burglary arrests every year.120 We cannot reasonably expect law enforcement to treat virtual reality hacking as a real-world crime, but applying the right label can make a difference and affect policy. Part of the problem is that cybercrime investigations are considered less than glamorous by law enforcement; investigating burglaries in virtual reality may seem more attractive.121 Likewise, “virtual reality burglaries” may sound more socially harmful and more pressing to the public and public officials than “incidents of computer misuse,” affecting public opinion and public policy. Furthermore, trespass and burglary may carry a stronger social stigma than cybercrime. Labeling virtual reality hackers as trespassers and burglars may serve to deter potential perpetrators. Finally, treating virtual reality hacking incidents as physical crimes may push virtual reality operators to invest more resources in securing these environments. 119   For guidelines detailing practical computer misuse prosecution considerations see Prosecuting Computer Crimes: Computer Crime and Intellectual Property Section Criminal Division: OLE Litigation Series, U.S. Dept. of Justice, justice.gov/criminal/cybercrime/docs/ccmanual.pdf. 120   From 1994 to 2011, arrests were made in roughly 10 percent of burglaries reported to police. See Household Burglary, 1994–2011, U.S. Dept. of Justice (June 2013), www.bjs.gov/content/pub/ pdf/hb9411.pdf. 121  Goodman, supra note 45, at 478.

BARFIELD PRINT.indd 360

13/11/2018 11:53

Beyond unauthorized access  361 This last point is especially salient. Characterizing virtual reality hacking cases as physical crimes may have a wider effect on the relationship between virtual reality users and virtual reality operators. From a legal point of view, these relationships are typically contractual, according relatively weak legal rights to users.122 The standard form technological service agreements typical of such relationships have been criticized as unfair, rarely read or understood, and legally and economically ineffective.123 These agreements therefore create an asymmetrical, unbalanced legal relationship. Recognition of the idea that hacking into a user’s personal virtual reality space is normatively akin to physical intrusion may have a balancing effect on these relationships. We can therefore expect a transition from unauthorized access in favor of physical intrusion norms to positively affect the security of virtual reality environments. This positive effect would be the result of incentivizing virtual reality operators to better secure these environments, empowering virtual reality users and motivating law enforcement to investigate and prosecute offenders. 5.2  Moving Beyond Unauthorized Access We have previously looked at the problems with unauthorized access: computer misuse statutes such as the CFAA use vague language leading to an overly broad scope; the unauthorized access legal regime has negative effects on digital creativity and digital freedoms; and despite their broad scope, computer misuse statutes are underenforced. We have also seen two approaches to solving these problems. One approach is to limit the scope of unauthorized access provisions. A more radical approach is to partly or fully abandon specialized computer misuse legislation in favor of the old common law doctrines of trespass and burglary. Acknowledging that the unauthorized access legal regime is a bad choice for virtual reality hacking cases creates a strong argument for abandoning the unauthorized access regime altogether. If this problematic legal regime becomes irrelevant in the face of an emergent technology, its relevance for existing technologies must be questioned. Attempting to fit virtual reality into this normative framework highlights the shortcomings of unauthorized access. Trespass and burglary are examples of clearly defined criminal behavior that take place within a real or virtual community. Unauthorized access to a computer is a conceptually abstract technological event. The natural function of criminal law is to prohibit and deter concrete socially harmful behavior. It is ill-suited to dealing with abstract technological functionality. Taking a real-world perspective on some computer crimes can easily begin with virtual reality environments and move toward

122   This is the case in virtual worlds. See Joshua Fairfield, Anti-Social Contracts: The Contractual Governance of Virtual Worlds, 53 McGill L. J. 427 (2008); Benkler, supra note 96, at 180. 123   See Robert L. Oakley, Fairness in Electronic Contracting: Minimum Standards for NonNegotiated Contracts, 42 Hous. L. Rev. 1041 (2005); see also Robert A. Hillman & Jeffrey J. Rachlinski, Standard Form Contracting in the Electronic Age, 77 N.Y.U. L. Rev. 429 (2002); Mark A. Lemley, Terms of Use, 91 Minn. L. Rev. 459 (2006); Shmuel I. Becher & Tal Z. Zarsky, E-Contract Doctrine 2.0: Standard Form Contracting in the Age of Online User Participation, 14 Mich. Telecomm. Tech. L. Rev. 303 (2008).

BARFIELD PRINT.indd 361

13/11/2018 11:53

362  Research handbook on the law of virtual and augmented reality other criminal behaviors, facilitating a paradigm shift in cybercrime law. Computer misuse legislation has moved too far from the real world, viewing cyberspace as something too alien for existing criminal law, forgetting that the law governs human behavior and not technological functionality. Cyberspace is growing more complex. The old distinctions are eroding. The clear lines that were once drawn between real-world crime and cybercrime are blurring—or perhaps they were never as clear as we thought. Virtual reality technology may herald the end of unauthorized access and a gradual return to classic criminal law; problems born of technological complexity sometimes call for simpler legal solutions.

SOURCES Cases Chicago v. Morales, 527 U.S. 41 (1999) United States v. Drew, 259 F.R.D. 449 (C.D. Cal. 2009) United States v. Kemmersell, 196 F.3d 1137 (10th Cir. 1999)

Statutes Computer Fraud and Abuse Act of 1986, 18 U.S.C. § 1030 (2016) Convention on Cybercrime, Nov. 23, 2001, E.T.S. No 185 Digital Millennium Copyright Act of 1998, 17 U.S.C. § 1201 (2016) Model Penal Code Public Company Accounting Reform and Investor Protection Act of 2002, 107 Pub. L. No. 204, 116 Stat. 745 U.S. Const. amend. IV

BARFIELD PRINT.indd 362

13/11/2018 11:53

12.  The law and ethics of virtual sexual assault John Danaher

In 1993, Julian Dibbell wrote an article in The Village Voice describing the world’s first virtual rape.1 The incident he described took place in a virtual world called LambdaMOO, which is a text-based virtual environment that still exists to this day. People in LambdaMOO create avatars (onscreen “nicknames”) and interact with one another through textual descriptions. Dibbell’s article described an incident in which one character (Mr Bungle) used a “voodoo doll”2 program to take control of two other user’s avatars and force them to engage in sexual acts with both his avatar and one another. At least one of the victims found the incident extremely traumatic, with Dibbell describing her later as having “postraumatic tears” streaming down her face when she spoke about it.3 In 2003, a similar incident took place in Second Life. Second Life is a well-known virtual world, created by the engineer and entrepreneur Philip Rosedale. It is visual rather than textual in nature. People create virtual avatars that can interact with other user’s avatars in a reasonably detailed visual virtual environment. In 2007, the Belgian Federal Police announced that they would be investigating a “virtual rape” incident that took place in Second Life back in 2003.4 Little is known about what actually happened, and the investigation does not appear to have progressed to a charge, but taking control of another character’s avatar and forcing it to engage in sexual acts was not unheard of in Second Life. Indeed, regular users noted that the use of voodoo doll program to engage in acts of “rape” was well known, though usually some form of “consent” was sought.5 In October 2016, another instance of sexual misconduct in a virtual environment was reported. In a widely discussed article, the journalist Jordan Belamire reported how she had been sexually assaulted while playing the game QuiVR, using the HTC Vive.6 The HTC Vive is an immersive virtual reality system. Users don a headset that puts a virtual 1   Dibbell, I. “A Rape in Cyberspace: How an Evil Clown, a Haitian Trickster Spirit, Two Wizards, and a Cast of Dozens Turned a Database into a Society,” The Village Voice, December 23, 1993. The article also appears as Chapter 1 in Dibbell’s My Tiny Life (New York: Holt Publishers, 1998). 2   A voodoo doll program is simply one that allows one user to take control of another user’s character or program. 3   Dibbell (n 1) id. 4   Lynn, R. “Virtual Rape Is Traumatising but Is it a Crime?” Wired Magazine, May 4, 2007. 5   On this, see the discussion of the Second Life case by Benjamin Duranske: “Reader Roundtable: ‘Virtual Rape Claim Brings Belgian Police to Second Life’,” April 24, 2007, Virtually Blind, available at http://virtuallyblind.com/2007/04/24/open-roundtable-allegations-of-virtualrape-bring-belgian-police-to-second-life/. I put the terms “rape” and “consent” in quotation marks because, as we shall see below, they are being discussed outside of their ordinary usage. Even the use of “rape” and “consent” in the same sentence is confusing since in many legal systems rape is simply defined as nonconsensual sex, so one simply could not consent to a rape. 6   Belamire, J. “My First Virtual Reality Groping,” Medium: Athena Talks, October 5, 2016, https://medium.com/athena-talks/my-first-virtual-reality-sexual-assault-2330410b62ee.

363

BARFIELD PRINT.indd 363

13/11/2018 11:53

364  Research handbook on the law of virtual and augmented reality environment into their visual field. The user interacts with that environment from a first-person perspective, with the overall goal of the headset being to create an immersive virtual experience (to get you “lost” in the virtual world).7 QuiVR is an archery game where players fight off marauding zombies. It can be played online with multiple users. Players appear to other users in a disembodied form as a floating helmet and a pair of hands. The only indication of gender comes through choice of name and voice used to communicate with other players. Jordan Belamire was playing the game in her home. While playing, another user, with the onscreen name of “BigBro442,” started to rub the area near where her breasts would be (if they were depicted in the environment). She screamed at him to “stop!” but he proceeded to chase her around the virtual environment and then to rub her virtual crotch. Each of these three incidents involves what might be termed “virtual sexual assault.” They each involve actions that take place in virtual environments in which virtual characters, controlled by humans, engage in representations of sexually explicit acts (sexual touching, intercourse, and so on) without the consent of at least one of their users. Clearly some of the participants were psychologically harmed by what happened. There are differences between the three cases. They involve different types of virtual environment, ranging from the simple text-based world of LambdaMOO to the more complex, immersive visual world of QuiVR. Nevertheless, the three incidents raise similar ethical and legal questions. Prima facie, there is something undesirable about the conduct involved. But how serious is it and what should we do about it? Should we really refer to such incidents as “virtual sexual assault”? Is this something that needs to be regulated and criminalized? Do existing laws on sexual assault and sexual harassment cover incidents of virtual sexual assault, or do we need a new set of criminal offenses to address this issue? These are the questions addressed and answered in this chapter. The chapter is organized into four main parts. In section 1, I propose a definition and classificatory scheme for understanding the phenomenon of virtual sexual assault. I argue that virtual sexual assault is complex due to the different modes of interaction that are possible in a virtual environment, and that there could be as many as six different types of virtual sexual assault that are worthy of our consideration. In section 2, I consider how seriously it is worth taking these six different types of virtual sexual assault, addressing the question of whether criminalization is warranted. I also consider whether existing laws on sexual assault could be interpreted so as to cover cases of virtual sexual assault. I follow this up in section 3 by looking at responsibility for virtual sexual assault. I ask whether it is right and proper for us to deem someone responsible for actions that they perform in a virtual world, particularly if those virtual actions have no obvious real-world effects. Finally, I address a variety of issues that arise in relation to consent (and its absence) in virtual sexual encounters. I do not propose any one single view of virtual sexual assault in this chapter. I think the issue is complex, with some forms of virtual sexual assault worthy of treatment as 7   The feeling of immersion alters the user’s experience of the virtual world. It is no longer separate or distant but phenomenologically real. The extent to which the mind can play these tricks of immersion on us, causing us to believe that we are located in different physical spaces or that our bodies extend into physical artifacts, is documented in the philosopher Thomas Metzinger’s book The Ego Tunnel (New York: Basic Books, 2010). I return to this important point later in the chapter.

BARFIELD PRINT.indd 364

13/11/2018 11:53

The law and ethics of virtual sexual assault  365 serious as any that of real-world form of sexual assault, and other forms worth taking less seriously. But my goal is not to offer the last word on any of the claims I make. It is, instead, to clarify and provoke future research and debate on this important topic. The analysis is intended to be general in scope, but on occasion I do rely on legal examples and case studies drawn from the jurisdictions with whose laws I am most familiar, particularly the law of sexual assault in England and Wales. These examples are provided simply to illustrate and explain the arguments being presented.

1.  CLASSIFYING VIRTUAL SEXUAL ASSAULT Before we can answer some of the interesting ethical and legal questions that arise from the phenomenon of virtual sexual assault, we must get a clearer picture of what that phenomenon entails. The legal philosopher Litska Strikwerda is one of the few people who has addressed this topic at any length,8 although she focuses specifically on virtual “rape.” What I propose here is an expansion and update of her definition and classificatory scheme. I start with the definition. Virtual sexual assault is a species of virtual act. Philip Brey has defined a virtual act as any act that is “initiated by a user within a virtual environment and is defined over objects and persons within that environment.”9 This is a useful starting point. Strikwerda argues that we can combine this definition of a virtual act with the definition of a real-world rape in order to create a satisfactory definition of virtual rape. She does so by adopting a definition of real-world rape as “the act of forcing sex upon an unwilling party,”10 and then defines virtual rape as the virtual act of forcing sex upon an unwilling party where the act is defined over persons in the virtual environment. For her, the critical question then becomes what type of virtual environment is used to perform the act of virtual rape. She distinguishes between two such environments: those that require users to interact using avatars (that is, onscreen textual or visual representations that they manipulate using a controller) or those that facilitate immersive interactions, such as the one described above for the game QuiVR. There are several problems with this suggested definition. First, Brey’s original definition of a virtual act limits its focus to virtual acts that are performed by human users of virtual characters. In other words, it imagines cases where there is a human inputting commands that influence the actions of onscreen characters. While this may be the most interesting and relevant case of a virtual act, it is not the only form that virtual acts can take. Many of the characters in virtual environments may be “wholly virtual” in nature, that is, computer-programmed characters with some degree of artificial intelligence. Since we are living through an era in which artificial intelligence is on the ascendency due to  8   See Strikwerda, L. “Present and Future Instances of Virtual Rape in Light of Three Categories of Legal Philosophical Theories of Rape” (2015) 28(4) Philosophy and Technology 491–510.  9   Brey, P. “The Physical and Social Reality of Virtual Worlds” in Grimshaw, M. (ed) The Oxford Handbook of Virtuality (Oxford: Oxford University Press, 2014), p. 49. This quote was originally sourced through Strikwerda, id. 10   Strikwerda (n 8), p. 493.

BARFIELD PRINT.indd 365

13/11/2018 11:53

366  Research handbook on the law of virtual and augmented reality advances in machine learning algorithms,11 there is every reason to expect that wholly virtual characters will play an increasingly important role in our virtual interactions in the future. This is important because it means that the distinguishing features of virtual sexual assault are not determined solely by the type of virtual environment used for the performance; they are also determined by the type of virtual character performing the act: Is it human controlled or not? Second, the definition of rape used by Strikwerda is both limiting and anachronistic. To describe “rape” as “forced sex” is contentious. Although there are jurisdictions that still insist upon a force requirement when it comes to proving the occurrence of rape, many jurisdictions now favor a definition of rape as nonconsensual sex. At the same time, Strikwerda’s definition of rape does not share the anatomical obsessions of some jurisdictions and so will seem misleading to many. For example, in England and Wales, rape is explicitly defined, in Section 1 of the Sexual Offences Act 2003, as nonconsensual “penetration of the vagina, anus, or mouth” by the penis. Anything that does not involve this penetrative interference is not counted as “rape.” Of course, it makes sense for Strikwerda to drop the anatomical obsessions, since virtual acts will never involve the penetration of one biological orifice by another biological organ—indeed, one could argue that the absence of direct biological contact is the essence of a virtual interaction—but this merely underscores the point I am making: It is anachronistic to focus solely on virtual rape if your interest is in unwanted sexual actions in virtual environments more generally. What I suggest, then, is a broader definition of virtual sexual assault: An unwanted, forced, or nonconsensual sexually explicit behavior performed by virtual characters, to one another, acting through representations in a virtual environment. This is a deliberately vague definition and requires some unpacking. It covers all manner of sexually explicit behaviors, not just those that we would classify as “rape.” That said, the definition allows for subtypes of virtual sexual assault, such as virtual rape. In this regard, I would suggest that virtual rape be defined as a sexually explicit behavior performed by virtual characters through representations in a virtual world that, if performed in the real world, would count as rape. This avoids placing any reliance on specific anatomical representations, which could vary across jurisdictions, but allows for this should it be required in a particular jurisdiction.12 The concept of “representation” is important to this definition of a virtual assault. This importance derives from the observation made previously about the absence of direct biological contact being the essence of the virtual interaction. The suggestion here is that what makes a virtual act different from a real-world act is that instead of the action being performed “flesh on flesh” (so to speak) it is performed through representations, that is, on-screen avatars or the like. There are complications to this, as we shall see below; nevertheless, acting via a representation is, I submit, central to what makes a virtual act “virtual.” 11   Alpaydin, E. Machine Learning: The New AI (Cambridge, MA: MIT Press, 2016); Domingos, P. The Master Algorithm (New York: Basic Books, 2015); Bostrom, N. Superintelligence: Paths, Dangers, Strategies (Oxford: Oxford University Press, 2014). 12   In other words, in the UK a virtual rape would only arise where a virtual character performs an act that represents the penile penetration of the vagina, anus, or mouth of another virtual character.

BARFIELD PRINT.indd 366

13/11/2018 11:53

The law and ethics of virtual sexual assault  367 Finally, the definition also hedges on the normative condition that turns a virtual sexual act from something that is acceptable into something that counts as assault. Although my general preference is that discussions of virtual sexual assault should focus on the idea of “consent,” there may be problems in applying this to all apparent incidents of virtual sexual assault, particularly if the characters involved, due to their being wholly virtual, are deemed incapable of consenting. This is why I include the possibility of unwanted or forced contact. It is important to realize that when it comes to interpreting and applying these normative conditions, representation still plays a key role. Although the virtual acts may involve a real-world human being who does not consent to or want their performance (or upon whom they are “forced”), that is merely a sufficient condition for something counting as an incident of virtual sexual assault. A representation that depicts an act that appears to be nonconsensual, unwanted, or forced will also count. The deliberate vagueness of the definition can be tempered by developing a classificatory scheme that makes sense of the different forms that virtual sexual assault can take. This helps us to draw important ethical and legal distinctions between the possible instances of virtual sexual assault, and allows us to hone in on the ones that raise the most significant concerns. We can start this exercise by distinguishing between virtual sexual assaults on the basis of the virtual characters involved in their performance: do they involve characters controlled by a human user or are they wholly virtual? We can also distinguish between virtual sexual assaults depending on who occupies the role of the victim and the perpetrator. Were human users both the victims and perpetrators? Or did a human user assault a wholly virtual character? Or vice versa? We can easily construct a two-by-two matrix for categorizing the different possibilities here, with one dimension for the human user–virtual character distinction and another for the perpetrator–victim distinction. But before we do so, we must add another layer of complexity to the classificatory scheme. As noted already, Strikwerda argues that we should categorize the different types of virtual rape depending on the mode of virtual interaction they involve. This seems appropriate. Prima facie, there is something different about an interaction that takes place through avatars that are controlled via a mouse or other controller—as is the case in a virtual world such as Second Life—and one that takes place via an immersive device like the HTC Vive as is the case in a virtual world like QuiVR. There is greater “distance” between the real world and the virtual world in the former case and this distance could be ethically and legally significant.13 This is particularly true when you realize that immersive interactions could involve the use of haptic technologies, that is, technologies that allow you to transmit touchlike sensations via a network. Companies have already started to create haptic devices that facilitate longrange virtual sexual interactions. The Dutch company Kiiroo,14 for instance, has a range of “teledildonic” devices that enable people to engage in virtual sexual experiences that replicate touchlike sensation. The users wear immersive helmets, watch their partners or pornographic actors, and use vibrators and artificial vaginas for sexual stimulation. The devices allow for interactivity, whereby one

13   The concept of “distance” and its ethical and legal importance is discussed at greater length below. 14   For information see www.kiiroo.com.

BARFIELD PRINT.indd 367

13/11/2018 11:53

368  Research handbook on the law of virtual and augmented reality character can stimulate another through the haptic device, at a distance. The users still act or are acted upon via representations—the onscreen depictions of their partners or the actors—but the gap between the real and the virtual is narrowed. The actions performed via the virtual representations are immediately transduced into effects on real human bodies: The users can feel completely psychologically immersed in the experience. This seems very different from the text-based interactions via avatars in LambdaMOO or the visual interactions in Second Life. In those cases the user never completely dissolves the gap between themselves and the virtual representations through which they act. So I propose to follow Strikwerda by distinguishing between virtual sexual assaults involving avatar interactions and those using immersive interactions. This makes it more difficult to construct a simple two-by-two matrix for classifying the different kinds of virtual sexual assault, since we are now dealing with three dimensions of difference (the human user–virtual character; perpetrator–victim; and avatar–immersive dimensions). We could solve the problem by constructing a cube to represent the three dimensions, but it is also still possible to work with the two-by-two grid by simply compromising on how you identify one of the dimensions of difference. I have done this in Figure 12.1. This diagram depicts what I call the “logical space” of virtual sexual assault.15 It shows all the possible manifestations of virtual sexual assault (given the three specified

Avatar-interaction

Human Perpetrator

Virtual Perpetrator

immersive-interaction

Virtual victim: human user virtually sexually assaults a wholly virtual agent using an avatar

Virtual victim: Human user virtually sexually assaults a wholly virtual agent using immersive tech.

Human victim: human user virtually sexually assaults an avatar controlled by another human being, using an avatar

Human victim: Human user virtually sexually assaults a virtual agent controlled by another human being, using immersive tech.

Virtual victim: wholly virtual avatar is virtually sexually assaulted by a wholly virtual agent.

Virtual victim - n/a this interaction is not distinct from an interaction between two wholly virtual avatars since the immersive tech only applies to fuse the real-world with the virtual world.

Human victim: avatar controlled by a human being is virtually sexually assaulted by a wholly virtual agent.

Human victim: human user virtually sexually assaulted by a wholly virtual agent controlled by another human being while wearing immersive tech.

Figure 12.1  Classifying virtual sexual assault 15   I adopt the terminology from the work of Christian List. See List, C. “The Logical Space of Democracy” (2011) 39(3) Philosophy and Public Affairs 262­–97 and List, C. and Vallentini, L. “Freedom as Independence” (2016) 126(4) Ethics 1043–4.

BARFIELD PRINT.indd 368

13/11/2018 11:53

The law and ethics of virtual sexual assault  369 ­ imensions). Not every location within this logical space is worthy of serious consideration d or analysis. In particular, I would argue that the interactions that involve wholly virtual characters as both perpetrator and victim do not merit consideration. Why not? I work with the assumption that wholly virtual characters are not, and will not for quite some time, count as moral victims or moral agents. That is to say, I assume that they cannot be harmed by what is done to them (at least, not “harmed” in any morally significant way) and they cannot be morally responsible for what they do. Consequently, wholly virtual incidents of sexual assault are purely representational in nature. They are best analyzed and treated as a subcategory of pornography, not as a type of “assault.” Omitting this type of virtual sexual assault still leaves six potentially significant locations within the logical space. Each of these locations involves either a human perpetrator, human victim, or both. For ease of reference, I will label all six of these locations in the following manner: Human user virtually sexually assaults a wholly virtual character using an avatar Type A  (worthy of consideration because the human perpetrator is a moral agent and so the act may say something about their moral character). Type B Human user virtually sexually assaults a virtual avatar that is controlled by another human being, using their own avatar (worthy of consideration because both the perpetrator and victims are moral agents and moral patients). Human user virtually sexually assaults a wholly virtual character using immerType C  sive technology (worthy of consideration because the human perpetrator is a moral agent and so the act may say something about their moral character). Human user virtually sexually assaults a virtual avatar controlled by another Type D  human (wearing immersive technology) while also using immersive technology (worthy of consideration because both the perpetrator and victims are moral agents and patients). Wholly virtual character virtually sexually assaults a virtual avatar controlled Type E  by a human user (worthy of consideration because the human victim is a moral patient and so can be harmed by the interaction.) Type F Wholly virtual character virtually sexually assaults an avatar controlled by a human user who is using immersive technology (worthy of consideration because the victim is a moral patient and so can be harmed by the interaction). What I propose to consider over the next two sections is exactly how seriously each of the six types ought to be treated. I do that by first considering the harm that might be involved in these cases and then considering responsibility for virtual acts.

2.  SHOULD WE CRIMINALIZE VIRTUAL SEXUAL ASSAULT? Let’s return to our opening examples. When you read about the “rapes” in LambdaMOO and Second Life, or the “groping” in QuiVR, what was your reaction? Did you sense that something wrong was happening? Was it wrong enough to warrant criminalization? Before answering those questions, I would suggest considering a real-world case that shares certain similarities with the cases of virtual sexual assault identified above.

BARFIELD PRINT.indd 369

13/11/2018 11:53

370  Research handbook on the law of virtual and augmented reality The case in question is the English case of R v. Devonald [2008] EWCA Crim 527. The case involved a man named Stephen Devonald who posed on the internet as a 20-year-old woman named “Cassey.” He did so in order to make contact with the victim, a 16-yearold boy who was the former boyfriend of Devonald’s daughter. Using the false persona, Devonald encouraged the victim to masturbate in front of a webcam. The goal, apparently, was to film the victim and post the video online in order to cause him some embarrassment. This was to exact revenge for the way in which the victim had treated Devonald’s daughter. Devonald was convicted under Section 4 of the Sexual Offences Act 2003, which makes it an offense to cause another to engage in sexual activity without their consent. The case is interesting for a number of reasons, not least of which is the novel way in which the court interpreted what it meant to consent to a sexual act. (The court held that the Devonald had deceived the victim as to the nature and purpose of the act in question and thereby vitiated his consent to the act. We will return to issues relating to consent and deception later in this chapter.) For now, I want to highlight three features of the case that are instructive for present purposes. The first is that the case involved a type of virtual interaction between a perpetrator and a victim: They interacted via onscreen representations, with Devonald using the fake avatar of “Cassey” to persuade the victim to engage in masturbation. Second, despite acting via virtual representations, the interaction clearly had real-world consequences and repercussions: the victim’s masturbatory act occurred in the real world, showing how virtual interactions can sometimes spill over into reality. Finally, the circumstances of the case, despite having novel features, already fell under the scope of an existing criminal offense, namely the offense of causing another to engage in sexual activity without consent. There was no need to create a new offense simply because the actions were mediated through technological artifacts. What is more, I doubt anyone felt as though it was wrong to apply that offense to this scenario: It seems perfectly appropriate to do so. This holds some important lessons for the present inquiry. In asking whether the incidents of virtual sexual assault identified above warrant criminalization, we need to be cognizant of the blurry boundary between the virtual and the real, and the fact that existing criminal offenses already cover some virtual actions. We also need to be cognizant of the fact that we already live much of our lives online, and that many of our simple day to day interactions are mediated through virtual environments. Consequently, although it may initially seem odd to many people, the idea of subjecting virtual acts to criminal sanction is neither novel nor exceptional. Granting this, we may still worry that some incidents of virtual sexual assault slip through the gaps of existing laws, and that even if they are covered by existing offenses, these offenses are not serious enough to address the gravity of the harm caused by the relevant incidents. This suggests that instead of simply looking to the different possible forms of virtual sexual assault and how they might match up with the existing law, we should first develop a principled framework for answering questions pertaining to the appropriate criminalization of virtual acts. Over the remainder of this section I will try to develop such a framework and then use it to make two arguments about how we might go about criminalizing certain instances of virtual sexual assault. 2.1  When Should We Criminalize Virtual Acts? I think there is a simple test we can apply to determine the appropriateness of criminalizing virtual acts. The test is this: a virtual act X can be criminalized when it is real and when

BARFIELD PRINT.indd 370

13/11/2018 11:53

The law and ethics of virtual sexual assault  371 its properties and effects satisfy the conditions that generally warrant criminalization. This simple test masks some complexities. Two followup questions must be asked: When are virtual acts (or their effects) real? And what are the conditions of an act that would ordinarily warrant its criminalization? In answer to the first question, we can be guided by the work of the aforementioned Philip Brey. He has a developed a sophisticated framework for understanding the reality of the virtual world. As he points out, there is much confusion about the status of objects and events in virtual environments. Some people think that if you say that an action or object exists virtually you are, effectively, saying that it is not real. This makes the phrase “virtual reality” sound strangely oxymoronic. But this, of course, is incorrect. Virtual entities clearly are real qua virtual entities. To use one of Brey’s examples, a virtual apple (that is, a representation of an apple in a virtual environment) does not exist qua real apple (you cannot bite into it and taste its flesh), but it does exist qua virtual apple. It really does exist within the virtual environment. This is to say that the virtual apple really exists as a simulation or representation. But, and this is the important point, that is all the virtual apple’s existence will ever be: a simulacrum. The same goes for any virtual object or event that represents a real-world object or event that necessarily has physical properties or dimensions. Tables, chairs, rocks, water, and natural gas can only ever be simulated in virtual reality; they cannot exist qua real tables, chairs, rocks, water, and natural gas in the virtual environment. Objects and events that are necessarily physical are easy cases. Other virtual objects and events share a more confusing relationship with reality. Consider something like a clock. I have a virtual clock on my laptop right now. I can read the time off it (it’s 14:35, in case you were wondering). This virtual clock is different from the clock upon my wall: It has no physical or material instantiation (apart from its symbolic representation in the microchips of my computer). But this doesn’t make it any less real than the clock upon my wall. There is no significant ontological difference between the two. This is because the status of clocks is determined solely by their functionality, not by the presence or absence of specific physical properties. Now consider something even more complex, like money. The naive view—one that many have had over the course of history (though this may not be true for future ­generations)—is that money is a physical object: It is some valuable commodity like gold and silver, or it is the notes and coins that we carry in our pockets. This naive view is obviously false. Money is a functional kind, like a clock. Its reality is determined by the function it performs in recording debts and obligations, and in paying for goods and services.16 But money is a different functional kind from a clock. Although the clock’s status is determined by its functionality, the success or failure of that functionality depends on its ability to track a real-world phenomenon: the passage of time. There is no similar real-world phenomenon determining the functionality of money. Money is a social construction. It exists and has the functions it has through an exercise of collective imagination and will. Through collective action we can confer the status of money on

16   For lengthy defences of this conception of money, see Ingham, G. The Nature of Money (London: Polity Press, 2004); Graeber, D. Debt: The First 5000 Years (New York: Melville Press, 2011); Martin, F. Money: The Unauthorised Biography (London: The Bodley Head, 2013).

BARFIELD PRINT.indd 371

13/11/2018 11:53

372  Research handbook on the law of virtual and augmented reality pretty much anything, and nowadays we confer the status of money on digitally recorded bank balances and cryptocurrency ledgers. Indeed, most money that exists today—that one regularly uses to pay for goods and services—is purely virtual in nature. It never has existed, and never will exist, in the form of physical currency. This complicates our understanding of the relationship between the real and the virtual quite a bit. It seems that there are some virtual objects and events that can only ever exist as simulations or representations; however, there are others that, although represented in a virtual environment, are every bit as real as their physical counterparts. Brey argues that we can use John Searle’s theory of social reality to understand what is going on here.17 Searle’s theory holds that there is a distinction between physical facts (that is, physical objects, events, states of affairs) and socially constructed facts. The former are ontologically objective and epistemically objective: They exist in the absence of human minds and we can all come to know of their existence through mutually agreed upon methods of inquiry. The apple from earlier on would be a classic example. Socially constructed facts are different. They are ontologically subjective and epistemically objective. They only exist in the presence of human minds (because they are created through acts of collective will) but this does not mean that we can simply imagine into existence whatever social objects we like. There has to be collective agreement on their status and this gives social objects  much of the intersubjective persistence and stability that we associate with “real” objects. Hence they are epistemically objective. Money really exists and we all agree that it exists in certain amounts and distributions. My bank balance is an epistemically objective fact: I cannot imagine extra money into my account, no matter how hard I try. Searle argues that this happens because we confer the status of money on certain representations through the collective creation of a constitutive rule, that is, a rule of the form “X counts as Y in context C.” A significant number of the objects, events, and activities with which we engage every day are social in nature and are made possible through collective constitutive rules. Getting married, being promoted, passing an exam, receiving an award, and being qualified/credentialed are all examples of this. The upshot of this is that we have a principled way of determining whether a virtual object or event is real or not. It is real, according to Brey, if it is a purely functional object or it falls within the definition of Searle’s social facts; it is merely a simulation if it falls within the definition of Searle’s physical facts. There is one additional wrinkle to this. Brey notes that virtual acts require a somewhat different analysis. Whenever a virtual action involves at least one human participant— and, as noted above, we are only interested in these cases—it can have intravirtual and extravirtual effects.18 The intravirtual effects are those that only affect the virtual representations. The extravirtual effects are those that affect the human beings who control or participate in the virtual actions. The intravirtual effects can only be real if the 17   See Brey (n 9) and Brey, P. “The Social Ontology of Virtual Environments” (2003) 62(1) American Journal of Economics and Sociology 269–82. See also Searle, J. The Construction of Social Reality (London: Penguin, 1996); Searle, J. Making the Social World (Oxford: Oxford University Press, 2010). 18   The terminology comes from the work of Johnny Soraker and his The Value of Virtual Worlds and Entities: A Philosophical Analysis of Virtual Worlds and Their Impact on Well-Being (Inskamp: Enschede 2010). It is referenced in both Brey (n 9) and Strikwerda (n 8).

BARFIELD PRINT.indd 372

13/11/2018 11:53

The law and ethics of virtual sexual assault  373 virtual representations fall within the scope of Searle’s analysis. The extravirtual effects are always, uncontroversially, real, either because they have physical and mental effects on the human user or because they have effects on the extravirtual social reality in which the user operates. So much for the nature of virtual reality. We now must confront the second subquestion. I said earlier that virtual sexual assault might warrant criminalization if it is real and if it meets the conditions that ordinarily warrant criminalization. We have some sense of the former now, but what about the latter? What are the conditions that warrant criminalization? Without seeking to defend my views at length, I here adopt a modified version of a position known as “legal moralism.” This might seem like an inauspicious starting point. Legal moralism is often associated with historically regressive attitudes toward the criminalization of private conduct, often sexual in nature. For instance, the most significant debate about legal moralism in the twentieth century was the debate between the legal theorist H.L.A. Hart and the judge Patrick Devlin about the criminalization of homosexuality.19 In that debate, Devlin defended what has been referred to as the moralistic view, according to which the criminalization of homosexuality was legitimate. Hart defended a narrower, liberal view, according to which it was not. He held that harm to others was an essential condition for the criminalization of any act. Since h ­ omosexuality did not cause harm to others, it should not be criminalized. If legal moralism is to be understood along the lines set by the Hart–Devlin debate, then I quite agree that it would be an inauspicious place from which to begin. If it were, I would be bound to the notion that actions which cause disgust or give offense warrant criminalization, and this is not a position I hold. Fortunately, it does not have to be understood along these lines. Indeed, if anything, the Hart–Devlin debate is misleading as to the true nature of legal moralism. Devlin’s view was really one of legal populism, not moralism, since it was about ensuring social cohesiveness through the law; and Hart’s view was itself a species of legal moralism, since he presumably believed that causing harm to others was immoral. My understanding of legal moralism is much narrower.20 It is as follows: A necessary condition for the criminalization of any act is that it be immoral in some way. This is quite a modest view and, I would argue, a sensible one. Immorality comes in many flavors. If we follow the work of social psychologists such as Jonathan Haidt it could come in as many as six different flavors, including loyalty to authority, fairness, harm, disgust, freedom, and sacredness.21 Whatever the case may be, the essence of legal moralism is that an act must be some flavor of immoral in order for it to warrant criminalization. Legal moralism, as defined, only identifies a necessary condition for criminalization, not a sufficient one. Identifying a sufficient condition is much trickier and, indeed, there are likely to be many sufficient conditions. Few if any defenders of legal moralism hold that all immoral conduct warrants criminalization. The more common view is that certain subtypes of immoral conduct warrant criminalization. Following Anthony Duff, I hold 19   Devlin, P. The Enforcement of Morals (Oxford: Oxford University Press 1965); and Hart, H.L.A. Law, Liberty and Morality (Stanford, CA: Stanford University Press, 1963). 20   For details of this understanding of legal moralism, see Simester, A.P. and Von Hirsch, A. Crimes, Harms and Wrongs (Oxford: Hart Publishing, 2011), pp. 22–30. 21   Haidt, J. The Righteous Mind (London: Penguin 2013).

BARFIELD PRINT.indd 373

13/11/2018 11:53

374  Research handbook on the law of virtual and augmented reality that one feature of these subtypes is that they must implicate a public wrong, that is, an immoral act that demands public attention, scrutiny, and accountability.22 Lying to your friends about where you were last night might be immoral, but it would not be a public wrong. Lying on your tax return is a different matter. Furthermore, and following the classic liberal view, I hold that the most suitable candidate immoral actions are those that cause harm to others, provided that the harm crosses a certain threshold of seriousness and is a matter of public concern. I would also allow for the possibility of some immoral actions that warrant criminalization and yet do not cause harm to others. So, for example, desecrating the bodies of the dead or destroying natural objects might warrant criminalization, even if no one is harmed in these actions.23 Of course, the most contentious issue is whether actions that cause harm to the self warrant criminalization. My own personal view is that they do not, though some actions may cause harm to one’s moral character, which could have downstream effects on others and so could, in certain circumstances, warrant criminalization. Where does that leave us in relation to virtual sexual assault? Recall that the position being defended here is that virtual sexual assault warrants criminalization if it is real and if it meets the conditions that ordinarily warrant criminalization. So we need to ask ourselves: Are virtual sexual assaults real and do they cause harm or implicate public wrongs? The answer to the latter seems relatively straightforward—they obviously can cause harm to others, since users can be psychologically traumatized or even physically interfered with, depending on the mode of virtual interaction—but the answer to the former is more contentious. There are effectively three positions which can be taken on the nature of sexual assault:



Sexual assaults are essentially physical: An action only deserves to be called “sexual” if it involves sexually motivated or stimulated physical touching. This would mean that virtual sexual assaults are not real sexual assaults; they are mere simulations of sexual assault. Sexual assaults are essentially social: An action acquires the status of being “sexual” through collective will and intention. This would mean that virtual sexual assaults could be every bit as real as real-world sexual assaults. Sexual assaults can be both physical and social: Some sexual acts are essentially physical in nature (for example, sexual intercourse); others are perhaps largely social in nature (for example, sexual harassment or obscene communications); and many include a mix of both physical and social properties.

The first position tallies with the law in many countries. As mentioned previously, under the Sexual Offences Act of 2003 in England and Wales, rape is defined as the intentional penile penetration of the vagina, anus, or mouth; the related offense of assault by penetration is defined as the penetration of the vagina or anus with something other than the

22   Duff, A. “Towards a Modest Legal Moralism” (2014) 8 Criminal Law and Philosophy 217–35. 23   On these examples, see Scoccia, D. “In Defense of ‘Pure’ Legal Moralism” (2013) 7 Criminal Law and Philosophy 513–30.

BARFIELD PRINT.indd 374

13/11/2018 11:53

The law and ethics of virtual sexual assault  375 penis. Both offenses require that certain physical conditions be met: Particular biological orifices must be penetrated by particular objects. The same goes for the more general offense of sexual assault. This is defined under Section 3 of the 2003 Act as intentional sexual touching, which again seems to require some physical contact. If the first position was adopted, then virtual sexual assault could only be said to occur when a human user was assaulted by another virtual agent (human controlled or not) while wearing a haptic suit that enabled them to satisfy the relevant physical conditions (such as penetration or touching). This does not mean that the mere simulations of sexual assault could not be criminalized—if it causes extravirtual harm to others, or to the individual performing it, that are tantamount to public wrongs, then it might be properly criminalized—but it does mean that it could not be criminalized qua sexual assault. It would have to be criminalized as something else. The second position is more at home with certain traditional and contemporary feminist understandings of sexual assault. As Strikwerda points out,24 sexual offenses such as rape were traditionally understood as property-related crimes: the crime being sexual interference with the property of a husband or father. Property is a social construct: An object, person, or piece of land is only your property if society confers that status upon it through collective will. One could view a virtual avatar as a type of property and hence virtual sexual assault as a real form of sexual assault. But even if you do conceive of sexual assault in these property-related terms, you would still be tasked with distinguishing sexual interferences with property rights from nonsexual ones. For example, what, traditionally, distinguished trespass from rape? Most countries answered this by appealing to certain essentially physical characteristics of sexual acts, which suggests that this traditional, conservative view is, in practice, similar to the first position when it comes to viewing sexual acts as essentially physical. It would need to be updated in order to cover virtual sexual assault. A more promising way to defend the second position is to use certain ideas and concepts drawn from feminist theory.25 It is, of course, a mistake to view feminism as a single coherent theoretical framework. There are many feminisms and they sometimes contradict. Nevertheless, there is a prominent strain of feminist theory, with adherents in legal theory, that argues that many of the characteristics and concepts we hold to be “sexual” are social constructs. Why is it sexually provocative and immodest for a woman to bare her arms in some cultures and not in others? The answer lies in socially constructed norms of sexuality. This constructionist view also carries over into how certain feminists think we should understand sexual assault. One suggestion is that sexual assault is best understood as a nonconsensual interference with someone’s sexual agency, where agency is understood as a social construct.26 This is not an uncommon view, with several feminist scholars arguing that sexual agency is a construct that emerges from competing discourses, prejudices, and beliefs in society about what an individual can and cannot permissibly do. The typical view is that female sexual agency is undermined and limited by   Strikwerda (n 8).   Strikwerda (n 8) also examines this line of argument. 26   Abrams, K. “From Autonomy to Agency: Feminist Perspectives on Self-Direction” (1998) 40 William and Mary Law Review 805–46; Turkheimer, D. “Rape On and Off Campus” (2015) 65 Emory Law Journal 101–46. 24 25

BARFIELD PRINT.indd 375

13/11/2018 11:53

376  Research handbook on the law of virtual and augmented reality patriarchal ideologies—that we live in a “rape culture” that normalizes interferences with female sexual agency and constantly polices its boundaries. If you adopt this view, then there is no reason why virtual sexual assault couldn’t be every bit as serious an interference with sexual agency as real-world sexual assault. The third position is probably closest to my own. This holds that sexual acts can have different mixes of physical and social properties. Some sexual offenses are properly viewed as essentially physical—penetrative sexual assault and rape are the most obvious cases. Some are essentially social in nature—sexual harassment being a possible example since, although it is usually defined in terms of unwanted or inappropriate sexual advances, the definition and determination of what counts as an unwanted or inappropriate sexual advance is largely a product of social convention. And, finally, some are just mixes of physical and social properties—simple sexual assault is arguably of this form because, although typically defined as “sexual touching,” the definition and determination of what counts as “sexual” touching is again a matter of social convention. This third position is easily reconciled with the idea that there is a range of sexual offenses, differentiated not only in terms of the seriousness of the harm they cause but also in terms of their metaphysical constitution (physical, social, or a mix of both). According to this view, some forms of virtual sexual assault could be captured by existing offenses and others might require the creation of new offenses, which are still properly classified as being “sexual” in nature. 2.2  How Should We Go about Criminalizing Virtual Sexual Assault? To this point, I have tried to outline a framework for thinking about virtual sexual assault and its possible criminalization. Is there anything more concrete and prescriptive to be said? How exactly should we treat the six types of virtual sexual assault identified in the previous section? I have two arguments to make in relation to this. First, I think it is probably a mistake to criminalize virtual sexual assault of types A and C. These are cases where a human user sexually assaults a wholly virtual character, either through an avatar or using immersive technology. In neither case is there a human victim who could be harmed, physically or mentally, by the action. The only person being harmed is the perpetrator themselves, perhaps by corrupting their moral character. As such, these instances of virtual sexual assault would be equivalent to classic vice crimes, where the criminalization is justified by its effect on the perpetrator’s moral virtue. My reason for being skeptical about the wisdom of criminalizing this conduct is not because I think the conduct in question is perfectly morally acceptable. On the contrary, I think someone who revels in committing acts of virtual sexual assault is doing something that is contrary to moral norms. But I think criminalization is not an effective or appropriate strategy for addressing this kind of immoral conduct, for much the same reasons why I do not think that throwing drug addicts in prison is an effective strategy for addressing their problems. What might be more appropriate would be to criminalize the makers or manufacturers of virtual reality technology for creating worlds that facilitate or encourage these types of virtual sexual assault, but in many instances this would be an overreach. Prohibition might conflict with freedom of expression and it might be unfair to the manufacturer if the user was the one who took advantage of the relative freedom/autonomy they were provided in the virtual environment to commit the relevant acts. In other words,

BARFIELD PRINT.indd 376

13/11/2018 11:53

The law and ethics of virtual sexual assault  377 unless the virtual environment was explicitly designed to facilitate fantasies of rape and sexual assault, and its doing so is completely gratuitous and devoid of any artistic or public merit, criminalization of makers and manufacturers would also seem a step too far.27 Regulatory interventions might be more appropriate, with the design of industry safeguards and standards being an obvious first step toward addressing problems that may arise. This is true not only in relation to cases involving wholly virtual victims, but regarding all other cases too. Second, I think that virtual sexual assault cases involving human victims (that is, Types B, D, E, and F) could be justifiably criminalized, but the approach toward each type could vary. Some could be adequately captured by existing laws, others might require the creation of new laws. The reason for being more confident about the criminalization of cases involving human victims is because they would seem to fall under the classic “harm to others” rationale for criminalization. Furthermore, we can be reasonably confident that people are harmed by virtual sexual assault. The harm would be primarily psychological as opposed to physical in nature,28 but this is no different from real-world forms of sexual assault. Legal provisions on sexual assault do not typically require physical harm. Furthermore, when the assault is carried out on a victim wearing haptic technology, there would be physical effects on the body, making this type of virtual sexual assault almost indistinguishable from real-world sexual assault. This also suggests that forms of virtual sexual assault that are mediated through the use of haptic technologies would already be dealt with by existing laws on sexual assault. In the absence of haptic technologies, things become a little bit more difficult. Sexually assaulting a person’s onscreen avatar might be covered by sexual harassment laws or obscene communication laws. Sexually explicit text messages, emails, and videos that are shared and communicated online can, for example, already form the basis of a sexual harassment claim, so it is no great imaginative leap to suggest that sexually assaulting someone’s avatar could do the same. The problem is that treating it as a type of harassment or obscene communication may not do justice to the seriousness of the offense. Sexual harassment is usually required to have taken place in the workplace, and so a good deal of virtual sexual assault might be missed unless the legal definition of harassment was expanded. Furthermore, you could argue that there is an important qualitative moral difference between making sexually explicit advances/communications via a virtual medium and actually sexually touching or penetrating someone’s onscreen persona. If so, then we may need to create a new class of offense—virtual sexual assault—that does not require physical contact before it can be said to have taken place. Alternatively, we could take existing legal definitions of sexual assault and broaden them so that they no longer have a physical contact element to them.

27   The computer game Rapelay might be a good example of a game that does go too far in this respect and is deservedly banned. But note that this raises a question that is widely discussed in the ethics of computer games, namely: why do we seem to have such different attitudes toward sexual violence in video games, as opposed to nonsexual violence? See Luck, M. “The Gamer’s Dilemma: An Analysis of the Arguments for the Distinction between Virtual Murder and Virtual Paedophilia” (2009) 11 Ethics and Information Technology 31–6; Luck, M. and Ellerby, N. “Has Bartel Resolved the Gamer’s Dilemma?” (2013) 15 Ethics and Information Technology 229–33. 28   Distinguishing between the physical and psychological is philosophically problematic, but we can set aside debates about mind–body dualism for now.

BARFIELD PRINT.indd 377

13/11/2018 11:53

378  Research handbook on the law of virtual and augmented reality

3. CRIMINAL RESPONSIBILITY FOR VIRTUAL SEXUAL ASSAULT The previous section considered whether virtual sexual assault ought to be criminalized. Students of criminal law theory will have noted that in attempting to address this issue I focused on the so-called actus reus elements of an offense, that is, what makes for a criminal act. I did not consider what it would take for someone to be justifiably held criminally responsible for those actions, the so-called mens rea or mental side of things. It is best to treat these as analytically distinct. The justification for criminalization is grounded largely in the external properties of an action (that is, its objective character and its effects on the world, including the minds of others). Nevertheless, there is little point in criminalizing an action if you cannot justifiably hold someone responsible for it. Is this possible in the case of virtual sexual assault? To answer that, it is worth distinguishing between types of virtual sexual assault in which a human acted as the perpetrator and cases in which it was a wholly virtual agent that perpetrated the act. 3.1  Responsibility When There Is a Human Perpetrator In cases involving human perpetrators, responsibility might seem like a nonissue. Humans have minds. They can act knowingly, intentionally, and recklessly, through their virtual representations, and hence they can satisfy the conditions of blameworthiness that are typically applied in the law. If they intentionally or recklessly perpetrated a virtual sexual assault, it would be right and proper to hold them responsible for doing so. But we have to be careful with this. The responsibility for virtual acts is more complex than responsibility for real-world acts. There are two reasons for this. The first is that the virtual environment itself might be structured and constrained by other agents (that is, it might limit your possibilities for action), and this might call into question how free or autonomous the virtual action really was. The second is that people often distance themselves from their virtual personas. This psychological distancing might call into question their responsibility for what happens in the virtual world. Geert Gooskens has addressed both of these issues in his work,29 and I want to assess his arguments in what follows. In relation to the first point, Gooskens argues that some freedom of action in a virtual environment is necessary if we are going to ascribe responsibility and blame to the actions of people in that environment. He points out that the virtual environments associated with traditional videogames don’t always afford their players a great degree of freedom. Indeed, classically, games only allowed players to exercise a very narrow form of instrumental rationality. The games would have clearly defined goals or objectives, and a limited set of pathways (maybe even only one pathway) to achieving those goals or objectives. Players had to follow those pathways in order to journey further in the virtual world. You could argue that this does not allow for the necessary kind of freedom and so it would be wrong to ascribe responsibility for actions in that environment. This has repercussions for how we view responsibility for virtual sexual assault. For instance,

29   Gooskens, G. “The Ethical Status of Virtual Actions” (2010) 17(1) Ethical Perspectives 59–78.

BARFIELD PRINT.indd 378

13/11/2018 11:53

The law and ethics of virtual sexual assault  379 in the controversial early 1980s videogame Custer’s Revenge, players had to commit a racially motivated sexual assault in order to succeed in the game. The game had one clearly defined ­objective—rape a Native American woman tied to a stake—and to succeed in the game you had to manipulate your character to avoid attacks and reach the woman. As such, this game only allowed the player to exercise a restricted form of instrumental rationality. Applying Gooskens’ analysis, one could argue that this would not be sufficient to ascribe responsibility to the player for that virtual sexual assault. This example involves a virtual environment in which the “victim” of the virtual assault is not ultimately controlled by a human, but presumably the same reasoning would apply if the victim was ultimately human: If the virtual environment allows for a restricted form of instrumental ­rationality—for example, if you have to sexually assault the other player’s avatar to progress in the virtual world—then you cannot rightly be held responsible for it. The problem with this analysis, however, is that it ignores the player’s decision to play a game with such a clearly defined goal in the first place. Surely freedom over that choice would meet the necessary condition for responsibility? If you sign up to a virtual world with such limited conditions of engagement, surely you can rightly be held responsible for following through on them? This seems right to me, but then it raises the question of whether there could ever be a sexual assault in such an environment, since presumably the other human player would have signed up to those conditions of engagement too, which might suggest that they consented to the outcome. I discuss this issue in more detail in the final section. In any event, much of this argument is moot since more recent video games and virtual environments allow for a much more robust type of ethical rationality to be exercised by the players. Increasingly, there are multiple ways of achieving a given game objective. For example, in a game such as those in the Grand Theft Auto series you don’t usually have to mow down pedestrians or kill sex workers in order to complete the missions. You exercise freedom of choice to perform those actions. Furthermore, the kinds of virtual worlds discussed at the outset of this chapter, and the kinds that are foremost in people’s minds when they imagine “virtual reality,” are openended environments, ones in which you choose your own objectives. These environments would seem to meet the necessary condition of freedom that is needed for responsibility. In relation to the second issue—that of psychological distance—Gooskens makes use of certain ideas from phenomenology, particularly the ideas of Edmund Husserl, to argue against responsibility for virtual actions. Phenomenology is a branch of philosophy that deals with the contours of subjective experience. Husserl was one of the leading phenomenologists and he developed a particular phenomenology of the “image world”, that is, the world of representations, both real and fictional. These representations would include photographs or paintings of real or fictional events and people, and also videogame representations involving real or fictional agents and worlds. Gooskens argues that acting in a virtual world is a kind of image consciousness (roughly: an intention to do things through the medium of an image). This is consistent with my earlier characterization of virtual action as action-via-a-representation. Image consciousness, according to Husserl, is characterized by the presence of an “as-if ” modifier, that is, the situation is such that you are actually perceiving or intending something of type X, but it is “as-if ” you are perceiving or intending something of type Y. Consider the following example:

BARFIELD PRINT.indd 379

13/11/2018 11:53

380  Research handbook on the law of virtual and augmented reality Grandfather’s photograph: I am holding in my hand a small black-and-white photograph of my grandfather and looking directly at it. What am I perceiving? I am actually perceiving a grey two-dimensional figure on a piece of card; but it is as-if I am perceiving my grandfather.

That example should be pretty straightforward. The point is that the things that are perceived or intended through image consciousness are “beings-as-if,” that is, they are not really there, but only virtually there. The key Husserlian move is to argue that image consciousness not only covers the objective correlates of our perceptions and intentions—that is, the things we are looking at and the actions we are performing—but also the mental states underlying those experiences. In other words, when looking at my grandfather’s photograph, it is not just that the photograph is only an “as-if ” representation of my grandfather, it is also that my subjective state of perception is an “as-if ” state of perception. Husserl illustrates the extreme implications of this view with the example of a picture of human suffering. He argues that we cannot really feel real pity by viewing this photograph, but only “as-if ” pity. Gooskens argues that the same analysis can be applied to the mental conditions of responsibility for actions in a virtual world. In other words, he holds if you intend to commit a virtual rape or sexual assault through a virtual representation, it is not the case that you actually intend to do this, it is only “as-if ” you intend to do so. Consequently, we cannot really ascribe responsibility or blame for actions in a virtual environment. This strikes me as being a deeply implausible view, particularly when you take into account the complex relationship between the real and the virtual outlined earlier in this chapter. If it is the case that some virtual actions and events are every bit as real as their real-world equivalents, it is difficult to see how the as-if modifier can be rightly applied to their occurrence. If I take bitcoin from someone’s online wallet, it seems fair to say that I have really stolen from them. It is not just “as-if ” I have stolen from them. There is also ample evidence from psychology and neuroscience to suggest that Husserl’s understanding of image consciousness is wrong. People’s sense of ownership over their bodies can be, and often is, projected out into the world of representations and artifacts. The famous “Rubber Hand” illusion, for example, shows that people can incorporate a fake rubber hand into their phenomenological representation of their own body.30 And as the philosopher Thomas Metzinger observes in his review of the evidence: “Owning” your body, its sensations, and its various parts is fundamental to the feeling of being someone. Your body image is surprisingly flexible. Expert skiers, for example, can extend their consciously experienced body image into the tips of their skis. Race-car drivers can expand it to include boundaries of the car; they do not have to judge visually whether they can squeeze through a narrow opening or avoid an obstacle—they simply feel it.31

All of this suggests that the “as-if ” modifier propounded by Husserl may not be present. If so, any attempt to deny responsibility for a virtual sexual assault by making the Husserlian move is likely to fall flat. 30   Botvinick, M. and Cohen, J. “Rubber Hand ‘Feels’ Touch that Eyes See” (1998) 391 Nature 756. 31   Metzinger (n 7) 75.

BARFIELD PRINT.indd 380

13/11/2018 11:53

The law and ethics of virtual sexual assault  381 Interestingly, Gooskens seems to be aware of the troubled relationship between the virtual and the real when it comes to the moral assessment of at least some kinds of virtual acts. He argues we can rightly feel moral discomfort at certain kinds of virtual acts, but we cannot properly assign moral blame for their performance. In broad outline, his argument is very straightforward. He argues that we can rightly feel moral discomfort whenever a performer of a virtually immoral act blurs the boundary between their real-self and imageworld-self: In other words, when the “as-if ” modifier begins to lose its grip. Gooskens uses some nice examples to flesh this out. First, he takes us away from the hi-tech world of video games and back to the more genteel and low-tech world of the stage play. Actors are the quintessential performers of virtual acts and usually we have no problem with them performing acts that would otherwise be highly immoral. Thus, we do not balk at Anthony Hopkins playing the sadistic cannibal Hannibal Lecter. But there is some complexity to this. Usually, upon learning that an immoral act is part of a play or performance, we feel a great sense of relief. To use Gooskens’ example, imagine walking into your flat to find your actor roommate apparently forcing himself on a woman, but then seeing the director of the play issuing various directions from the couch. In this case, what would initially be a feeling of grave moral concern would shift to one of relief. Can the emotions ever run in the other direction, that is, from a feeling of relief (or ease) to a feeling of discomfort? Gooskens argues that it can. Imagine if you are watching the stage performance of your roommate’s play. Everyone is captivated by his intense and realistic performance during the rape scene. Afterwards, you congratulate him on this and he tells you that the reason why he was so intense and realistic is that he was actually sexually aroused during the scene. Gooskens suggests that in this case you would feel moral discomfort. Your roommate appears to have blurred the boundary between his real-self and his imageworld-self in a most disturbing manner. If we accept this analysis, however, don’t we end up with the view that it is often going to be right and proper to hold people responsible for virtual sexual assault? After all, the distinction between the real-self and the imageworld-self is an increasingly fuzzy one in most people’s minds. In many instances, I would argue, it simply collapses. I don’t feel separation from the online version of myself and any attempt to disclaim responsibility for my online actions would ring hollow to my peers. Gooskens’ claim is that we can only feel moral discomfort because the “as-if ” modifier prevents us from thinking that anything intrinsically wrong has occurred in virtuality. But this conclusion seems untenable, at least in the cases of virtual sexual assault in which we are interested, because they have more than a whiff of reality about them. As Gooskens himself concedes, if an action has some real-world consequences it may still count as being intrinsically wrong. This would seem to be true of virtual sexual assault in which there is a human victim. In sum, I would argue that the ascription of responsibility for human perpetrators poses no significant problems. 3.2  Responsibility for Virtual Perpetrators What about cases in which the perpetrator of the sexual assault is wholly virtual? In the cases in which we are interested, these will involve a human victim, so there is the possibility of genuine harm and hence a justification for criminalization, but the issues surrounding responsibility are more complex. If the wholly virtual agent committing the

BARFIELD PRINT.indd 381

13/11/2018 11:53

382  Research handbook on the law of virtual and augmented reality assault is specifically programmed by a human to perform such an act, then it will be easy to trace a line of responsibility back to that human programmer. If the wholly virtual agent has some degree of autonomy or artificial intelligence, if it is programmed by an organization or team of humans, and if its behavior is an unanticipated or unexpected outcome of its autonomy and intelligence, the situation might be a little more difficult. In that case, we may have a “responsibility gap” opening up between the acts of the virtual agent and the decisions made by its programmers and creators. This responsibility gap problem has been widely discussed in the debate about autonomous robots. Indeed, the possibility of such a gap is probably the leading objection to the creation of robotic weapons systems.32 Many legal and ethical scholars think that the use of such systems would undermine accountability and responsibility for acts of war. Similar discussions have also begun in relation to responsibility for the misdeeds of autonomous cars.33 I would argue that when it comes to determining the responsibility for acts of virtual sexual assault committed by wholly virtual agents, we will end up following and learning from this debate. Although I cannot do justice to all the positions defended in the existing responsibility gap literature, I can at least give a sense of the solutions proposed to the problem. Many of them point out that the gap can be plugged by existing legal doctrines and concepts. It is not as though the law has never had to deal with problems arising from corporate agency or nonhuman agency before. There are already laws that allow for corporations to be held criminally responsible for certain actions, as well as laws that impose duties on humans to control nonhuman agents (such as animals). Recklessness and negligence are well-recognized standards of criminal responsibility for offenses as serious as manslaughter. The recklessness standard states that someone can be held responsible for an act if they recognized the risk of its occurrence but took the risk anyway. The negligence standard states that someone can be held responsible for an act if they ought to have recognized the risk of its occurrence. Neha Jain has recently argued that both standards could be applied to the creators and/or deployers of autonomous weapons systems if and when those systems do something illegal.34 A similar argument could be made in relation to creators and programmers of virtual agents who commit virtual sexual assault. As a last resort, it might also be possible to apply a strict liability standard to the creators and manufacturers, thereby obviating the need to prove any foresight or lack of care on their part. That said, we should be careful here. The use of strict liability standards in criminal law is highly controversial, and usually reserved for so-called regulatory offenses. That may not be much of an issue, however, since the perpetration of sexual assault by a wholly virtual agent would probably require the creation of an entirely

32   For the classic discussion of this problem, see Sparrow, R. “Killer Robots” (2007) 24(1) Journal of Applied Philosophy 62–77; for a lengthy analysis of the issues, see the contributions on responsibility in Bhuta, N., Beck, S., Geiß, R., Liu, H.-Y., and Kreß, C. (eds) Autonomous Weapons Systems: Law, Ethics, Policy (Cambridge: Cambridge University Press, 2016). 33   I discuss some of these issues in Danaher, J. “Robots, Law and the Retribution Gap” (2016) 18(4) Ethics and Information Technology 299–309; see also Nyholm, S. “Attributing Agency to Automated Systems: Reflections on Human–Robot Collaborations and Responsibility-Loci” (2017) Science and Engineering Ethics DOI: 10.1007/s11948-017-9943-x. 34   Jain, N. “Autonomous Weapons Systems: New Frameworks for Individual Responsibility” in Bhuta et al (eds) Autonomous Weapons Systems (n 32).

BARFIELD PRINT.indd 382

13/11/2018 11:53

The law and ethics of virtual sexual assault  383 new category of offense. And, furthermore, it may be the case that civil law is a better remedy to the kind of harm that is caused by such an act.

4.  CONSENT AND VIRTUAL SEXUAL ASSAULT It is worth closing by reflecting on consent and its importance in our approach to virtual sexual assault. Consent is, according to many, a kind of “moral magic.”35 It is what converts morally impermissible acts into permissible ones. This is particularly true in relation to sexual offenses. In many instances, the absence of consent is baked into the definition of sexual assault and rape. It is what differentiates acceptable sexual activity from immoral sexual activity. But what does it mean to consent to sexual activity? This is a topic that could fill many books.36 In broad outline, consent is a subjective attitude toward the performance of an act (roughly, an agreement that the act be performed, which is not necessarily equivalent to a desire that it be performed) that is communicated through objective behavior. So you consent to sexual activity when you subjectively agree to its performance and you communicate that subjective agreement through your behavior. When it comes to determining the law and ethics of sexual consent, the communication of consent is all-important. We can never know what a person is really thinking. We can only go by their outward behavior. What objective behavior counts as an effective communication of consent? Must you clearly and unambiguously state your agreement to the sexual activity? Or will a raised eyebrow and alluring glance be enough? Many of the historical problems with sexual assault have centered on the tendency to infer consent from outward behaviors that are highly ambiguous and frequently misinterpreted. And much of the modern debate about sexual assault is concerned with the desirability of moving away from such misinterpretations to affirmative consent standards, that is, a legal norm of consent that requires clear and unambiguous signals of assent to sexual relations.37 Given the centrality of consent to our modern conception of sexual assault, it stands to reason that consent will take a similarly central location in our understanding of virtual sexual assault. People already engage in sexual behaviors in virtual environments and will continue to do so in the future. And, just as is the case in the real world, developing a set of norms around which objective behaviors and signals “count” when it comes to consent to virtual sexual activity will be crucial when it comes to distinguishing the permissible forms of virtual sexual activity from the impermissible forms. For the most part, the debates about real-world consent and virtual-world consent will and should follow similar tracks. It is possible for people to communicate their agreement to sexual activity through their virtual characters and avatars, either by text or speech, and   I take this from Hurd, H. “The Moral Magic of Consent” (1996) 2(2) Legal Theory 121–46.   See, for example, Wertheimer, A. Consent to Sexual Relations (Cambridge: Cambridge University Press, 2003); Westen, P. The Logic of Consent (London: Ashgate, 2004). 37   Little, N. “From No Means No to Only Yes Means Yes: The Rational Results of an Affirmative Consent Standard in Rape Law” (2005) 58(4) Vanderbilt Law Review 1321–64; Dougherty, T. “Yes Means Yes: Communication as Consent” (2015) 43(3) Philosophy and Public Affairs 224–53. 35 36

BARFIELD PRINT.indd 383

13/11/2018 11:53

384  Research handbook on the law of virtual and augmented reality it is desirable to create a set of norms that encourages them to do so in clear and unambiguous terms (if they wish to engage in virtual sexual activity). In this respect, I would argue that if we are encouraging a shift toward affirmative consent standards in the real world (and it seems as if we are), we should encourage the same shift in the virtual world. Nevertheless, despite my enthusiasm for treating the two domains in the same way, there may be ways in which consent-relevant communications between sexual partners in virtual worlds could be more problematic than consent-relevant communications in the real world. I want to close by discussing two such problems. The first is the problem of deception. Many legal systems accept that certain forms of deception undermine consent to sexual relations, and there are ethicists who argue that nearly all forms of deception should.38 In the legal system with which I am most ­familiar—that of England and Wales—deception as to the “nature and quality of the act” and deception as to the identity of the perpetrator (if this involves impersonation of someone known to the victim) are so-called irrebuttable presumptions of nonconsent in cases of sexual assault.39 In other words, if it can be shown that a defendant deceived a victim as to the sexual nature of the activity in which they engaged, or if they can be shown to have impersonated someone known to the victim, it is impossible for them to argue that they had a reasonable belief in the consent of the victim. Other forms of deception may also undermine consent, but do not provide an irrebuttable presumption of nonconsent. The two forms of deception that provide irrebuttable presumptions obviously do occur in the real world. Deception as to the nature and quality of a sexual act typically arises where some doctor or medical official deceives someone into undergoing an examination that provides the perpetrator with sexual gratification but is presented to the victim under the pretense of medical necessity.40 It can arise in other contexts too. The case of R v. Devonald (discussed previously) was successfully prosecuted because the court held that there was deception as the nature of the act in question: the victim thought his masturbation was for mutual gratification when it was actually for the purposes of embarrassment and revenge. Impersonation of someone known to the victim is relatively rare in the real world, and the reported cases usually involve someone climbing into the bed of a victim when it is dark and difficult to see who is really there. Nevertheless, it does happen.41 I mention these examples because one concern we may have about virtual sexual activity is that it is more susceptible to consent-undermining forms of deception than real-world forms. Since people are not interacting “flesh on flesh,” but instead through virtual representations, it could be much easier, for example, for one party to impersonate someone known to a victim. The usual techniques of identity-checking that apply in the real world are more easily circumvented in virtual environments. The account associated with your virtual avatar could be hacked or you might carelessly share the relevant information or login details with someone else. If the poorly named practice of “fraping” is anything to go by, impersonating another may end up being relatively common. Similarly, deception as to the nature and quality of a virtual sexual act might be more common in virtual   Dougherty, T. “Sex, Lies and Consent” (2013) 123(4) Ethics 717–44.   See section 76 of the Sexual Offences Act 2003. 40   See, for example, the cases of R v. Tabussum [2000] 2 Cr App R 328 and R v. Green [2002] EWCA Crim 1501. 41   See, for example, the case of R v. Elbekkay [1995] Crim LR 163 (CA). 38 39

BARFIELD PRINT.indd 384

13/11/2018 11:53

The law and ethics of virtual sexual assault  385 interactions than real-world interactions. The case of R v. Devonald already suggests as much, showing how easy it is for there to be deception as to the purpose of what is taking place when the parties to the act are not in direct physical contact with one another. More fancifully, some people have suggested that sexual versions of the “Chinese Room” thought experiment could be realized in virtual environments.42 The Chinese Room thought experiment was originally proposed by John Searle as a way of arguing that computers could never think in the same way as humans (never have consciously directed thoughts). The experiment asks you to imagine a man inside a room with a big book of instructions. Chinese symbols are fed into the room and the big book of instructions tells the man how to respond to each and every one of these symbols. Unbeknown to the man inside the room, the net effect of this is for somebody outside the box to be able to carry out a conversation (in Chinese) with the room. One party—the man inside the room—is completely ignorant as to what is taking place. A similar level of ignorance could arise with virtual sexual activity. One human user of a virtual environment could input instructions and commands to their avatar that, unbeknown to them, have the net effect of sexually assaulting somebody else’s character, or vice versa. Virtual sexual activity’s susceptibility to these forms of deception probably does not have any major ethical or normative significance. If we think that certain forms of deception vitiate consent in the real world, then they should also vitiate consent in virtual worlds. That said, the relative ease with which deception might occur in virtual worlds suggests that there is added risk to virtual sexual encounters, and perhaps a need for higher standards or norms of consent, such as additional checks and verifications of identity. A second problem with consent in virtual worlds arises from the possibility of resisting or avoiding unwanted sexual contact. Historically, many jurisdictions have included a “force” requirement in their definitions of sexual assault and rape. The idea was that a sexual act was only nonconsensual or unwanted if the victim actively resisted its performance and so required the perpetrator to use force to complete the act. Active resistance could manifest in many ways, such as trying to escape the perpetrator’s clutches or actual physical violence. The force requirement placed a very onerous duty on victims of sexual assault, and was not, according to many critics of the law, fair to the lived experience of assault victims, many of whom could freeze or become psychologically paralyzed by the trauma of the assault. Fortunately, there has been a general shift away from the force requirement in the past half century—though it lingers in some places—and an acceptance of the fact that victims of sexual assault do not need to actively resist unwanted sexual advances in order to convince a court that they are victims of a sexual crime. There is an interesting normative and sociocultural question to be asked regarding why this shift has happened. Is it simply because we have become convinced that consent should be an active, affirmative agreement and should never be inferred from outward passivity? Or is it because we feel it is impractical (and hence unfair) to place a duty on victims of sexual assault to actively resist their attackers? The answer may make some difference when it comes to assessing the moral seriousness of virtual sexual assault.

42   Sanberg, A. “The Chinese Pleasure Room: Ethics of Technologically Mediated Interaction,” Practical Ethics Blog, September 20, 2016, http://blog.practicalethics.ox.ac.uk/2016/09/the-chin​ ese-pleasure-room-ethics-of-technologically-mediated-interaction/.

BARFIELD PRINT.indd 385

13/11/2018 11:53

386  Research handbook on the law of virtual and augmented reality One common rebuttal to the claim that virtual sexual assault should be taken seriously—“common” in the sense that whenever I have discussed the topic it has been brought up—is that it is much easier to resist or avoid virtual sexual assault than it is to resist or avoid real-world sexual assault. In the real world, it may be hard to get away from your attacker. Their physical presence may cause you to freeze up or shut down and hence become passive and nonresistant. In a virtual world, it will be much easier to get away: just take off your headset or switch off your computer. You are instantly free from the environment in which the assault was taking place.43 Consider once more the story of Jordan Belamire, discussed in the introduction to this chapter. While we may sympathize with what happened to her and agree that the actions of her perpetrator were abominable, she could have quickly escaped from his clutches by taking off her headset. The upshot of this is the following: If we think that the practicality of resistance is what matters when it comes to nonconsent, and if we agree that it is much easier to resist in a virtual environment, we might think that a force requirement should be included in the definition of what counts as a virtual sexual assault. We might go even further and start to think that there are virtual equivalents of “asking for it.” So if someone enters a virtual world (while, say, wearing a haptic sexsuit) in which virtual sexual assault is known to be common, or if they fail to absent themselves from an environment when such behavior becomes apparent, they were “asking for it” and do not deserve our sympathy. In other words, we might argue that it is only those cases where someone is locked into a virtual environment against their will that are worth taking seriously. There are three reasons why I think we should reject such a view. The first is simply that even if the practicality of resistance is the factor that has moved us away from a force requirement, there is no reason to think that the psychological forces that make resistance impractical (and hence unfair) in the real world would not apply to the virtual world. Jordan Belamire’s description of her virtual reality groping points out how realistic, and traumatic it was to be immersed in the virtual world, suggesting the psychological distance between her virtual self and real self was not all that great. Second, and more importantly, the practicality of resistance should not be the decisive factor in any event. The reason for rejecting a force requirement should, in my opinion, lie in the preferability of open, affirmative consent standards. These should apply to the virtual world as well and we should not read consent into lack of virtual resistance. Finally, because the boundary between the virtual and the real is increasingly blurred, and because we now live much of our lives through virtual-world representations of ourselves, it seems unfair to insist that victims of virtual sexual assault should absent themselves from virtual worlds when things start to get unpleasant. Just as women should not be required to modulate their dress for fear of being sexually assaulted on city streets, so too should people not be required to flee from a virtual assaulter. The duties and responsibilities should fall on the perpetrator, not the victim.

43   Jaclyn Seelagy notes that the degree of customizability of the user experience and the manufacturer’s safeguards against unwanted interactions would also be relevant factors in thinking about both consent and the appropriate subject of responsibility in the event of a nonconsensual interaction. See Seelagy, J. “Virtual Violence” (2016) 64 UCLA Law Review Discourse 412­–33.

BARFIELD PRINT.indd 386

13/11/2018 11:53

The law and ethics of virtual sexual assault  387

5. CONCLUSION Virtual sexual assault is an emerging phenomenon and is likely to become more prevalent in the future as we continue to live more of our lives in virtual environments. It is important that we start to think about the seriousness of this type of assault, its appropriate classification, and the desirable legal and ethical responses to its occurrence. Building on initial work done by others, particularly the work of Litska Strikwerda, I have tried to provide a framework for thinking about these important issues. I have not tried to defend any particular view, rather I have presented a series of arguments about how we should understand and respond to this phenomenon. The main conclusions of these arguments can be summarized as follows. First, I have argued that virtual sexual assault needs to be classified along three dimensions (victim–perpetrator dimension; avatar–immersive interaction dimension; human user–virtual character dimension). Using this classificatory scheme we can identify six major types of virtual sexual assault that are worth taking seriously either because they involve a human perpetrator, or a human victim, or both. Second, I have argued that virtual sexual assault may warrant criminalization if (1) it has properties that make it “real” and (2) those properties would ordinarily warrant criminalization. Following the work of Philip Brey, I argued that there is a complex relationship between the virtual world and the real world, and that certain actions and events that take place in a virtual world are every bit as real as their real-world equivalents. This is true when the virtual acts have real-world consequences and/or when they are social, as opposed to physical, in nature. This means that some instances of virtual sexual assault are real, because virtual sexual assault can have real-world consequences and there are some grounds for thinking that certain aspects of sexual activity are social, as opposed to physical, in nature. What is more, the properties of virtual sexual assault that are real are ones that make it apt for criminalization, because it can involve harm to others. The actual approach to criminalization is complex, with certain forms of virtual sexual assault likely to be covered by existing laws, and others necessitating new laws. Third, I have argued that it is possible to hold people responsible for virtual sexual assault. Where the virtual assault is perpetrated by a human acting in a virtual environment, there is a responsible moral agent to whom responsibility can be ascribed. Any claims to the effect that there is some psychological/moral distance between the real self and the virtual self that make such ascriptions of responsibility impossible should be resisted. Furthermore, in cases where there is no immediate human perpetrator, it may be possible to hold the designers and manufacturers of the virtual environment liable for the actions that take place within it. In this respect, there is much to be learned from the “responsibility gap” debate in relation to robotic weapons systems. Finally, I argued that our approach to consent in the case of virtual sexual assault should mimic our approach to consent in the case of real-world sexual assault, although we should acknowledge a greater potential for deception in virtual sexual interactions, while at the same time rejecting the claim that the ease with which someone can resist virtual sexual assault is a relevant factor in its normative assessment.

BARFIELD PRINT.indd 387

13/11/2018 11:53

388  Research handbook on the law of virtual and augmented reality

SOURCES Cases R v. Devonald [2008] EWCA Crim 527 R v. Elbekkay [1995] Crim LR 163 (CA) R v. Green [2002] EWCA Crim 1501 R v. Tabussum [2000] 2 Cr App R 328

BARFIELD PRINT.indd 388

13/11/2018 11:53

13.  Criminal liability for intellectual property offenses of artificially intelligent entities in virtual and augmented reality environments Gabriel Hallevy4*

1.  INTRODUCTION: THE LEGAL PROBLEM 1.1 The Simplicity of Criminal Law Issues for Offenses Committed in Virtual and Augmented Reality Environments by Human Offenders: The Factual Mistake Doctrine Virtual worlds are created by computer code that generates each aspect of the virtual environment, which includes the visual scene, the 3D spatialized sound presented to the user, and the tactile and force feedback that the user may experience interacting with virtual objects. Criminal offenses may be committed in any environment, including real, virtual, or augmented. From a criminal law perspective, when considering a human offender immersed within a virtual environment and performing criminal activities, the nature of the virtual environment may not be a major factor considered by the court; instead, the law will generally focus on the offender’s understanding of what can be termed the physical reality of the crime (which consists of what is true about the ­physical world and the person’s beliefs about the physical world); that is, the “real” ­reality associated with the offense, and not the virtual reality associated with the offense. When the virtual or augmented reality experienced by the user is different than the physical reality in a certain level of detail, the question of criminal liability then focuses on the offender’s belief, which is the basis for determining the offender’s fault. Most legal systems around the world prefer to use the offender’s belief as the basis for fault rather than what actually happened (that is, the factual reality of the crime). To legal actors in these jurisdictions, it seems more just and fair to impose criminal liability on the basis of the offender’s subjective view of reality. This is, in short, the factual mistake doctrine (“mistake of fact”): We prefer the offender’s factual mistake (subjective) over the factual situation (objective) that actually occurred as the basis for the offender’s fault. For example, say an adult is consensually having sex with a teenager, but truly thinks the teenager is an adult. If we take the factual reality as the basis for fault, this is clearly a case of statutory rape in most jurisdictions. In contrast, if we take the offender’s sincere belief that the partner was an adult, it is clearly not a criminal offense at all. As mentioned

*  I thank Professor Woodrow Barfield and Professor Marc Blitz for their helpful comments and insights on this issue. I also thank the participants of the Beyond IP seminar at Yale Law School for their helpful comments.

389

BARFIELD PRINT.indd 389

13/11/2018 11:53

390  Research handbook on the law of virtual and augmented reality above, most legal systems have chosen the offender’s belief (true and sincere belief) as the basis for attributing fault.1 Of course, this perspective can be used against the offender as well. For example, in order to arrest a pedophile, an undercover police detective may try to convince the pedophile that she (the detective) is a child. If the offender truly believes this false fact, and acts accordingly, he is then charged with the matching sex offense as if it were committed against a child, although according to the factual reality (what actually happened), it was not. The human offender’s (mistaken?) belief about the nature of his actions may be generated and caused by different circumstances. As discussed in this chapter, when he is interacting with a another person in virtual and augmented reality, such technologies (and the limits of what an offender knows about others there) may contribute to the criminal behavior. If this is the case, criminal law issues for offenses committed in virtual and augmented reality environments by human offenders are not very complicated and may be straightforward to solve. The “factual mistake doctrine” simplifies the legal situation, in that the offender is being charged, convicted, and sentenced, or exonerated, due to what he actually thought the factual reality was.2 Extending the argument, what if the offender is not human, such as a virtual avatar? Basically, we may use the factual mistake doctrine with any given offender, including corporations or any other nonhuman offender. However, when we examine the combination of virtual and augmented reality with artificial intelligence technology, there are additional challenges confronting criminal law doctrine. 1.2 The Legal Problem due to the Combination of Virtual and Augmented Reality with Artificial Intelligence Technology Most current virtual and augmented reality devices use artificial intelligence (AI) technology for a range of activities (for example, the purchase and selling of goods), some of which may violate criminal law statutes, such as if fraud is involved. With the use of AI, virtual reality is able to reach a level of realism such that it can imitate factual reality more accurately than systems operating without AI; this is due to the sophistication of the algorithms and analytical techniques which comprise AI. When artificially intelligent software is directed to imitate factual reality, there is a high probability that AI might copy various representations of “reality” which may receive protection through intellectual property (IP) or other rights (such as a right of publicity). One way that the imitation of reality (which may be eligible for protection under IP rights) may be done is through the act of copying protected software into the virtual reality framework. However, it is important to note that the infringement of the rights associated with the imitation of 1   See, e.g., United States v. Short, 4 U.S.C.M.A. 437, 16 C.M.R. 11 (1954); United States v. Lampkins, 4 U.S.C.M.A. 31, 15 C.M.R. 31 (1954); State v. Ebbeler, 283 Mo. 57, 222 S.W. 396 (1920); McQuirk v. State, 84 Ala. 435, 4 So. 775 (1888). 2   According to the transformation-of-fault doctrine, if the offender has entered into this situation intentionally and knowingly, this may act as evidence for not actually thinking that it was the factual reality. See more in Gabriel Hallevy, The Matrix of Insanity in Modern Criminal Law (2015).

BARFIELD PRINT.indd 390

13/11/2018 11:53

Criminal liability of AI entities  391 reality may be perpetuated by the artificially intelligent software itself, and not done under the direction of its operator; this level of independence by the AI entity poses significant challenges for various legal schemes. Although the main legal protection granted to IP rights is civil, legislators may also define certain types of protections through the statutes of criminal law.3 This legal response reflects society’s wide interest in protecting IP rights. Thus, when IP rights are violated, and the violation fulfills the basic requirements of a relevant offense, society may indict the violator, regardless of the violator’s identity (corporation or human). But what if the violator of IP rights that exist in virtual or augmented reality is an AI entity; for example, a virtual avatar or digital assistant controlled by AI? Can society indict an AI entity? Most modern AI entities do in fact have the technological capability of violating IP rights, so this is not a moot question.4 As an example, trademark and copyright law may be implicated if an AI entity uses the trademarked representation of a celebrity’s virtual appearance for commercial gain. Clearly, when we consider the vast number of applications of virtual reality, along with advances in the technology of AI, there is an increased chance that AI will play a more significant role in violating the various IP rights associated with virtual and augmented reality environments. In general, virtual reality may include both changes in the factual reality and in the legal reality of a given situation. For example, by changing the place of action, the action may be subordinated to the domestic law of that particular place. So if, under the laws of state A, IP infringements are not considered offenses, and a certain action has taken place in state A, then under most rules on conflict of laws no offense has been committed. Accordingly, if the virtual reality environment is believed to be in state A, then the virtual reality controls not only factual changes, but legal changes as well. However, a legal mistake is not an easy way to effect criminal liability. In most legal systems, the offender should prove that all reasonable measures have been taken in order to prevent the mistake, and still the mistake has occurred.5 Nevertheless, the technological principles involved in changing either the factual reality, the legal reality, or both are identical. When the AI entity “believed” it was doing something legal or “believed” it did not do it at all, under the relevant circumstances of virtual reality, there is no significant difference between these situations. Therefore, we may assume both factual and legal mistakes in this chapter as part of the virtual or augmented reality experienced by a participant. Illegal copying of software (for example, the software creating a virtual sword in Second Life) by another software program equipped with AI technology is a prime example of how AI could violate IP or criminal law. In this scenario, an interesting question is: Could the AI entity itself be criminally liable, that is, beyond the criminal liability of the manufacturer, the end user, or owner, and beyond their civil liability? This chapter suggests a possible positive answer. Using the current definitions of criminal liability,   See, e.g., 35 U.S.C.S. § 292 and 17 U.S.C.S. § 506.   See examples in Gabriel Hallevy, When Robots Kill—Artificial Intelligence under Criminal Law (Northeastern Univ. Press 2013): for intentional offenses at pp.38–83, for negligence offenses at pp.84–103, and for strict liability offenses at pp.104–19. 5   See, e.g., Ratzlaf v. United States, 510 U.S. 135, 114 S.Ct. 655, 126 L.Ed.2d 615 (1994); United States v. Mitchell, 209 F.3d 319 (4th Cir. 2000); United States v. Fuller, 162 F.3d 256 (4th Cir. 1998). 3 4

BARFIELD PRINT.indd 391

13/11/2018 11:53

392  Research handbook on the law of virtual and augmented reality the imposition of criminal liability upon AI entities for committing IP offenses in virtual environments seems feasible. However, it should be noted that in the view of the author, as a practical matter, no criminal liability should be imposed on AI systems, at least as yet. However, if basic definitions of criminal law are not changed, the odd consequence of AI being involved in IP violations, but not liable for such violations under current legal doctrine, is inevitable. With this background, the modern criminal law is modular, as criminal liability may be imposed when all of its requirements are fulfilled. Further, each requirement has its own definitions and terms to be fulfilled, as defined by the basic concepts of criminal law. The way in which the modular nature of criminal law relates to IP and AI entities within virtual and augmented reality worlds is discussed more fully below. Given the focus on criminal law, the major question addressed in this chapter is: who is to be held responsible for IP offenses that occur in virtual and augmented reality worlds when committed by AI entities controlling the actors in virtual worlds? This is an important question because the twenty-first century world is seeing rapid technological development. In the technological world of the twenty-first century, simple daily activities are being undertaken by robots and computers equipped with AI, and more and more people are accessing virtual and augmented reality technology. As long as humans used computers as mere tools, there was no real difference under the law between computers and hammers, chairs, or pencils; for example, under the common law, each was viewed as the property of the owner. However, when software became more sophisticated, some commentators began to argue that computers were able to “think,” which then took humans out of the decisionmaking loops of the system. A problem began to develop (especially for assigning tort or criminal liability) when computers evolved from “thinking” machines into thinking machines. Thus emerged the challenges AI began to pose for established legal doctrine. Artificial intelligence research began in the middle of the twentieth century,6 and since then AI entities have become an integral part of modern human life, functioning in a much more sophisticated manner than other daily tools. Could they become a challenge or even a danger to society under the tenants of criminal law? It is already the case that, at least in some ways, AI does pose “dangers” to society;7 and this concern has caught the attention of authors and researchers. For example, in 1950, Isaac Asimov set down three fundamental laws of robotics in his science fiction masterpiece I, Robot.8 These laws were: (1) a robot may not injure a human being or, through inaction, allow a human being to come to harm; (2) a robot must obey the orders given it by human beings, except where such orders would conflict with the First Law; and (3) a robot must protect its own existence, as long as such protection does not conflict with the First or Second Laws.9 These three fundamental laws discussed in the context of robotics are obviously contradictory.

  N.P. Padhy, Artificial Intelligence and Intelligent Systems 3–29 (2005).   For example, in 1981, a 37-year-old Japanese employee of a motorcycle factory was killed by an artificial intelligence robot working near him. The robot erroneously identified the employee as a threat to its mission, and calculated that the most efficient way to eliminate this threat was by pushing him into an adjacent operating machine. Using its very powerful hydraulic arm, the robot smashed the surprised worker into the operating machine, killing him instantly, and then resumed its duties with no one to interfere with its mission. 8   Isaac Asimov, I, Robot (1950). 9   Id. 6 7

BARFIELD PRINT.indd 392

13/11/2018 11:53

Criminal liability of AI entities  393 Isaac Asimov himself wrote in 1964 that “[t]here was just enough ambiguity in the Three Laws to provide the conflicts and uncertainties required for new stories, and, to my great relief, it seemed always to be possible to think up a new angle out of the 61 words of the Three Laws.”10 Of relevance to this chapter is the question: Do Asimov’s laws of robotics relate with equal force to intelligent entities (such as virtual avatars) existing in virtual and augmented reality? And if so, how should the criminal law respond to the appearance of increasingly intelligent virtual entities in virtual and augmented reality? With Asimov’s laws in mind, what if a person orders a robot (or virtual avatar) to hurt another person (or digital entity) for the good of the other person? What if the robot is in police service and the commander of the mission orders it to arrest a suspect and the suspect resists arrest? Or what if the robot is in medical service and is ordered to perform a surgical procedure on a patient, and the patient objects, but the medical doctor insists that the procedure is for the patient’s own good and repeats the order to the robot? Whatever the solutions to such hypothetical scenarios under the fictitious statutes developed by Asimov, his fundamental laws of robotics relate on their face only to robots. AI software not installed in a robot would not be subject to Asimov’s laws, even if these laws had any real legal significance. Nevertheless, although Asimov was a prolific science fiction author, his Three Laws were written for the technological realm that was possible according to his vision of the future. Of course, artificial intelligence technology has been vastly developed since then, and as a consequence technologically advanced societies are facing some of the problems Asimov predicted, even in the worlds found in virtual and augmented reality. 1.3  Some Virtual Reality IP Issues and Disputes Before discussing various models of criminal liability relating to AI entities involved in IP disputes in virtual and augmented reality, as background I briefly outline some IP issues that relate more generally to virtual and augmented reality. Under copyright law, it is established that there is liability for infringement in the cases where strikingly similar or nearly identical copying occurs for virtual objects.11 Thus, from IP law, there is little question that copyright protection will extend in some form to virtual works and even to virtual objects. Therefore, for virtual worlds, an interesting question to pose is: What level of similarity for works produced by an AI entity operating within a virtual world will be sufficient for copyright infringement? And if copying of some aspect of the virtual world is an illegal act, would the AI entity have the necessary mens rea to be found culpable for a criminal law violation? The criminal law, of course, has a major role to play for many actions occurring in virtual worlds, and several jurisdictions have begun to address the growing issue of criminal disputes in virtual and augmented reality. For example, among others, the European Network and Information Security Agency and the Australian Institute of Criminology have both considered the issue of crime in virtual worlds.12 One particular crime—the   Isaac Asimov, The Rest of Robots (1964).   For a basic review of copyright law see: http://legal-dictionary.thefreedictionary.com/ copyright+infringement. 12   See Wayne Rumbles, Theft in the Digital: Can You Steal Virtual Property? 17 Canterbury Law Review, 354, 364 (2011). 10

11

BARFIELD PRINT.indd 393

13/11/2018 11:53

394  Research handbook on the law of virtual and augmented reality theft of virtual property which may have external value—falls within the Australian Institute of Criminology’s interests, but formal criminal intervention would only be initiated if there was an appreciable and measurable effect on the real-world victim. Given the numerous scenarios which can occur when AI entities operate in virtual and augmented reality worlds, it is necessary to consider whether the AI entity itself can be liable for its actions. The current chapter presents three models of criminal liability for AI entities in virtual and augmented reality, among others to clarify the issue of culpability for IP crimes committed by the different actors involved in virtual and augmented reality worlds. It should be noted that criminal actions which occur in virtual environments have the potential to bleed into the real world. Consider the case of Qui Chengwei, who was a player in the virtual world of Legend of Mir II.13 By playing the game, he had earned a Dragon Sabre; he subsequently loaned this to another man, Zhu Caoyuan, who sold the weapon without the former’s permission. If the Dragon Sabre had been a physical item, Caoyuan would have been liable for theft by conversion. However, when the injured individual initially sought the assistance of the police, he was told that the theft was not a crime, since virtual property was not covered as a protectable asset under the then current law. In response to the theft, Chengwei attacked the alleged thief and killed him. Even though Chengwei was sentenced to death for a very real murder,14 the case is important to the area of virtual theft. It shows that a user (or their avatar?) can gain possession of another user’s virtual goods, either with restricted permission or by subterfuge, and that doing so may be a crime. Following the Chengwei case, and in the wake of an increasing number of cases of virtual theft, China’s criminal law actively began to protect virtual property. This led to two 17-year-old boys being sentenced for virtual property theft in 2005. The above discussion is just the tip of the iceberg in terms of disputes that are beginning to occur in virtual and augmented reality. The remainder of this chapter focuses on criminal liability for AI entities across a range of criminal actions and scenarios. 1.4  Criminal Liability With the above brief discussion in mind, the main questions with regard to AI criminal liability with respect to virtual and augmented reality worlds are: (1) which kind of laws or ethics are correct; and (2) who is to decide? In order to cope with such problems as they relate to humans, society devised systems of criminal law, which embodies the most powerful legal social control in modern civilization. Fear of AI entities often arises because AI entities are not considered to be subject to the law, specifically to criminal law.15 Will these same fears and concerns transfer to the use of AI in virtual and augmented reality worlds? That is: To what extent will real-world law bleed into virtual and augmented reality worlds, especially when events occurring within such worlds are more and more under the direction of AI operating autonomously from a human? 13   Allen Chein, A Practical Look at Virtual Property, 80 St John’s Law Review, 1059 (2006) (provides a review of the Qui Chengwei case and others). 14   Note that the sentence was commuted to life in prison. 15   The apprehension that AI entities evoke may have arisen due to Hollywood’s depiction of AI entities in numerous films, as discussed in the following paragraph. Treatment of AI entities has included clear subordination to human legal social control and to criminal law.

BARFIELD PRINT.indd 394

13/11/2018 11:53

Criminal liability of AI entities  395 The apprehension evoked by AI may have come about as a result of Hollywood’s depiction of AI entities in numerous films, such as 2001: A Space Odyssey (1968)16 and the modern trilogy The Matrix (1999, 2003, 2003),17 in which AI entities are not subject to the law. However, it should be noted that Hollywood did treat AI entities in an empathic way as well, depicting them as human, as almost human, or as wishing to be human (the AI agent in the movie Her also received empathic treatment).18 This kind of treatment included, of course, clear subordination to human legal social control, and to criminal law. In the past, people were similarly fearful of corporations and their power to commit a spectrum of crimes, but since corporations came to be considered legal entities subject to criminal and corporate law, that kind of fear has been reduced significantly.19 Therefore, the modern question relating to AI entities becomes: Does the growing intelligence of AI entities subject them, as any other legal entity, to legal social control?20 This chapter attempts to work out a legal solution to the problem of the criminal responsibility of AI entities, specifically with regard to IP offenses that occur in virtual and augmented reality worlds (such as the act of illegal copying or the theft of virtual property). On this issue, consider an AI software which functions as an expert system, but copies protected information as it independently learns that it is much easier and faster to copy information rather than independently creating it. For this act of copying, could the AI be punishable? To address this question, a description of the attributes of an AI entity will be presented; based on that description, the chapter will then propose and introduce three possible models of criminal responsibility for an AI entity: (1) perpetration-by-another responsibility; (2) natural-probable-consequence responsibility; and (3) direct responsibility. These possible models could be applied separately to IP violations occurring within virtual and augmented reality worlds, but in many situations a coordinated combination of the models (using all or some of them) is required in order to complete the legal structure of criminal responsibility. Once I examine the possibility of legally imposing criminal responsibility on AI entities, then the question of punishment must be addressed. How can an AI entity serve a sentence of incarceration? How can probation, a pecuniary fine,

  2001: A Space Odyssey (Metro-Goldwyn-Mayer 1968).   The Matrix (Village Roadshow Pictures 1999); The Matrix Reloaded (Village Roadshow Pictures 2003); The Matrix Revolutions (Village Roadshow Pictures 2003). 18   See, e.g., A.I. Artificial Intelligence (Amblin Entertainment 2001). 19   See, e.g., John C. Coffee, Jr, “No Soul to Damn: No Body to Kick”: An Unscandalized Inquiry Into the Problem of Corporate Punishment, 79 Mich. L. Rev. 386 (1981); Steven Box, Power, Crime and Mystification 16–79 (1983); Brent Fisse & John Braithwaite, The Allocation of Responsibility for Corporate Crime: Individualism, Collectivism and Accountability, 11 Sydney L. Rev. 468 (1988). 20   See generally L. Thorne McCarty, Reflections on Taxman: An Experiment in Artificial Intelligence and Legal Reasoning, 90 Harv. L. Rev. 837 (1977); E. Donald Elliott, Holmes and Evolution: Legal Process as Artificial Intelligence, 13 J. Legal Stud. 113 (1984); Bruce G. Buchanan & Thomas E. Headrick, Some Speculation about Artificial Intelligence and Legal Reasoning, 23 Stan. L. Rev. 40 (1971); Antonio A. Martino, Artificial Intelligence and Law, 2 Int’l J. L. & Info. Tech. 154 (1994); Edwina L. Rissland, Artificial Intelligence and Law: Stepping Stones to a Model of Legal Reasoning, 99 Yale L. J. 1957 (1990). 16 17

BARFIELD PRINT.indd 395

13/11/2018 11:53

396  Research handbook on the law of virtual and augmented reality and so on be imposed on an AI entity? Consequently, it is necessary to formulate viable forms of punishment in order to impose criminal responsibility practically on AI entities. Thus, for instance, how can the provision of the US Code which relates to false marking (. . . shall be fined not more than $500 for every such offense)21 be implemented when the offender is an AI system? And will it make a difference if the offender is a virtual entity operating within virtual and augmented reality?

2.  AI ENTITIES For some years, there has been significant controversy about the very essence of what constitutes an AI entity.22 Futurologists have even proclaimed the birth of a new species, machina sapiens, which they postulate will share with humans the position of most intelligent creature on earth (and maybe even surpass humans).23 However, some critics have argued that a “thinking machine” is an oxymoron.24 They claim that machines (including computers), with their foundations of cold logic, can never be as insightful or creative as humans.25 This controversy raises basic questions of the very essence of humanity (do human beings function as thinking machines?) and of AI (can there be thinking machines?).26 And if AI can be a “thinking machine,” how does this development affect criminal law for virtual entities that are intelligent and interact with humans and with each other in virtual and augmented reality environments?

21   35 U.S.C.S. § 292 provides: “(a) Whoever, without the consent of the patentee, marks upon, or affixes to, or uses in advertising in connection with anything made, used, offered for sale, or sold by such person within the United States, or imported by the person into the United States, the name or any imitation of the name of the patentee, the patent number, or the words ‘patent,’ ‘patentee,’ or the like, with the intent of counterfeiting or imitating the mark of the patentee, or of deceiving the public and inducing them to believe that the thing was made, offered for sale, sold, or imported into the United States by or with the consent of the patentee; or Whoever marks upon, or affixes to, or uses in advertising in connection with any unpatented article, the word ‘patent’ or any word or number importing that the same is patented, for the purpose of deceiving the public; or Whoever marks upon, or affixes to, or uses in advertising in connection with any article, the words ‘patent applied for,’ ‘patent pending,’ or any word importing that an application for patent has been made, when no application for patent has been made, or if made, is not pending, for the purpose of deceiving the public—Shall be fined not more than $500 for every such offense. Only the United States may sue for the penalty authorized by this subsection. (b) A person who has suffered a competitive injury as a result of a violation of this section may file a civil action in a district court of the United States for recovery of damages adequate to compensate for the injury. (c) The marking of a product, in a manner described in subsection (a), with matter relating to a patent that covered that product but has expired is not a violation of this section.” 22   See, e.g., Terry Winograd, Thinking Machines: Can There Be? Are We? in The Foundations of Artificial Intelligence 167 (Derek Partridge & Yorick Wilks eds, 2006). 23   Id at 167. 24   Id at 167–8. 25   Id at 168–9. 26   For the formal foundations of AI see, for example, Teodor C. Przymusinski, Non-Monotonic Reasoning versus Logic Programming: A New Perspective, in The Foundations of Artificial Intelligence, supra note 22, at 49; Richard W. Weyhrauch, Prolegomena to a Theory of Mechanized Formal Reasoning, in The Foundations of Artificial Intelligence, supra note 22, at 72.

BARFIELD PRINT.indd 396

13/11/2018 11:53

Criminal liability of AI entities  397 As a basic observation, there are five attributes that one would expect an intelligent entity (within a virtual environment or not) to possess.27 The first is communication.28 One can communicate with an intelligent entity. The easier it is to communicate with an entity, the more intelligent the entity seems; this observation holds true for virtual avatars and for virtual entities that perform commercial transactions for humans. To illustrate this idea: One can communicate with a dog, but not about Einstein’s theory of relativity. Further, one can communicate with a child about Einstein’s theory, but it requires a discussion in terms that a child can comprehend. The second attribute of artificial intelligence is internal knowledge:29 An intelligent entity is expected to have some knowledge about itself. For a virtual avatar, the more it possesses self-knowledge, the more human it seems and the more people will react to it as if it were human. The third attribute of artificial intelligence is external knowledge:30 An intelligent entity is expected to know about the outside world, to learn about it, and to utilize that information. This aspect of artificial intelligence is especially relevant for augmented reality, where virtual images are projected within a real-world scene. The fourth attribute of artificial intelligence is goal-driven behavior:31 An intelligent entity is expected to take action in order to achieve its goals. We see this attribute of artificial intelligence occurring more frequently in virtual environments as virtual agents appear to engage in goal-driven behavior, often in the role of digital assistant for a person or corporation. The fifth aspect of artificial intelligence is creativity:32 An intelligent entity is expected to have some ability to take alternate actions when the initial action fails. For example, a fly that tries to exit a room and bumps into a window panel lacks certain types of intelligence and thus engages in the same behavior repeatedly. However, given advances in AI, when an AI robot bumps into a window, its sensors detect the window and, using logic, it tries to exit using a door. Most AI entities possess these five attributes by definition.33 And some twenty-first-century types of AI entities possess even more attributes that enable them to act in far more sophisticated ways.34 In fact, AI entities have a wide variety of applications and embodiments. For example, AI entities can be designed to imitate the physical capabilities of a human being, and these 27   Roger C. Schank, What is AI Anyway? in The Foundations of Artificial Intelligence, supra note 22, at 3. 28   Id at 4–5. 29   Id at 5. 30   Or “world knowledge.” See id. at 5. 31   Hans Welzel, Das Deutsche Strafrecht—Eine Systematische Darstellung (1969). 32  Schank, supra note 27, at pp. 5–6. 33   Id at 4–6. 34   In November 2009, during the Supercomputing Conference in Portland, Oregon (SC 09), IBM scientists and others announced that they had succeeded in creating a new algorithm named “Blue Matter,” which possesses the thinking capabilities of a cat. Chris Capps, “Thinking” Supercomputer Now Conscious as a Cat, Unexplainable.net, Nov. 19, 2009, www.unexplainable. net/artman/publish/article_14423.shtml; see generally Supercomputing Conference 09, http://sc09. supercomputing.org/; What Is Blue Matter? www.hpcwire.com/hpc/1096657.html. This algorithm collects information from very many units with parallel and distributed connections. The information is integrated and creates a full image of sensory information, perception, dynamic action and reaction, and cognition. This platform simulates brain capabilities, and eventually is supposed to simulate real thought processes. The final application of this algorithm contains not only analog and digital circuits, metal, and plastics, but also protein-based biologic surfaces. Id.

BARFIELD PRINT.indd 397

13/11/2018 11:53

398  Research handbook on the law of virtual and augmented reality capabilities can be improved. Further, an AI entity is capable of being physically faster (in the virtual and physical world) and stronger (in the physical world) than a human being. The AI software controlling an AI entity allows it to perform many complicated calculations faster and simultaneously, or to “think” faster than a human. An AI entity is also capable of learning and of gaining experience, and experience is a useful way of learning. Based on the above characteristics, AI entities and AI software are used in a wide range of applications in industry, military services, medical services, science, and even in games played in virtual and augmented reality.35 All these attributes create the essence of an AI entity,36 and all these attributes offer challenges to the establishment of criminal law doctrine for intelligent entities operating in virtual and augmented reality environments.

3. THREE MODELS OF THE CRIMINAL LIABILITY OF ARTIFICIAL INTELLIGENCE ENTITIES FOR COMMISSION OF IP OFFENSES Say that an AI system is required to complete the preparation of software for a given task designed for public use. For example, the system may be designed as a debugger which then “fills in” the necessary software ingredients that may be required or missing for a particular application. The copying or replacement of software may occur when an AI entity copies a protected part of another software. Even though the AI system may make some changes in the copied software, this act of copying is still considered to be an IP violation, an act which is protected as a criminal offense in addition to civil protection. As policy, the state acts to protect society from such actions, and the administrative directive is to indict the violator in criminal proceedings.37 Under criminal law, in order to impose criminal liability upon a person, two main elements must exist. The first is the factual element (actus reus) to a crime, while the other is the mental element (mens rea). The actus reus requirement is expressed mainly by acts or omissions.38 Sometimes, other external elements are required in addition to conduct, such as the specific results of that conduct and the specific circumstances underlying the conduct.39 In most IP offenses the actual conduct which infringes IP rights (such as illegal copying) is the actus reus requirement (for example, the actual copying of the information).

35   See, e.g., William B. Schwartz, Ramesh S. Patil, & Peter Szolovits, Artificial Intelligence in Medicine: Where Do We Stand, 27 Jurimetrics J. 362 (1987); Richard E. Susskind, Artificial Intelligence, Expert Systems and Law, 5 Denning L.J. 105 (1990). 36   See also, e.g., Yorick Wilks, One Small Head: Models and Theories, in The Foundations of Artificial Intelligence, supra note 22, at 121; Alan Bundy & Stellan Ohlsson, The Nature of AI Principles, in The Foundations of Artificial Intelligence, supra note 22, at 135; Thomas W. Simon, Artificial Methodology Meets Philosophy, in The Foundations of Artificial Intelligence, supra note 22, at 155. 37   For more examples see Timothy L. Butler, Can a Computer Be an Author? Copyright Aspects of Artificial Intelligence, 4 Comm. Ent. 707 (1982). 38   W. H. Hitchler, The Physical Element of Crime, 39 Dick. L. Rev. 95 (1934); Michael S. Moore, Act and Crime: The Philosophy of Action and Its Implications for Criminal Law (1993). 39   John William Salmond, Salmond on Jurisprudence 505 (Glanville Williams ed., 11th

BARFIELD PRINT.indd 398

13/11/2018 11:53

Criminal liability of AI entities  399 Additionally, the mens rea requirement has various levels of mental elements. The highest level is expressed by knowledge, and sometimes that knowledge is accompanied by a requirement of intent or specific intention.40 Lower levels of mens rea are expressed by negligence (a “reasonable person should have known” standard),41 or by strict liability offenses.42 In most legal systems the mental element requirement of IP offenses is expressed through knowledge, that is, the offender is required to know or to be aware of the offending nature of the illegal conduct, that is, knowledge that an actor is violating protected IP rights. As AI improves and becomes more autonomous from humans, an interesting question is whether the AI agent can be said to possess the requisite mens rea necessary to commit a crime, such as the act of copying protected software. The above mens rea and actus reus requirements are the only criteria or capabilities which are required in order to impose criminal responsibility, not only for humans but for any other kind of entity, including corporations and (theoretically) AI entities. Additionally, any entity might possess further mental capabilities, such as creativity; and in fact some AI entities are thought to express creativity. However, in order to impose criminal responsibility, the existence of actus reus and mens rea defining the specific offense is adequate. No further capabilities are required for the imposition of criminal liability. Thus, on the one hand, an uncreative thief is still a thief, if the actus reus and mens rea of theft are proven. On the other hand, an ant is capable of biting creatively, but it is incapable of formulating the mens rea requirement; therefore, a creatively biting ant bears no criminal responsibility. As another example, a parrot is capable of repeating the words it hears, but it is incapable of formulating the mens rea requirement for libel.43 Would the same reasoning hold true for virtual entities operating within virtual and augmented reality worlds, and how would the role of increased intelligence for virtual entities influence the court’s thinking? The imposition of criminal responsibility on any kind of entity is dependent on proving the above two elements (mens rea, actus reus) beyond a reasonable doubt. When it has been proven that a person committed the criminal act knowingly or with criminal intent, that person is held criminally responsible for that offense.44 The relevant question concerning the criminal responsibility of AI entities operating within virtual and augmented reality ed. 1957); Glanville L. Williams, Criminal Law: The General Part § 11 (2nd ed. 1961); Oliver W. Holmes, Jr, The Common Law 54 (1923); Walter Wheeler Cook, Act, Intention, and Motive in the Criminal Law, 26 Yale L.J. 645 (1917). 40   J. Ll. J. Edwards, The Criminal Degrees of Knowledge, 17 Mod. L. Rev. 294 (1954); Rollin M. Perkins, “Knowledge” as a Mens Rea Requirement, 29 Hastings L.J. 953 (1978); United States v. Youts, 229 F.3d 1312 (10th Cir. 2000); United States v. Spinney, 65 F.3d 231 (1st Cir. 1995); State v. Sargent, 594 A.2d 401 (Vt. 1991); State v. Wyatt, 482 S.E.2d 147 (W. Va. 1996); People v. Steinberg, 595 N.E.2d 845 (N.Y. 1992). 41   Jerome Hall, Negligent Behavior Should Be Excluded from Penal Liability, 63 Colum. L. Rev. 632 (1963); Robert P. Fine & Gary M. Cohen, Comment, Is Criminal Negligence a Defensible Basis for Criminal Responsibility? 16 Buff. L. Rev. 749 (1966). 42   Jeremy Horder, Strict Liability, Statutory Construction and the Spirit of Liberty, 118 L.Q.R. 458 (2002); Francis Bowes Sayre, Public Welfare Offenses, 33 Colum. L. Rev. 55 (1933); A.P. Simester, Is Strict Liability Always Wrong? in Appraising Strict Liability (2005) 21–50. 43   See, e.g., Ormsby v. United States, 273 F.2d 977 (1921). 44   Of course, if the legal action is civil and it is meant for civil remedies, the burden of proof is not subjected to the standards of criminal law.

BARFIELD PRINT.indd 399

13/11/2018 11:53

400  Research handbook on the law of virtual and augmented reality worlds is: How can these entities fulfill the two requirements of criminal responsibility? This chapter proposes the possibility of imposing criminal responsibility on AI entities using three possible models of responsibility: the perpetration-by-another responsibility model; the natural-probable-consequence responsibility model; and the direct responsibility model. The following sections explain these three possible models. 3.1  Perpetration-by-Another Responsibility The perpetration-by-another responsibility model does not consider the AI entity as possessing any human attributes. The AI entity, such as a virtual avatar, is considered an innocent agent. Accordingly, due to this legal viewpoint, a machine is a machine, a digital entity is a digital entity, and neither under the law are considered human or considered culpable under criminal law. However, one cannot ignore an AI entity’s capabilities (which are improving), as mentioned above.45 But pursuant to the perpetration-by-another model, these capabilities are insufficient to deem the AI entity a perpetrator of an offense. These capabilities resemble the parallel capabilities of a mentally limited person, such as a child, or of a person who is mentally incompetent and thus lacks a criminal state of mind. When an offense is committed physically by an innocent agent (such as a virtual avatar performing agency duties for a person or corporation), the person who sent or activated the innocent agent is criminally responsible as a perpetrator-by-another.46 In such cases, the innocent agent is regarded as a mere instrument, albeit a sophisticated instrument, while the party orchestrating the offense is the real perpetrator as a principal in the first degree and is held accountable for the conduct of the innocent agent. The perpetratorby-another’s responsibility is determined on the basis of the innocent agent’s conduct,47 as well as the perpetrator-by-another’s own mental state.48 In such situations, the derivative question relative to AI entities such as virtual avatars is: Who is the perpetrator-by-another? There are two candidates: The first is the programmer of the AI software and the second is the end user. For example, a programmer of AI software might design a program in order to commit IP offenses via the AI entity. In this case, the same programmer may design software for an operating AI entity. Say an AI entity is intended to be placed in a software lab, and its software is designed to copy information from other IP-protected software in the lab. In this scenario, although the AI entity committed the offense, the programmer is deemed the perpetrator. Another person who might be considered the “perpetrator-by-another” is the user of the AI entity. Even though the user did not program the software, he uses the AI entity, including its software, for his own purpose and benefit. For example, under robotics, a user may purchase a servant-robot, which is designed to execute any order given by its master. The specific user is identified by the AI entity as the master, and the master orders   See supra section 2.   Morrisey v. State, 620 A.2d 207 (Del. 1993); Conyers v. State, 790 A.2d 15 (Md. 2002); State v. Fuller, 552 S.E.2d 282 (S.C. 2001); Gallimore v. Commonwealth, 436 S.E.2d 421 (Va. 1993). 47   Dusenbery v. Commonwealth, 263 S.E.2d 392 (Va. 1980). 48   United States v. Tobon-Builes, 706 F.2d 1092 (11th Cir. 1983); United States v. Ruffin, 613 F.2d 408 (2d Cir. 1979); see also Gabriel Hallevy, Victim’s Complicity in Criminal Law, 2 Int’l J. Punishment & Sentencing 72 (2006). 45 46

BARFIELD PRINT.indd 400

13/11/2018 11:53

Criminal liability of AI entities  401 the AI entity to copy certain protected information. The AI entity executes the order exactly as ordered. This example is not substantially different from a person who orders his dog to attack a trespasser, or any person who orders a virtual digital assistant to access another’s financial information. In the latter case, the AI entity committed the offense, but the user is deemed the perpetrator.49 In the scenarios just given, the actual offense was committed by the AI entity. The programmer or the user did not perform any action conforming to the definition of a specific offense; therefore, neither the programmer nor the user meets the actus reus requirement of the specific offense. The perpetration-by-another responsibility model considers the action committed by the AI entity as if it had been the programmer’s or the user’s action. The legal basis for this outcome is the instrumental usage of the AI entity as an innocent agent. Under different jurisdictions, no mental attribute required for the imposition of criminal responsibility is attributed to the AI entity.50 When programmers or users use an AI entity instrumentally as is often the case with virtual avatars, the commission of an offense by the AI entity is attributed to them. The mental element required for the specific offense already exists in their mind. For example, the programmer had criminal intent when he ordered the commission of arson, and the user had criminal intent when he ordered the commission of an assault, even though these offenses were actually committed through an AI entity. When an end user makes instrumental usage of an innocent agent to commit a crime, the end user is deemed the perpetrator. Thus, in the context of virtual and augmented reality we do not attribute an intelligent agent operating in either environment as a perpetrator of a crime. The perpetration-by-another responsibility model does not attribute any mental capability, or any human mental capability, to the AI entity. According to this model, there is no legal difference between an AI entity such as a virtual avatar and a screwdriver or an animal. When a burglar uses a screwdriver in order to open a window, he uses the screwdriver instrumentally, and the screwdriver is not thought to be criminally responsible. Lacking mental capacity, the screwdriver’s “action” is, in fact, the burglar’s. This is the same legal situation when using an animal instrumentally or a virtual avatar instrumentally. An assault committed by a dog by order of its master is, for legal purposes, an assault committed by the master. This particular legal model might be suitable for two types of scenarios. The first scenario is using an AI entity to commit an IP offense without using its advanced capabilities, which enable it to “think” or to think (such as when an AI entity decides to commit an IP offense based on its own accumulated experience or knowledge). The second scenario is using a very old version of an AI entity, which lacks the modern advanced “thinking” capabilities of modern AI entities.51 In both scenarios, under the law, the use of the AI entity is as

49   See, e.g., Regina v. Manley, (1844) 1 Cox Crim. Cas. 104; Regina v. Cogan, [1976] Q.B. 217, [1975] 2 All E.R. 1059, [1975] 3 W.L.R. 316, 61 Cr. App. Rep. 217, 139 J.P. 608; Gabriel Hallevy, Theory of Criminal Law vol. II 700–706 (2009). 50   The AI entity is used as an instrument and not as a participant, although it uses its features of processing information. See, e.g., Cary G. Debessonet & George R. Cross, An Artificial Intelligence Application in the Law: CCLIPS, A Computer Program that Processes Legal Information, 1 High Tech. L.J. 329 (1986). 51   For some of the modern advanced capabilities of the modern AI entities see generally

BARFIELD PRINT.indd 401

13/11/2018 11:53

402  Research handbook on the law of virtual and augmented reality i­ nstrumental usage. Still, it is usage of an AI entity by a user or programmer, due to its ability to execute an order to commit an offense. A screwdriver cannot execute such an order; a dog can. A dog cannot execute complicated orders; an AI entity can.52 But as AI improves in capabilities, the perpetration-by-another responsibility model is not suitable when an AI entity decides to commit an offense based on its own accumulated experience or knowledge.53 That is, this model is not suitable when the software controlling the AI entity was not designed to commit the specific offense, but was committed by the AI entity nonetheless. This model is also not suitable when the specific AI entity functions not as an innocent agent, but as a semi-innocent agent.54 However, the perpetration-by-another responsibility model might be suitable when a programmer or user makes instrumental usage of an AI entity, but without using the AI entity’s advanced capabilities. The legal result of applying this model is that the programmer and the user are fully criminally responsible for the specific offense committed, while the AI entity has no criminal responsibility whatsoever. 3.2  Natural-Probable-Consequence Responsibility The natural-probable-consequence responsibility model assumes deep involvement of the programmers or users in the AI entity’s daily activities, but without any intention of committing any offense via the AI entity. For this model, consider the following scenario: During the execution of its daily tasks, an AI entity commits an offense; for example, a virtual avatar may use defamatory language. In this case, the programmers or users had no knowledge of the offense until it had already been committed; they did not plan to commit the offense, and they did not participate in any part of the commission of that specific offense. Of course, this example raises the question: Who should be held liable for the offense? Another example of such a scenario is an AI entity that is designed to browse information on the internet to determine solutions for certain problems and which, at some point, copies IP-protected information from the internet and presents it as its own data. When the AI entity’s figures/data are published, it appears to be infringing the IP rights of its actual creator; this may involve criminal liability.55 The programmer may not have intended to violate any laws, but nonetheless, IP rights were infringed as a result of the AI entity’s actions. Moreover, even though these actions

Donald Michie, The Superarticulacy Phenomenon in the Context of Software Manufacture, in The Foundations of Artificial Intelligence, supra note 22, at 411–39. 52   Compare Andrew J. Wu, From Video Games to Artificial Intelligence: Assigning Copyright Ownership to Works Generated by Increasingly Sophisticated Computer Programs, 25 AIPLA Q.J. 131 (1997), with Timothy L. Butler, Can a Computer Be an Author? Copyright Aspects of Artificial Intelligence, 4 Comm. Ent. 707 (1982). 53   The programmer or user should not be held criminally responsible for the autonomous actions of the AI if he or she could not have predicted these actions. See infra sections 3.2 and 3.3. 54   Nicola Lacey & Celia Wells, Reconstructing Criminal Law—Critical Perspectives on Crime and the Criminal Process 53 (2nd ed. 1998). 55   17 U.S.C.S. § 506. See more in Butler, supra note 52; Diane L. Kilpatrick-Lee, Criminal Copyright Law: Preventing a Clear Danger to the U.S. Economy or Clearly Preventing the Original Purpose of Copyright Law? 14 U. Balt. Intell. Prop. L.J. 87 (2005); Eric Goldman, A Road to No Warez: The No Electronic Theft Act and Criminal Copyright Infringement, 82(2) Or. L. Rev. 369 (2003); Hardy, Criminal Copyright Infringement, 11(1) Wm & Mary Bill of Rts. J. 305 (2002–3).

BARFIELD PRINT.indd 402

13/11/2018 11:53

Criminal liability of AI entities  403 were done based on the code written by the programmer, still the programmer did not intend for a violation to occur. Another example of an AI entity engaging in criminal activity is AI software designed to detect threats from the internet and to protect a computer system from these threats. Perhaps a few days after the software is activated, it determines that the most efficient way to detect such threats is by entering websites it defines as dangerous and, once in the website, destroying any software it recognizes as a threat. When the software does this action, it is committing a computer offense, although the programmer did not intend for the AI entity to do so. In these examples, the first model considered—the perpetration-by-another model—is not “legally suitable” for assigning liability given the actions of the AI entity and the lack of knowledge by the human in the system. The perpetration-by-another model assumes mens rea, the criminal intent of the programmers or users to commit an offense via the instrumental use of some of the AI entity’s capabilities. However, this is not the legal situation in the above examples. In these cases, the programmers or users had no prior knowledge that an offense would be committed; they had not planned it, and had not intended to commit the offense using the AI entity. For such cases, the natural-probableconsequence responsibility model might create a suitable legal response. This model is based upon the ability of the programmers or users to foresee the potential commission of offenses. According to the natural-probable-consequence responsibility model, a person might be held accountable for an offense if that offense is a natural and probable consequence of that person’s conduct. Originally, the natural-probable-consequence responsibility model was used to impose criminal responsibility upon accomplices, that is, when one committed an offense which had not been planned by all and which was not part of a conspiracy. For this scenario, the established rule prescribed by courts and commentators is that accomplice responsibility extends to acts of a perpetrator that were a “natural and probable consequence”56 of a criminal scheme that the accomplice encouraged or aided.57 The natural-probable-consequence responsibility model has been widely accepted in accomplice responsibility statutes and recodifications and could be applicable to the actions of autonomous virtual avatars in virtual environments.58 The natural-probable-consequence responsibility model seems to be legally suitable for situations in which an AI entity committed an offense, while the programmer or user had no a priori knowledge of it, had not intended it, and had not participated in it. Further, the natural-probable-consequence responsibility model requires the programmer or user to be in a mental state of negligence, no more. In this model, programmers or users are not required to definitively know about any forthcoming commission of an offense as a

  United States v. Powell, 929 F.2d 724 (D.C. Cir. 1991).   William L. Clark & William L. Marshall, A Treatise on the Law of Crimes 529 (7th ed. 1967); Francis Bowes Sayre, Criminal Responsibility for the Acts of Another, 43 Harv. L. Rev. 689 (1930); People v. Prettyman, 926 P.2d 1013 (Cal. 1996); Chance v. State, 685 A.2d 351 (Del. 1996); Ingram v. United States, 592 A.2d 992 (D.C. 1991); Richardson v. State, 697 N.E.2d 462 (Ind. 1998); Mitchell v. State, 971 P.2d 813 (Nev. 1998); State v. Carrasco, 928 P.2d 939 (N.M. Ct. App. 1996); State v. Jackson, 976 P.2d 1229 (Wash. 1999). 58   State v. Kaiser, 918 P.2d 629 (Kan. 1996); United States v. Andrews, 75 F.3d 552 (9th Cir. 1996). 56 57

BARFIELD PRINT.indd 403

13/11/2018 11:53

404  Research handbook on the law of virtual and augmented reality result of their activity, but are required to know that such an offense is a natural, probable consequence of their actions. In a criminal context, a negligent person has no knowledge of the offense; however, a reasonable person in the same situation would have known about the possibility of an offense, because it is a natural and probable consequence of the situation.59 The programmers or users of an AI entity, those who should have known about the probability of the forthcoming commission of the specific offense, are criminally responsible for the specific offense, even though they did not actually know the offense would occur. This is the fundamental legal basis for criminal responsibility in negligence cases. Negligence is, in fact, an omission of awareness or knowledge. The negligent person omitted knowledge, not acts. The natural-probable-consequence responsibility model would permit responsibility to be predicated upon negligence, even when the specific offense requires a different state of mind.60 This has been accepted in the modern criminal law, and significantly reduces the mental element requirements in these situations, since the relevant accomplice did not really know about the offense but a reasonable person could have predicted it. Thus negligence is suitable for this class of activities. This point is not valid in relation to the person who personally committed the offense, but rather is considered valid in relation to the person who was not the actual perpetrator of the offense but was one of its intellectual forebears. Reasonable programmers or users should have foreseen the offense and prevented it being committed by the AI entity. However, the legal results of applying the natural-probable-consequence responsibility model to the programmer or user vary in two different types of factual cases. The first type of case is when the programmers or users were negligent while programming or using the AI entity and had no criminal intent to commit any offense. The second type of case is when the programmers or users programmed or used the AI entity knowingly and willfully in order to commit one offense via the AI entity, but the AI entity deviated from the plan and committed some other offense, in addition to or instead of the planned offense. The first type of case is a pure case of negligence. The programmers or users acted or omitted negligently; therefore, they should be held accountable for an offense of negligence, if there is such an offense in the specific legal system. Thus, as in the above example, where a programmer of an AI browser negligently programmed it to browse the internet with no restrictions on the IP rights, the programmer is negligent and responsible for the IP offense. Consequently, if a negligence IP offense exists as a specific offense in the relevant legal system, it is the most severe offense for which the programmer may be held accountable, as opposed to most IP offenses, which require at least knowledge or general intent. The second type of case resembles the basic idea of the natural-probable-consequence responsibility model in accomplice responsibility cases. The danger of the very associa59   Robert P. Fine & Gary M. Cohen, Comment, Is Criminal Negligence a Defensible Basis for Penal Liability? 16 Buff. L. Rev. 749 (1966); Herbert L.A. Hart, Negligence, Mens Rea and Criminal Responsibility, in Oxford Essays in Jurisprudence 29 (Anthony Gordon Guest ed. 1961); Donald Stuart, Mens Rea, Negligence and Attempts, Crim. L.R. 647 (1968). 60   Model Penal Code—Official Draft and Explanatory Notes sec. 2.06, pp. 31–2 (1985) (hereinafter “Model Penal Code”); State v. Linscott, 520 A.2d 1067 (Me. 1987).

BARFIELD PRINT.indd 404

13/11/2018 11:53

Criminal liability of AI entities  405 tion or conspiracy whose aim is to commit an offense is the legal reason for more severe accountability to be imposed upon the cohorts. For example, a programmer programs an AI entity to commit an IP offense of copying certain information (music), but the programmer did not program the AI entity to copy other information (video). During the execution of the planned IP offense, the AI entity copies video information as well. In such cases, the criminal negligence responsibility alone is insufficient. The social danger posed by such situations far exceeds the situations in which negligence is sufficient for the offense (through retribution, deterrence, rehabilitation, and incapacitation). As a result, according to the natural-probable-consequence responsibility model, if the programmers or users knowingly and willfully used the AI entity to commit an offense (of one kind) and if the AI entity deviated from the plan by committing another offense, in addition to or instead of the planned offense, the programmers or users should be held accountable for the additional offense, as if it had been committed knowingly and willfully. In the above example, the programmer shall be held criminally accountable for both offenses, planned and unplanned. The question still remains: What is the criminal responsibility of the AI entity itself when the natural-probable-consequence responsibility model is applied? In fact, there are two possible outcomes. If the AI entity acted as an innocent agent, without knowing anything about the criminal prohibition, it is not held criminally accountable for the offense it committed. Under such circumstances, the actions of the AI entity were not different from the actions of the AI entity under the first model (the perpetration-by-another responsibility model). However, if the AI entity did not act merely as an innocent agent, then, in addition to the criminal responsibility of the programmer or user pursuant to the natural-probableconsequence responsibility model, the AI entity itself should be held criminally responsible for the specific offense directly. The direct responsibility model of AI entities, the third model discussed in this chapter, is suitable for this outcome, and is described next. 3.3  Direct Responsibility The direct responsibility model does not assume any dependence of the AI entity on a specific programmer or user in the commission of an offense. It focuses on the activities of the AI entity itself.61 As discussed previously, criminal responsibility for a specific offense is mainly comprised of the factual element (actus reus) and the mental element (mens rea) of that offense. Any person successfully attributed with both elements of the specific offense is held criminally accountable for that specific offense. No other criteria are required in order to impose criminal responsibility. A person might possess further capabilities, but, in order to impose criminal responsibility, the existence of the factual element and the mental element required to impose responsibility for the specific offense is quite enough. 61   Cf, e.g., Steven J. Frank, Tort Adjudication and the Emergence of Artificial Intelligence Software, 21 Suffolk U. L. Rev. 623 (1987); Sam N. Lehman-Wilzig, Frankenstein Unbound: Towards a Legal Definition of Artificial Intelligence, 13 Futures 442 (1981); Marguerite E. Gerstner, Liability Issues with Artificial Intelligence Software, 33 Santa Clara L. Rev. 239 (1993); Richard E. Susskind, Expert Systems in Law: A Jurisprudential Approach to Artificial Intelligence and Legal Reasoning, 49 Mod. L. Rev. 168 (1986).

BARFIELD PRINT.indd 405

13/11/2018 11:53

406  Research handbook on the law of virtual and augmented reality In order to impose criminal responsibility on any kind of entity, the existence of these elements in the specific entity must be proven. Generally, when it has been proven that a person committed the offense in question with knowledge or intent, that person is held criminally responsible for that offense. The criminal responsibility of AI entities depends upon the following questions: How can these entities fulfill the requirements of criminal responsibility? Do AI entities differ from humans in this context? Clearly, the criminal law will be challenged by activities occurring in virtual and augmented reality—where incidents of theft and fraud, along with assault and bullying, are on the rise. But if such activities are the result of virtual avatars operating independently of a human or corporation, who is responsible? Currently, accusations of physical assaults on avatars are mainly handled within the virtual world— meaning the players patrol criminal behavior—though these cases too are starting to get attention from real-world police. An AI algorithm might have numerous features and qualifications far exceeding those of an average human, such as higher velocity of data processing (thinking), the ability to take into consideration many more factors in a shorter time span, and so on. Nevertheless, such features or qualifications are not required in order to impose criminal responsibility. Thus, while they do not negate criminal responsibility, they are not required for the imposition of criminal responsibility. When a human or corporation fulfills the requirements of both the factual element and the mental element, criminal responsibility is imposed. If an AI entity is capable of fulfilling the requirements of both the factual element and the mental element, and, in fact, actually fulfills them, there is nothing to prevent criminal responsibility from being imposed on that AI entity. Generally, the fulfillment of the factual element requirement of an offense is easily attributed to AI entities. As long as an AI entity controls a mechanical or other mechanism that moves its parts, any act might be considered as performed by the AI entity. Thus, when an AI entity activates its electric or hydraulic arm and moves it, this might be considered an act, if the specific offense involves such an act. For example, in the specific offense of assault, such an electric or hydraulic movement of an AI entity that hits a person standing nearby is considered to fulfill the actus reus requirement of the offense of assault. When an offense might be committed due to an omission, the legal analysis is even simpler. Under this scenario, the AI entity is not required to act at all. Its very inaction is the legal basis for criminal responsibility, as long as there had been a duty to act. If a duty to act is imposed upon the AI entity, and it fails to act, the actus reus requirement of the specific offense is fulfilled by way of an omission. Consider an avatar that is programmed to close a switch to avoid a deleterious consequence, but that fails to do so; its omission may fulfill the actus reus requirement for a particular offense. The attribution of the mental element of offenses to AI entities is the real legal challenge to established law in most cases. The attribution of the mental element differs from one AI technology to another. Most cognitive capabilities developed in modern AI technology, such as creativity, are immaterial to the question of the imposition of criminal responsibility. The only cognitive capability required for the imposition of criminal liability is embodied within the mental element requirement (mens rea). Creativity is a human feature that some animals also have, but creativity is a not a requirement for imposing criminal responsibility. Even the most uncreative persons may be held criminally responsible. The sole mental requirements needed in order to impose criminal responsibility are knowledge, intent, and negligence, as required by the specific offense and under the general theory

BARFIELD PRINT.indd 406

13/11/2018 11:53

Criminal liability of AI entities  407 of criminal law. As a result, AI entities do not have to create the idea of committing the specific offense, but in order to be criminally responsible they have only to commit the specific offense with knowledge as to the factual elements of that offense. Knowledge is defined as sensory reception of factual data and the understanding of that data.62 Most AI systems are well equipped for such reception. Sensory receptors of sights, voices, physical contact, touch, and so on are common in most AI systems. These receptors transfer the factual data received to central processing units that analyze the data. The process of analysis in AI systems parallels that of human understanding.63 The human brain understands the data received by eyes, ears, hands, and so on, by analyzing that data. Advanced AI algorithms are trying in many cases to imitate human cognitive processes. These processes are not so different.64 Specific intent is the strongest of the mental element requirements for a criminal offense.65 Specific intent is the existence of a purpose or an aim that a factual event will occur.66 The specific intent required to establish responsibility for murder is a purpose or an aim that a certain person will die.67 As a result of the existence of such intent, the perpetrator of the offense commits the offense; that is, he performs the factual element of the specific offense. This situation is not unique to humans. Some AI entities might be programmed to determine on their own a purpose or an aim and to take actions in order to achieve that self-derived purpose, and some advanced AI entities may determine for themselves a purpose and take relevant actions in order to achieve that purpose. In both cases this might be considered as specific intent, since the AI entity determined its own purpose and determined of its own volition the relevant actions in order to achieve that purpose. One might assert that many crimes are committed as a result of strong emotions or feelings (heat of passion crimes) that cannot be imitated by AI software such as a virtual

62   William James, The Principles of Psychology (1890); Hermann von Helmholtz, The Facts of Perception (1878). In this context knowledge and awareness are identical. See, e.g., United States v. Youts, 229 F.3d 1312 (10th Cir. 2000); State v. Sargent, 594 A.2d 401 (Vt. 1991); United States v. Spinney, 65 F.3d 231 (1st Cir. 1995); State v. Wyatt, 482 S.E.2d 147 (W. Va. 1996); United States v. Wert-Ruiz, 228 F.3d 250 (3rd Cir. 2000); United States v. Jewell, 532 F.2d 697 (9th Cir. 1976); United States v. Ladish Malting Co., 135 F.3d 484 (7th Cir. 1998); Model Penal Code, supra note 60, § 2.02(2)(b). 63   Margaret A. Boden, Has AI Helped Psychology? in The Foundations of Artificial Intelligence, supra note 22, at 108; Derek Partridge, What’s in an AI Program? in The Foundations of Artificial Intelligence, supra note 22, at 112; David Marr, AI: A Personal View, in The Foundations of Artificial Intelligence, supra note 22, at 97. 64   Daniel C. Dennett, Evolution, Error, and Intentionality, in The Foundations of Artificial Intelligence, supra note 22, at 190; B. Chandrasekaran, What Kind of Information Processing is Intelligence? in The Foundations of Artificial Intelligence, supra note 22, at 14. 65   Robert Batey, Judicial Exploration of Mens Rea Confusion, at Common Law and Under the Model Penal Code, 18 Ga. St. U. L. Rev. 341 (2001); State v. Daniels, 109 So. 2d 896 (La. 1958); Carter v. United States, 530 U.S. 255 (2000); United States v. Randolph, 93 F.3d 656 (9th Cir. 1996); United States v. Torres, 977 F.2d 321 (7th Cir. 1992); Frey v. State, 708 So. 2d 918 (Fla. 1998); State v. Neuzil, 589 N.W.2d 708 (Iowa 1999); People v. Disimone, 650 N.W.2d 436 (Mich. Ct. App. 2002); People v. Henry, 607 N.W.2d 767 (Mich. Ct. App. 1999). 66   Wayne R. LaFave, Criminal Law 244–9 (4th ed. 2003). 67   For the intent-to-kill murder see id at 733–4.

BARFIELD PRINT.indd 407

13/11/2018 11:53

408  Research handbook on the law of virtual and augmented reality avatar, not even by the most advanced software. Such feelings include love, affection, hatred, jealousy, and so on. This observation might be accurate in relation to the AI technology of the beginning of the twenty-first century, but advances in creative AI entities and with emotions are rapidly occurring. Even so, such feelings are rarely required in specific offenses. Most specific offenses are satisfied by knowledge of the existence of the external element. Few offenses require specific intent in addition to knowledge. Almost all other offenses are satisfied by much less than that (negligence, recklessness, strict liability). Perhaps in a very few specific offenses that do require certain feelings (such as crimes of racism or hate),68 criminal responsibility cannot be imposed upon AI entities, which have no such feelings; however, in any other specific offense, lack of certain feelings is not a barrier to imposing criminal liability. In most legal systems IP offenses require no specific intent; knowledge alone is their mental element requirement. If a person fulfills the requirements of both the factual element and the mental element of a specific offense, then the person is held criminally responsible. Why should an AI entity operating within a virtual world that fulfills all elements of an offense be exempt from criminal responsibility? One might argue that some segments of human society are exempt from criminal responsibility even if both the external and internal elements have been established. Such segments of society include infants and the mentally ill—should AI entities form a third category of exemption for criminal liability? Again, specific provisions in criminal law exempt infants from criminal responsibility.69 The social rationale behind the infancy defense is to protect infants from the harmful consequences of the criminal process and to handle them in other social frameworks.70 Do such frameworks exist for AI entities? The original legal rationale behind the infancy defense was the fact that infants are as yet incapable of comprehending what was wrong in their conduct (doli incapax).71 However, children can be held criminally responsible

68   See, e.g., Elizabeth A. Boyd et al, “Motivated by Hatred or Prejudice”: Categorization of Hate-Motivated Crimes in Two Police Divisions, 30 Law & Soc’y Rev. 819 (1996); Crimes Motivated by Hatred: The Constitutionality and Impact of Hate Crimes Legislation in the United States, 1 Syracuse J. Legis. & Pol’y 29 (1995). 69   See, e.g., Minn. Stat. § 9913 (1927); Mont. Rev. Code § 10729 (1935); N.Y. Penal Law § 816 (1935); Okla. Stat. § 152 (1937); Utah Rev. Stat. 103-i-40 (1933); State v. George, 54 A. 745 (Del. 1902); Heilman v. Commonwealth, 1 S.W. 731 (Ky. 1886); State v. Aaron, 4 N.J.L. 269 (Sup. Ct. 1818); McCormack v. State, 15 So. 438 (Ala. 1894); Little v. State, 554 S.W.2d 312 (Ark. 1977); Clay v. State, 196 So. 462 (Fla. 1940); In re Devon T., 584 A.2d 1287 (Md. Ct. Spec. App. 1991); State v. Dillon, 471 P.2d 553 (Idaho 1970); State v. Jackson, 142 S.W.2d 45 (Mo. 1940). 70   Frederick J. Ludwig, Rationale of Responsibility for Young Offenders, 29 Neb. L. Rev. 521 (1950); In re Tyvonne, 558 A.2d 661 (Conn. 1989); Andrew Walkover, The Infancy Defense in the New Juvenile Court, 31 UCLA L. Rev. 503 (1984); D. Keith Foren, Note, In Re Tyvonne M. Revisited: The Criminal Infancy Defense in Connecticut, 18 Quinnipiac L. Rev. 733 (1999); Michael Tonry, Rethinking Unthinkable Punishment Policies in America, 46 UCLA L. Rev. 1751 (1999); Andrew Ashworth, Sentencing Young Offenders, in Principled Sentencing: Readings on Theory and Policy 294, 294–306 (Andrew von Hirsch et al, eds, 3rd ed. 2009); Franklin E. Zimring, Rationales for Distinctive Penal Policies for Youth Offenders, in Principled Sentencing: Readings on Theory and Policy 316; Andrew von Hirsch, Reduced Penalties for Juveniles: The Normative Dimension, in Principled Sentencing: Readings on Theory and Policy 323. 71   Sir Edward Coke, The Third Part of the Institutes of the Laws of England 4 (6th ed. 2001) (1681).

BARFIELD PRINT.indd 408

13/11/2018 11:53

Criminal liability of AI entities  409 if the presumption of mental incapacity was refuted by proof that the child was able to distinguish between right and wrong, that is, between permitted and forbidden actions in terms of criminal offenses.72 Could such logic be similarly applied to AI entities? Most AI algorithms are capable of distinguishing between permitted and forbidden actions, and therefore they may distinguish between a conduct which is a commission of an offense (forbidden) or one which is not (permitted).73 Similarly, mentally ill people are presumed to lack the fault element of the specific offense, due to their mental illness (doli incapax).74 Mentally ill people are considered unable to distinguish between right and wrong (that is, they lack the cognitive capabilities possessed by normal persons)75 and to control impulsive behavior.76 When an AI algorithm functions properly, there is no reason for it not to use all of its capabilities to analyze the factual data received through its receptors. However, an interesting legal question would be whether a defense of insanity might be raised in relation to a malfunctioning AI algorithm when its analytical capabilities become corrupted as a result of that malfunction (of course, issues of standing and other rules of civil procedure will have to be considered for AI entities). When an AI entity establishes all elements of a specific offense, both factual and mental, there is no reason to prevent imposition of criminal responsibility upon it for that offense. The criminal responsibility of an AI entity does not replace the criminal responsibility of the programmers or the users, if criminal responsibility is imposed on the programmers and/or users by any other legal path. Criminal responsibility is not to be divided, but rather added. The criminal responsibility of the AI entity is imposed in addition to the criminal responsibility of the human programmer or user. However, the criminal responsibility of an AI entity is not dependent upon the criminal responsibility of the programmer or user of that AI entity. As a result, if the specific AI entity was programmed or directed by another AI entity, the criminal responsibility of the programmed or used AI entity is not influenced by that fact. The programmed or used AI entity shall be held criminally accountable for the specific offense pursuant to the direct responsibility model, unless it was an innocent agent. In addition, the programmer or user of the AI entity shall be held criminally accountable for that very offense pursuant to one 72   Sir Matthew Hale, Historia Placitorum Coronae: The History of the Pleas of the Crown 23, 26 (1736); McCormack v. State, 15 So. 438 (Ala. 1894); Little v. State, 554 S.W.2d 312 (Ark. 1977); In re Devon T., 584 A.2d 1287 (Md. Ct. Spec. App. 1991). 73   Padhy, supra note 6, at 277–333. 74   See, e.g., Benjamin B. Sendor, Crime as Communication: An Interpretive Theory of the Insanity Defense and the Mental Elements of Crime, 74 Geo. L. J. 1371, 1380 (1986); Joseph H. Rodriguez, Laura M. LeWinn, & Michael L. Perlin, The Insanity Defense under Siege: Legislative Assaults and Legal Rejoinders, 14 Rutgers L. J. 397, 406–7 (1983); Homer D. Crotty, The History of Insanity as a Defence to Crime in English Criminal Law, 12 Cal. L. Rev. 105 (1924). 75   See, e.g., Edward de Grazia, The Distinction of Being Mad, 22 U. Chi. L. Rev. 339 (1955); Warren P. Hill, The Psychological Realism of Thurman Arnold, 22 U. Chi. L. Rev. 377 (1955); Manfred S. Guttmacher, The Psychiatrist as an Expert Witness, 22 U. Chi. L. Rev. 325 (1955); Wilber G. Katz, Law, Psychiatry, and Free Will, 22 U. Chi. L. Rev. 397 (1955); Jerome Hall, Psychiatry and Criminal Responsibility, 65 Yale L. J. 761 (1956). 76   See, e.g., John Barker Waite, Irresistible Impulse and Criminal Liability, 23 Mich. L. Rev. 443, 454 (1925); Edward D. Hoedemaker, “Irresistible Impulse” as a Defense in Criminal Law, 23 Wash. L. Rev. 1, 7 (1948).

BARFIELD PRINT.indd 409

13/11/2018 11:53

410  Research handbook on the law of virtual and augmented reality of the three responsibility models, according to its specific role in the offense. The chain of criminal responsibility might continue if more parties are involved, whether they are human or AI entities. There is no reason to eliminate the criminal responsibility of an AI entity or of a human, which is based on complicity between them. An AI entity and a human might cooperate as joint perpetrators, as accessories and abettors, and so on, and the relevant criminal responsibility might be imposed on them accordingly. Since the factual and mental capabilities of an AI entity are sufficient to impose criminal responsibility on it, if these capabilities satisfy the legal requirements of joint perpetrators, accessories and abettors, and so on, then the relevant criminal responsibility as joint perpetrators, accessories and abettors, and so on, should be imposed, regardless of whether the offender is an AI entity or a human. Not only may positive factual and mental elements be attributed to AI entities; all relevant negative fault elements are attributable to AI entities too. Most of these elements are expressed by the general defenses in criminal law, for example, self-defense, necessity, duress, intoxication, and so on. For some of these defenses (justifications)77 there is no material difference between humans and AI entities, since they relate to a specific situation (in rem), regardless of the identity of the offender. For example, an AI entity serving under the local police force is given an order to arrest a person, and unbeknown to the AI entity this order is illegal. If the executer is unaware, or could not reasonably become aware, that an otherwise legal action is illegal in this specific instance, the executor of the order is not criminally responsible.78 In that case, there is no discernable difference whether the executor is human or an AI entity. For other defenses (excuses and exemptions),79 some applications should be adjusted. For example, the intoxication defense is applied when the offender is under the physical influence of an intoxicating substance, such as alcohol or drugs. The influence of alcohol on an AI entity would be nonexistent, but the influence of an electronic virus that is infecting the operating system of the AI entity might be considered parallel to the influence of intoxicating substances on humans. Some other factors might be considered as being parallel to insanity or loss of control. It might be summarized that the criminal responsibility of an AI entity according to the direct responsibility model is not different from the relevant criminal responsibility

77   John C. Smith, Justification and Excuse in the Criminal Law (1989); Anthony M. Dillof, Unraveling Unknowing Justification, 77 Notre Dame L. Rev. 1547 (2002); Kent Greenawalt, Distinguishing Justifications from Excuses, 49 Law & Contemp. Probs. 89 (1986); Kent Greenawalt, The Perplexing Borders of Justification and Excuse, 84 Colum. L. Rev. 1897 (1984); Thomas Morawetz, Reconstructing the Criminal Defenses: The Significance of Justification, 77 J. Crim. L. & Criminology 277 (1986); Paul H. Robinson, A Theory of Justification: Societal Harm as a Prerequisite for Criminal Liability, 23 UCLA L. Rev. 266 (1975); Paul H. Robinson & John M. Darley, Testing Competing Theories of Justification, 76 N.C. L. Rev. 1095 (1998). 78   Michael A. Musmanno, Are Subordinate Officials Penally Responsible for Obeying Superior Orders which Direct Commission of Crime? 67 Dick. L. Rev. 221 (1963). 79   Peter Arenella, Convicting the Morally Blameless: Reassessing the Relationship Between Legal and Moral Accountability, 39 UCLA L. Rev. 1511 (1992); Sanford H. Kadish, Excusing Crime, 75 Cal. L. Rev. 257 (1987); Andrew E. Lelling, A Psychological Critique of Character-Based Theories of Criminal Excuse, 49 Syracuse L. Rev. 35 (1998).

BARFIELD PRINT.indd 410

13/11/2018 11:53

Criminal liability of AI entities  411 of a human. In some cases, some adjustments are necessary, but substantively it is the very same criminal responsibility, which is based upon the same elements and examined in the same ways. 3.4 Combination The responsibility models described above are not competing models. These models might be applied in combination in order to create a full image of criminal responsibility in the specific context of AI entity involvement. In fact, none of the possible models are mutually exclusive. Thus, applying the second model is possible as a single model for the specific offense, and it is possible as one part of a combination of two of the legal models or all three of them. When the AI entity plays the role of an innocent agent in the perpetration of a specific offense, and the programmer is the only person who directed that perpetration, the application of the perpetration-by-another responsibility model (the first responsibility model introduced) is the most appropriate legal model for that situation. In that same situation, when the programmer is itself an AI entity (when an AI entity programs another AI entity to commit a specific offense), the direct responsibility model (the third model) is most appropriately applied to the criminal responsibility of the AI entity’s programmer. The third responsibility model in the above situation is applied to the programmer in addition to the first model, and not in lieu thereof. Thus, in such situations, the AI entity programmer should be criminally responsible, pursuant to a combination of the perpetration-by-another responsibility model and the direct responsibility model. If the AI entity plays the role of the physical perpetrator of the specific offense, but that very offense was not planned to be perpetrated, then the application of the naturalprobable-consequence responsibility model might be appropriate. The programmer might be deemed negligent if no offense had been deliberately planned to be perpetrated. Similarly, the programmer might be held fully accountable for that specific offense if another offense had indeed been deliberately planned, but the specific offense that was perpetrated had not been part of the original criminal scheme. Nevertheless, when the programmer is not human, the direct responsibility model must be applied in addition to the simultaneous application of the natural-probable-consequence responsibility model; the same is the case when the physical perpetrator is human while the planner is an AI entity. The coordination of all three responsibility models creates a dense network of criminal liability which minimizes opportunities to escape from the imposition of criminal responsibility. The combined and coordinated application of these possible models reveals a new legal situation in the specific context of AI entities and criminal law. As a result, when AI entities and humans are involved, directly or indirectly, in the perpetration of a specific offense, it will be far more difficult to evade criminal responsibility. The social benefit to be derived from such a legal policy is of substantial value. All entities—human, legal, or AI—become subject to criminal law. If the clearest purpose of the imposition of criminal responsibility is the application of legal social control in a specific society, then the coordinated application of all possible models is necessary in the very context of AI entities.

BARFIELD PRINT.indd 411

13/11/2018 11:53

412  Research handbook on the law of virtual and augmented reality

4.  PUNISHING AI Let us assume arguendo that an AI entity is indicted, tried, and convicted. After the conviction, the court is required to sentence that AI entity.80 The most appropriate punishments under IP law are fines and imprisonment.81 The question is: How can an AI entity practically serve a sentence which includes ingredients of a fine (who pays it? an AI system does not independently have money) or imprisonment (the AI system does not necessarily have a tangible appearance)? In instances where there is no physical body to arrest, especially in cases of AI software that was not installed in a physical body (such as a robot), what is the practical meaning of incarceration? Further, where no bank account is available for the sentenced AI entity, what is the practical significance of fining it? Should the punishment be imposed within the virtual reality environment or outside it? A basic question toward punishing such software is: Why would we not disable the computer or reprogram or redesign it in a way that assures the software future conduct does not repeat its criminal conduct? First, disabling any hardware does not necessarily effect the software, as the software may be set on a network or on a cloud. Second, reprogramming any entity with the capability of AI (a learning machine) is not really effective in the long run. The machine would probably learn how to bypass the redesign with no special difficulty, given the required time and technological skills that may be taken from the network. Thus, the above questions still stand and require solutions, both within the virtual reality environment and outside it. A preliminary question toward punishing AI entities relates to the general purposes of punishment, that is, retribution, deterrence, rehabilitation, and incapacitation.82 Are these general purposes of punishment relevant to nonhuman entities? From the above four purposes of punishment, which are relevant to artificial intelligence technology? Retribution is aimed to satisfy society more than it is purposed to the offender. Causing suffering to the offender, itself, has no prospective value. The suffering may deter the offender, but that is part of the general purpose of deterrence, not retribution. Retribution may supply some catharsis to society and victims through causing suffering to the offender. But punishing machines through retribution, in this context, would be meaningless and impractical. 80   For an overview of the criminal process see Gabriel Hallevy, The Defense Attorney as Mediator in Plea Bargains, 9 Pepp. Disp. Resol. L.J. 495 (2009). 81   See, e.g., 17 U.S.C.S. § 506, House Report No 94-1476: “Four types of criminal offenses actionable under the bill are listed in section 506: willful infringement for profit, fraudulent use of a copyright notice, fraudulent removal of notice, and false representation in connection with a copyright application. The maximum fine on conviction has been increased to $10,000 and, in conformity with the general pattern of the Criminal Code (18 U.S.C.), no minimum fines have been provided. In addition to or instead of a fine, conviction for criminal infringement under section 506(a) can carry with it a sentence of imprisonment of up to one year. Section 506(b) deals with seizure, forfeiture, and destruction of material involved in cases of criminal infringement . . . Section 506(a) contains a special provision applying to any person who infringes willfully and for purposes of commercial advantage the copyright in a sound recording or a motion picture. For the first such offense a person shall be fined not more than $25,000 or imprisoned for not more than one year, or both. For any subsequent offense a person shall be fined not more than $50,000 or imprisoned not more than two years, or both.” 82   For these general purposes of punishment see Gabriel Hallevy, The Right to Be Punished: Modern Doctrinal Sentencing 15–56 (2013).

BARFIELD PRINT.indd 412

13/11/2018 11:53

Criminal liability of AI entities  413 Some people, when they are in a hurry and their car cannot start, get angry. In their anger they may hit the car, kick, it or even shout at it. Punishing machines—any machine, from cars to highly sophisticated artificial intelligence robots, or even a virtual avatar—for retribution would not be different than kicking a car. It may ease some people’s anger, but not more than that. A machine does not suffer, and as long as retribution is based on suffering, retribution would not be very relevant to punishing intelligent avatars. This is true for both classic and modern (“just desert”) approaches of retribution. Moreover, functionally, if retribution functions as a lenient factor of sentencing in order to prevent revenge, it just strengthens the retribution’s irrelevance to artificial intelligence sentencing. Revenge is assumed to cause additional suffering to the offender on top of the official punishment; however, since machines do not experience suffering, the choice between revenge and retribution is meaningless for them. Deterrence is meant to prevent the commission of another offense, through intimidation. For machines, at the moment, intimidation is a feeling they cannot experience. The intimidation itself is based on future suffering imposed in the case of committing the offense. Since machines do not currently experience suffering, as aforesaid, the reason for intimidation, besides the intimidation itself, is also annihilated when considering the appropriate punishment for robots. However, both retribution and deterrence may be relevant punishments for the human participants in the commission of the offense (for example, users and programmers). As to rehabilitation, artificial intelligence systems may experience decisionmaking processes and take decisions that might be seem unreasonable. Sometimes the artificial intelligence system may need external direction in order to refine the decisionmaking process. This may be part of the machine learning process. Rehabilitation functions in exactly the same way for humans and therefore it may be applicable to artificial intelligence systems as well. The rehabilitation of humans causes them to make better decisions in their daily lives, according to society’s point of view. So may the criminal process result in the same outcome for artificial intelligence systems. The punishment, under this approach, would be directed to refine the machine learning process. The artificial intelligence system, after being rehabilitated, would be able to form better and more accurate decisions after adding more limitations to its discretion and refining the process through machine learning. Thus, the punishment, if adjusted correctly to the particular artificial intelligence system, would be part of the machine learning process. Through this process, directed by the rehabilitative punishment, the artificial intelligence system would have better tools to analyze factual data and deal with it. In fact, rehabilitative punishments have the same effect on humans, providing them better tools to face factual reality. Consequently, rehabilitation may be a relevant purpose of punishment for artificial intelligence systems as it is not based on intimidation or suffering, but is directed at the better performance of artificial intelligence systems. For humans, this consideration may be secondary in most cases; however, for artificial intelligence systems it may be a primary purpose of punishment. Nevertheless, rehabilitation is not the only consideration that may be relevant for artificial intelligence systems; incapacitation is relevant as well. If an artificial intelligence system commits offenses during its activation and has no capability of changing its ways through inner changes (for example, through machine learning), only incapacitation may supply an adequate answer. Whether the artificial ­intelligence

BARFIELD PRINT.indd 413

13/11/2018 11:53

414  Research handbook on the law of virtual and augmented reality system understands the meaning of its activity or not, and whether the artificial intelligence system is equipped with proper tools to perform inner changes or not, delinquency must still be prevented. In such situations society must remove the artificial intelligence system’s physical capabilities to commit further offenses. The particular artificial intelligence system must be taken out of the circle of delinquency, regardless of its skills. Substantively, this is what the society does with equivalent cases of human offenders.83 It may be concluded that with artificial intelligence systems the two relevant considerations of punishments are rehabilitation and incapacitation. Both reflect the extreme edges of sentencing, and both serve criminal law purposes for nonhuman offenders. When the artificial intelligence system possesses capabilities to perform inner changes that affect its activity, it seems that rehabilitation would be the relevant consideration rather than incapacitation. However, when the artificial intelligence system does not possess such capabilities, incapacitation would be relevant. Thus, the relevant punishment for the relevant case is adjusted to the personal characteristics of the offender, as is acceptable for human offenders. Similar legal problems were raised when the criminal responsibility of corporations was recognized.84 Some legal scholars asked how any of the legitimate penalties imposed upon humans could be applicable to corporations. The answer was relatively simple. When a punishment can be imposed on a corporation as it is on humans, it is imposed without change. When the court adjudicates a fine, the corporation pays the fine in the same way that a human pays the fine and in the same way that a corporation pays its bills in a civil context. However, when punishment of a corporation cannot be carried out in the same way as with humans, an adjustment is required. Such is the legal situation vis-à-vis AI entities. The punishment adjustment considerations are concentrated on the theoretical foundations of the specific punishment. Thus, in order to impose a fine upon an AI entity, there is a necessity to understand the theoretical foundations of the fine, so that it would be possible to adjust it from humans to AI entities without changing its essence. These adjustment considerations are applied in a similar manner and are composed three stages. Each stage may be explained by a question: What is the fundamental significance of the specific punishment for a human? How does that punishment affect AI entities? What practical penalties may achieve the same significance when imposed on AI entities? The most significant advantage of these punishment adjustment considerations is that the significance of the specific punishment remains identical when imposed on humans 83   Martin P. Kafka, Sex Offending and Sexual Appetite: The Clinical and Theoretical Relevance of Hypersexual Desire, 47 Int’l J. of Offender Therapy and Comparative Criminology 439 (2003); Matthew Jones, Overcoming the Myth of Free Will in Criminal Law: The True Impact of the Genetic Revolution, 52 Duke L. J. 1031 (2003); Sanford H. Kadish, Excusing Crime, 75 Cal. L. Rev. 257 (1987). 84   Gerard E. Lynch, The Role of Criminal Law in Policing Corporate Misconduct, 60 Law & Contemp. Probs. 23 (1997); Richard Gruner, To Let the Punishment Fit the Organization: Sanctioning Corporate Offenders Through Corporate Probation, 16 Am. J. Crim. L. 1 (1988); Steven Walt & William S. Laufer, Why Personhood Doesn’t Matter: Corporate Criminal Liability and Sanctions, 18 Am. J. Crim. L. 263 (1991); Coffee, supra note 19; Steven Box, Power, Crime and Mystification 16–79 (1983); Brent Fisse & John Braithwaite, The Allocation of Responsibility for Corporate Crime: Individualism, Collectivism and Accountability, 11 Sydney L. Rev. 468 (1988).

BARFIELD PRINT.indd 414

13/11/2018 11:53

Criminal liability of AI entities  415 and AI entities. This method of punishment adjustment considerations is referred to below in some of the penalties used in modern societies: capital punishment, incarceration, suspended sentencing, community service, and fines. Incarceration is one of the most popular sentences imposed in Western legal systems for serious crimes,85 and it is mentioned as an appropriate sanction in some IP criminal legislation.86 The significance of incarceration for humans is the deprivation of human liberty and the imposition of severe limitations on human free behavior, including freedom of movement and freedom to manage one’s personal life.87 The “liberty” or “freedom” of an AI entity includes the freedom to act as an AI entity in the relevant area. For example, an AI entity in medical service has the freedom to participate in surgeries, an AI entity in a factory has the freedom to manufacture, and so on.88 Considering the nature of a sentence of incarceration, the practical action that may achieve the same effect as incarceration when imposed on an AI entity is to put the AI entity out of use for a determinate period. During that period, no action relating to the AI entity’s freedom is allowed, and thus its freedom or liberty is restricted. Suspended sentencing is a very popular intermediate sanction in Western legal systems to increase the deterrent effect on offenders in lieu of actual incarceration.89 The significance of a suspended sentence for humans is the very threat of incarceration if the human commits a specific offense or a type of specific offense.90 If the human commits such an offense, a sentence of incarceration will be imposed for the first offense in addition to the sentence for the second offense. As a result, a human is deterred from committing another offense and from becoming a recidivist offender. Practically, a suspended sentence is imposed only in the legal records. No physical action is taken when a suspended sentence is imposed. As a result, in terms of imposing a suspended sentence, there is no difference in effect between humans and AI entities. The 85   See generally John Irwin, Prisons in Turmoil (Little, Brown 1980); Anthony J. Manocchio & Jimmy Dunn, The Time Game: Two Views of a Prison (Sage Publications 1982). 86   See, e.g., supra note 81. 87   See generally David J. Rothman, For the Good of All: The Progressive Tradition in Prison Reform, in History and Crime: Implications for Criminal Justice Policy 271 (James A. Inciardi & Charles E. Faupel eds 1980); Michael Welch, Ironies of Imprisonment (Sage Publications 2004); Roy D. King, The Rise and Rise of Supermax: An American Solution in Search of a Problem? 1 Punishment & Soc’y 163 (1999); Chase Riveland, Supermax Prisons: Overview and General Considerations (1999); Jamie Fellner & Joanne Mariner, Cold Storage: Super-Maximum Security Confinement in Indiana (Human Rights Watch 1997); Richard R. Korn, The Effects of Confinement in the High Security Unit in Lexington, 15 Soc. Just. 8 (1988); Holly A. Miller, Reexamining Psychological Distress in the Current Conditions of Segregation, 1 J. of Correctional Health Care 39 (1994); Frieda Bernstein, The Perception of Characteristics of Total Institutions and Their Effect on Socialization (1979);  Bruno Bettelheim, The Informed Heart: Autonomy in a Mass Age (Free Press 1960); Marek M. Kaminski, Games Prisoners Play: Allocation of Social Roles in a Total Institution, 15 Rationality & Soc’y 188 (2003). 88   Not all AI entities do understand freedom in the same capacity as humans. When the entire internal world of a specific AI entity is focused on its function at work, depriving it of the opportunity to work has the same “social” effect that incarceration has upon humans. 89   A.E. Bottoms, The Suspended Sentence in England 1967–1978, 21 Brit. J. Criminology 1 (1981). 90   Marc Ancel, Suspended Sentence (Heinemann Educational Books 1971); Marc Ancel, The System of Conditional Sentence or Sursis, 80 L. Q. Rev. 334 (1964).

BARFIELD PRINT.indd 415

13/11/2018 11:53

416  Research handbook on the law of virtual and augmented reality statutory criminal records of the state do not differentiate between a suspended sentence imposed on humans and one imposed on corporations or AI entities, as long as the relevant entity may be identified specifically and accurately. Community service is also a very popular intermediate sanction in Western legal systems in lieu of actual incarceration.91 In most legal systems, community service is a substitute for short sentences of actual incarceration. In some legal systems, community service is imposed coupled with probation so that the offender “pays a price” for the damages he caused by committing the specific offense.92 The significance of community service for humans is compulsory contribution of labor to the community. As discussed above, an AI entity can be engaged as a worker in very many areas. When an AI entity works in a factory, its work is done for the benefit of the factory owners or for the benefit of the other workers in order to ease and facilitate their professional tasks. In the same way that an AI entity works for the benefit of private individuals, it may work for the benefit of the community. When work for the benefit of the community is imposed on an AI entity as a compulsory contribution of labor to the community, it may be considered community service. Thus the significance of community service is identical, whether imposed on humans or AI entities. The adjudication of a fine is the most popular intermediate sanction in Western legal systems in lieu of actual incarceration.93 For humans, the significance of paying a fine is deprivation of some of their property, whether that property is money (a fine) or other property (forfeiture).94 When a person fails to pay a fine, or has insufficient property to pay the fine, substitute penalties are imposed on the offender, particularly incarceration.95 The imposition of a fine on a corporation is identical to the imposition of a fine on a person, since both people and corporations have property and bank accounts. However, most AI entities have no money or property of their own, nor have they any bank accounts. If an AI entity does have its own property or money, the imposition of

91   Andrew Willis, Community Service as an Alternative to Imprisonment: A Cautionary View, 24 Probation J. 120 (1977). 92   John Harding, The Development of the Community Service, in Alternative Strategies for Coping with Crime 164 (Norman Tutt ed., Blackwell Publishers 1978); Home Office, Review of Criminal Justice Policy (1977); Julie Leibrich, Burt Galaway, & Yvonne Underhill, Community Sentencing in New Zealand: A Survey of Users, 50 Fed. Probation 55 (1986); James Austin & Barry Krisberg, The Unmet Promise of Alternatives, 28 J. of Res. in Crime & Delinq. 374 (1982); Mark S. Umbreit, Community Service Sentencing: Jail Alternatives or Added Sanction? 45 Fed. Probation 3 (1981). 93   Gerhardt Grebing, The Fine in Comparative Law: A Survey of 21 Countries (1982). 94   See generally Judith A. Greene, Structuring Criminal Fines: Making an “Intermediate Penalty” More Useful and Equitable, 13 Just. Sys. J. 37 (1988); Nigel Walker & Nicola Padfield, Sentencing: Theory, Law and Practice (2nd ed. 1999); Manfred Zuleeg, Criminal Sanctions to Be Imposed on Individuals as Enforcement Instruments in European Competition Law, in European Competition Law Annual 2001: Effective Private Enforcement of EC Antitrust Law 451 (Claus-Dieter Ehlermann & Isabela Atanasiu eds 2001); Steve Uglow, Criminal Justice (Sweet & Maxwell 1995); Douglas C. McDonald, Judith A. Greene, & Charles Worzella, Day-Fines in American Courts: The Staten Island and Milwaukee Experiments (National Institute of Justice 1992). 95   Fiori Rinaldi, Imprisonment for Non-Payment of Fines (1976); Scottish Advisory Council on the Treatment & Rehabilitation of Offenders, Use of Short Sentences of Imprisonment by the Court (1960).

BARFIELD PRINT.indd 416

13/11/2018 11:53

Criminal liability of AI entities  417 a fine on it would be identical to the imposition of a fine on humans or corporations. For most humans and corporations, property is gained through labor.96 When paying a fine, the property, which is a result of labor, is transferred to the state. That labor might be transferred to the state in the form of property or directly as labor. As a result, a fine imposed on an AI entity might be collected as money, property, or labor for the benefit of the community. In virtual reality the fine may be payback of bitcoin currency, perhaps for a task performed within a virtual world. When the fine is collected in the form of labor for the benefit of the community, it is not different from community service as described above. Thus, most common penalties are applicable to AI entities. The imposition of specific penalties on AI entities does not negate the nature of these penalties in comparison with their imposition on humans. Of course, some general punishment adjustment considerations are necessary in order to apply these penalties, but still, the nature of these penalties remains the same relative to humans and to AI entities.

5. CONCLUSION If all of the specific requirements are met, criminal responsibility may be imposed upon any entity, whether human, corporate, or AI. Modern times warrant modern legal measures in order to resolve today’s legal problems. The rapid development of AI technology as used in virtual environments requires current legal solutions in order to protect society from possible dangers inherent in technologies not subject to the law, especially criminal law. Criminal law has a very important social function—that of preserving social order for the benefit and welfare of society. Threats to that social order may be posed by humans, corporations, or AI entities. Humans have been subject to criminal law, except when otherwise decided by international consensus. Thus, minors and mentally ill persons are not subject to criminal law in most legal systems around the world. Although corporations in their modern form have existed since the fourteenth century,97 it took hundreds of years to subordinate corporations to the law, and especially to criminal law. For hundreds of years, the law stated that corporations are not subject to criminal law, as inspired by Roman law (societas delinquere non potest—corporations cannot perform delinquency).98 It was only in the seventeenth century that an English court was progressive enough to impose criminal responsibility on a corporation.99 It was inevitable. Corporations participate fully in human life, and it was a major disadvantage not to

  John Locke, Two Treatises of Government (1689).   William Searle Holdsworth, A History of English Law 471–6 (1923). 98   W.S. Holdsworth, English Corporation Law in the 16th and 17th Centuries, 31 Yale L. J. 382 (1922); William Robert Scott, The Constitution and Finance of English, Scottish and Irish Joint-Stock Companies to 1720 462 (1912); Bishop Carleton Hunt, The Development of the Business Corporation in England, 1800–1867 6 (1936). 99   Case of Langforth Bridge (1634) 79 Eng. Rep. 919 (K.B.); see also The King v. Inhabitants of Clifton, (1794) 101 Eng. Rep. 280 (K.B.); Rex v. Inhabitants of Great Broughton, (1771) 98 Eng. Rep. 418 (K.B.); Stratford-upon-Avon Corporation, (1811) 104 Eng. Rep. 636 (K.B.); Liverpool (Mayor), (1802) 102 Eng. Rep. 529 (K.B.); Reg v. Saintiff, (1705), 87 Eng. Rep. 1002 (K.B.). 96 97

BARFIELD PRINT.indd 417

13/11/2018 11:53

418  Research handbook on the law of virtual and augmented reality subject them to human laws, since offenses are committed by corporations or through them. But corporations have neither body nor soul. Legal solutions were developed so that in relation to criminal responsibility, corporations would be deemed capable of fulfilling all the requirements of criminal responsibility, including factual elements and mental elements.100 These solutions were embodied in models of criminal responsibility and general punishment adjustment considerations. It worked. In fact, it is still working, and very successfully. Why should AI entities be different from corporations? Although corporations are not human entities, the effects and imposition of punishments on a corporation are still felt by human beings, which leads to deterrence. This is also the situation with AI entities when they are treated as criminally responsible, since no legal difference is found between humans and AI entities as to the imposition of criminal liability and the imposition of punishments. When AI entities are accepted as legal entities, all effects of criminal law shall be identical to those of humans. Of course, since AI entities are able to learn by themselves, they are also able to be deterred by learning what happened in similar cases, exactly as humans do.101

SOURCES Cases Carter v. United States, 530 U.S. 255 (2000) Chance v. State, 685 A.2d 351 (Del. 1996) Clay v. State, 196 So. 462 (Fla. 1940) Conyers v. State, 790 A.2d 15 (Md. 2002) Dusenbery v. Commonwealth, 263 S.E.2d 392 (Va. 1980) Frey v. State, 708 So. 2d 918 (Fla. 1998) Gallimore v. Commonwealth, 436 S.E.2d 421 (Va. 1993) Heilman v. Commonwealth, 1 S.W. 731 (Ky. 1886) In re Devon T., 584 A.2d 1287 (Md. Ct. Spec. App. 1991) Ingram v. United States, 592 A.2d 992 (D.C. 1991) Little v. State, 554 S.W.2d 312 (Ark. 1977) Liverpool (Mayor), (1802) 102 Eng. Rep. 529 (K.B.) McCormack v. State, 15 So. 438 (Ala. 1894) Mitchell v. State, 971 P.2d 813 (Nev. 1998) Morrisey v. State, 620 A.2d 207 (Del. 1993) People v. Disimone, 650 N.W.2d 436 (Mich. Ct. App. 2002) People v. Henry, 607 N.W.2d 767 (Mich. Ct. App. 1999) People v. Prettyman, 926 P.2d 1013 (Cal. 1996) Reg v. Saintiff, (1705), 87 Eng. Rep. 1002 (K.B.) Regina v. Cogan, [1976] Q.B. 217, [1975] 2 All E.R. 1059, [1975] 3 W.L.R. 316, 61 Cr. App. Rep. 217, 139 J.P. 608 Regina v. Manley, (1844) 1 Cox Crim. Cas. 104 Richardson v. State, 697 N.E.2d 462 (Ind. 1998) State v. Aaron, 4 N.J.L. 269 (Sup. Ct. 1818) State v. Carrasco, 928 P.2d 939 (N.M. Ct. App. 1996) State v. Daniels, 109 So. 2d 896 (La. 1958)

100   Frederick Pollock, Has the Common Law Received the Fiction Theory of Corporations?, 27 L. Q. Rev. 219 (1911). 101   For the rationale of learning from case to case see Gabriel Hallevy, Rethinking the Legitimacy of the Anglo-American High Courts’ Judicial Review of Determining Factual Findings in Courts of the First Instance in Criminal Cases, 5 High Ct. Q. Rev. 20 (2009).

BARFIELD PRINT.indd 418

13/11/2018 11:53

Criminal liability of AI entities  419 State v. Dillon, 471 P.2d 553 (Idaho 1970) State v. Fuller, 552 S.E.2d 282 (S.C. 2001) State v. George, 54 A. 745 (Del. 1902) State v. Jackson, 142 S.W.2d 45 (Mo. 1940) State v. Jackson, 976 P.2d 1229 (Wash. 1999) State v. Kaiser, 918 P.2d 629 (Kan. 1996) State v. Linscott, 520 A.2d 1067 (Me. 1987) State v. Neuzil, 589 N.W.2d 708 (Iowa 1999) State v. Sargent, 594 A.2d 401 (Vt. 1991) State v. Wyatt, 482 S.E.2d 147 (W. Va. 1996) Stratford-upon-Avon Corporation, (1811) 104 Eng. Rep. 636 (K.B.) United States v. Andrews, 75 F.3d 552 (9th Cir. 1996) United States v. Jewell, 532 F.2d 697 (9th Cir. 1976) United States v. Ladish Malting Co., 135 F.3d 484 (7th Cir. 1998) United States v. Randolph, 93 F.3d 656 (9th Cir. 1996) United States v. Ruffin, 613 F.2d 408 (2nd Cir. 1979) United States v. Spinney, 65 F.3d 231 (1st1 Cir. 1995) United States v. Tobon-Builes, 706 F.2d 1092 (11th Cir. 1983) United States v. Torres, 977 F.2d 321 (7th Cir. 1992) United States v. Wert-Ruiz, 228 F.3d 250 (3d Cir. 2000) United States v. Youts, 229 F.3d 1312 (10th Cir. 2000)

Statutes 17 U.S.C.S. § 506 35 U.S.C.S. § 292 Minn. Stat. § 9913 (1927) Mont. Rev. Code § 10729 (1935) N.Y. Penal Law § 816 (1935) Okla. Stat. § 152 (1937) Utah Rev. Stat. 103-i-40 (1933)

BARFIELD PRINT.indd 419

13/11/2018 11:53

PART IV APPLYING THE LAW TO DIFFERENT APPLICATIONS OF VIRTUAL AND AUGMENTED REALITY

BARFIELD PRINT.indd 420

13/11/2018 11:53

14.  Advertising legal issues in virtual and augmented reality S.J. Blodgett-Ford, Woodrow Barfield, and Alexander Williams*

Any marketing material or product that can be shown in a print advertisement, a movie, or a television show in the physical world can also be displayed in a virtual reality world, such as a video game or an immersive virtual reality (“VR”) experience. Augmented reality (“AR”) technology can be used to replace or supplement physical advertising with projections of virtual ads into the real space surrounding the consumer.1 Ads can even be customized based on the identity of the person nearby. For example, one company promises that “[w]ith our SDK you can customize digital signs based upon who is nearby them. This can be triggered from a mobile app, a customer reward card or [a beacon]. With or without connectivity to the Internet.”2 A number of commercial enterprises specialize in creating virtual and augmented reality advertising.3 For example, the NBA created a well-received virtual reality documentary of the 2016 finals: “Viewers got so close they could see the sweat on the face” of the coach.4 Although there are practical challenges to using virtual reality and augmented reality advertising,5 it seems likely that costs will continue to decline,6 and marketers will find ways to use the medium in powerful ways. Advertising in virtual and augmented reality implicates a variety of legal issues and may prompt future legislative and regulatory actions, as well as efforts at industry self-regulation. A good rule of thumb is that virtual and augmented reality marketing

*  The authors wish to thank Professor Marc Blitz of the Oklahoma City University School of Law for his helpful review and comments. 1   Mike Butcher, The World of Minority Report Is About to Be Unlocked, TechCrunch (Apr. 19, 2017), https://techcrunch.com/2017/04/19/the-world-of-minority-report-is-about-be-unlockedand-the-key-is-inside-your-phone/. Unless otherwise indicated, all websites cited in this chapter were last visited between June and August 2017. 2  https://newaer.com/. 3   United States-based PVI, Inc. (Princeton Video Imaging) markets the L-Vis system (LiveVideo Imaging System). See Digital Broadcasting.com, Princeton Video Image, Inc. Storefront, PVI Overview In-Studio Broadcast Services, SMT, www.sportvision.com/pvi-overview (last visited Feb. 27, 2017). 4   Adi Robertson, The NBA and Oculus Just Released One of the Longest, Most Dynamic VR Films So Far, The Verge (Sep. 14, 2016, 8:00 AM), www.theverge.com/2016/9/14/12907390/ nba-oculus-vr-video-follow-my-lead-cavaliers-warriors-finals. 5   Janet Morrissey, Virtual Reality Leads Marketers down a Tricky Path, The New York Times (Mar. 5, 2017), www.nytimes.com/2017/03/05/business/media/virtual-reality-leads-­marketers-downa-tricky-path.html. 6   Tim Merel, The Reality of VR/AR Growth, TechCrunch, Jan. 11, 2017, https://techcrunch. com/2017/01/11/the-reality-of-vrar-growth/.

421

BARFIELD PRINT.indd 421

13/11/2018 11:53

422  Research handbook on the law of virtual and augmented reality involves, at a minimum, the same legal risks of traditional marketing plus risks related to the particular technology, such as products liability, personal injury and trespass, and privacy issues involved in geolocational tracking and targeting. Moreover, like traditional advertising, virtual and augmented reality advertising may be treated as “commercial speech” entitled to less protection under the First Amendment than other types of expression and, as such, may be subject to additional regulations. For example, content-based restrictions may be permissible with regard to advertising for regulated products such as alcohol, tobacco and pharmaceuticals, and process-based restrictions may relate to spam, telemarketing, and other consumer protections. In this chapter, after offering a quick primer on VR and AR ads, we will present a high-level survey of issues such as products liability (section 2), nuisance and trespass (section 3), intellectual property (patent, trademark, copyright, and trade secrets) (section 4), right of publicity (section 5), and unfair and deceptive trade practices (section 6). Privacy law issues (including privacy-related issues related to “geofencing” and geolocation) are highly relevant and are discussed in a separate chapter. Before addressing the legal issues, we provide a few baseline definitions regarding virtual and augmented reality advertisements.

1.  A VR AND AR ADVERTISING PRIMER We use the term “virtual reality” advertising and marketing to refer to advertising and marketing that takes place entirely within an intangible or simulated environment, such as in a video game or a “virtual world” such as the Second Life advertising,7 or immersive 3D advertisements (or “close to” 3D on Facebook 360).8 We define advertising and marketing broadly, and use the terms interchangeably for simplicity (although we appreciate there can be significant differences). For example, Volvo’s advertising includes an application that delivers an entire holographic showroom to users.9 Customers wearing a head-mounted display may be able to experience vehicle interiors and color combinations in three dimensions virtually.10 eBay Australia partnered with a retailer to create a virtual reality department store where shoppers can browse and select products by simply gazing at the item using a virtual reality headset.11 Ikea partnered with Apple in the use of aug-

 7   Henry Blodget, Thinking of Advertising in Second Life? Don’t Bother, Business Insider (Jul. 20, 2007, 6:38 AM), www.businessinsider.com/2007/7/thinking-of-adv; Jo Yardley, Second Life Advertising Hits New Low, Wordpress (Mar. 19, 2014), https://joyardley.wordpress.com/2014/03/19/ second-life-advertising-hits-new-low/.  8   https://virtualsky.com/ (“Immersive advertising at scale on all virtual reality platforms”).  9   Adi Robertson, Microsoft and Volvo’s New Hololens Is Fascinating and Frustrating, The Verge (Nov. 20, 2015, 12:19 AM), https://www.theverge.com/2015/11/20/9767886/micro​soft-volvo-hololensmi​xed-reality-car-showroom. 10   Robert Hof, Google Cardboard will Bring VR—and VR Ads—to the Masses, Forbes (Dec. 23, 2015, 09:49 AM), https://www.forbes.com/sites/roberthof/2015/12/23/googles-cardboard-willbring-vr-and-vr-ads-to-the-masses/#5f3cba4a38a6; Adobe VR example: Janko Roettgers, Adobe is Developing Advertising Solutions for Virtual Reality, Variety (Feb. 26, 2017, 09:01 PM), http:// variety.com/2017/digital/news/adobe-vr-video-advertising-1201996008/. 11   Ariel Bogle, eBay Launches a World-First Virtual Reality Department Store, Mashable

BARFIELD PRINT.indd 422

13/11/2018 11:53

Advertising legal issues in virtual and augmented reality  423 mented reality in their stores so that shoppers can see how a sofa will look in their living room.12 A sponsor’s ad campaign gave fans at the US Open a chance to “play” tennis against Maria Sharapova using virtual reality headsets.13 Virtual reality headsets with eye tracking capabilities can track how consumers react to signage, displays, products, and packaging. Such data could provide extremely valuable insight on the efficiency of marketing techniques utilized by retailers. More virtual reality devices have been entering the consumer market. Samsung’s Gear, Facebook’s Oculus Rift, PlayStation virtual reality and HTC’s Viv are examples, and Mattel even updated its classic View-Master toy as a virtual reality headset. Virtual reality developers are also working to ensure the technology no longer comes with the side effect of motion sickness, which has hindered its appeal. The capacity for virtual reality, combined with 360-degree video, to put the viewer at the center of an experience, such as transporting them to a holiday resort or music or sporting event, opens new territory for the marketing of brands. By contrast, the term “virtual advertising”14 has, in the past at least, been used by the advertising industry to refer to the placement of ads in live or recorded commercial television. For the purposes of our analysis, we are not addressing this traditional type of “virtual advertising.”15 Instead, for virtual reality advertising we focus on the concept of placing an advertisement solely within a virtual environment, in which case the ad and the world experienced are completely computer-generated, and for augmented reality advertising we address situations in which virtual images are seen to appear in the physical world.16 We refer to an “augmented reality” advertisement as a digital layer of information overlaid on the tangible environment such that the ad can be seen with the use of a camera on a smartphone, or viewed using a head-mounted see-through display, or even as a holographic image.17 Similarly, “mediated reality” is a mix of computer-generated virtual information combined with visual information from the real world and includes overlaying

(May 18, 2016), http://mashable.com/2016/05/18/ebay-virtual-reality-shopping/#coJ4XcDYAEqB. 12   Jon Martindale, Ikea Shopping Experience Will Be Augmented with AR App for iOS 11, www.digitaltrends.com/mobile/ios-11-ikea-augmented-reality-app/ (June 19, 2017). 13   Kristina Monllos, You Can Play Tennis Against Maria Sharapova at This Year’s U.S. Open, AdWeek (Aug. 18, 2015), www.adweek.com/brand-marketing/you-can-play-tennis-against-mar​ ia-sharapova-years-us-open-166445/. 14   Jason Tuohey, Invasion of the Video Game Ads: Ad Networks Target Online Gamers as Next Big Audience for Product Placements (November 12, 2004), Medill News Service, www.pcworld. com/article/118428/article.html. 15   These traditional online ads (banners, pop-up ads, and so on) do raise interesting “ambush marketing” issues and ad fraud concerns: Denis Sinegubko, Injection of Unwanted Google AdSense Ads, Sucuri (Jan. 10, 2017), https://blog.sucuri.net/2017/01/injection-unwanted-googleadsense-ads.html; George Slefo, Google Investigation: Ad Injection Is Infesting Millions of Devices, Advertising Age (Aug. 04, 2016), http://adage.com/article/digital/google-ad-injection-affectingmillions/305321/; Benjamin Edelman and Wesley Brandi, The Ad Networks and Advertisers that Fund Ad Injectors, Benedelman.org, www.benedelman.org/injectors/. 16   See generally Seth Grossman, Comment, Grand Theft Oreo: The Constitutionality of Advergame Regulations, 115 Yale L.J. 227 (2005). 17   See generally, W. Barfield (ed.) Fundamentals and Augmented Reality and Wearable Computers (2015).

BARFIELD PRINT.indd 423

13/11/2018 11:53

424  Research handbook on the law of virtual and augmented reality virtual objects on “real life,” thus taking away or otherwise visually altering or occluding real-world objects, such as the characters in the Pokémon Go mobile game application.18 Augmented and mediated reality both project computer-generated information in the real world and do not, specifically, change the physical characteristics of any real-world object. In principle, any augmented reality advertisement could also be thought of as mediating reality insofar as it might modify the appearance or meaning of a real sign or overwrite a real object. In addition, many applications represent the concept of augmenting and mediating reality at the same time because they project information or replace ads in real space. For the purposes of uniformity, all of these processes shall be referred to collectively as “augmented reality” throughout this chapter unless otherwise noted.19 One type of augmented reality technology uses a transparent head-mounted visual display through which the consumer directly views the world, along with a virtual ad typically in the form of an image or text projected into the physical world.20 In contrast, with an opaque headset visual display the consumer cannot directly view the world. Instead, the consumer sees the world through a video projection shown on the headset display, along with a digital advertisement superimposed over the video.21 An augmented reality display may allow the wearer to either view or occlude unwanted real images, such as a billboard advertisement containing an undesirable ad.22 In this way, the person wearing the computer (or with a smart phone holding the technology) is “mediating” reality.23 Additionally, with augmented reality systems,24 including smartphones, when wirelessly networked, a person can access information from the internet at any time and in any place.25 Further, if a virtual ad is paired to sensors placed on objects in the environment, and if a GPS sensor is worn by the user (or a mobile device with geolocation capability 18   Id; Steve Mann and Hal Niedzviecki, Cyborg: Digital Destiny and Human Possibility in the Age of the Wearable Computer at 265 (2001). 19   The 2017 movie Ghost in the Shell has powerful images of augmented reality advertisements in the urban environment. 20   Jannick Rolland & Henry Fuchs, Optical versus Video See-Through Head-Mounted Displays, in Lawrence Erlbaum Associates, Fundamentals of Wearable Computers and Augmented Reality, 113–50 (Woodrow Barfield and Tom Caudell eds, 2001). 21   See generally Woodrow Barfield & Thomas Caudell, Basic Concepts in Wearable Computers and Augmented Reality, in Fundamentals of Wearable Computers and Augmented Reality, 3–26 (Woodrow Barfield & Tom Caudell eds, 2001). 22  https://techcrunch.com/2017/04/19/the-world-of-minority-report-is-about-be-unlockedand-the-key-is-inside-your-phone/; Steve Mann & James Fung, EyeTap Devices for Augmented, Deliberately Diminished, or Otherwise Altered Visual Perception of Rigid Planar Patches of Real World Scenes, 11 Presence: Teleoperators and Virtual Environments, 158–75 (2002). See also memo from the European Broadcasting Union Legal Department, Consultation on possible requirement to use virtual imaging techniques on television to block out alcohol advertising at sports venues, www.ebu.ch/CMSimages/en/leg_pp_reply_csa_alcohol_advertising_tcm6-4355.pdf (last visited Feb. 27, 2017). 23   Steve Mann, Mediated Reality, M.I.T. M.L. Technical Report 260, Cambridge, MA (1994); Steve Mann, Intelligent Image Processing (2001); Steven Feiner, Blair MacIntyre, & Doree Seligmann, Knowledge-Based Augmented Reality, 36 Communications of the ACM 52–62 (1993). 24   By use of the term “system,” the authors are referring to a head-worn display and wearable computer, which is used to generate an image or text. 25   See generally Barfield & Caudell, supra note 21. A person can also access information from the environment using appropriate sensor technology.

BARFIELD PRINT.indd 424

13/11/2018 11:53

Advertising legal issues in virtual and augmented reality  425 is used), the projected virtual ad can be associated with a particular product, brand, or service at a particular location and at a particular time in real space.26 With this baseline, we now turn to various legal issues raised by advertising and marketing in virtual reality and augmented reality, including: products liability; trespass and nuisance; intellectual property (patent, trademark, copyright and trade secrets); right of publicity; unfair and deceptive trade practices, and other legislative and regulatory areas, including industry self-regulation.

2.  PRODUCTS LIABILITY: PERSONAL INJURY At first glance, it is not obvious that there would be any real-world products liability issues for virtual reality advertisements. By definition, such ads are intangible. How can they cause a personal injury? In fact, certain types of visual content can cause nausea,27 as well as seizures. This has allegedly been experienced with such content as a flashing Vine ad on Twitter and a Pokémon TV episode in Japan.28 A tweet to journalist Kurt Eichenwald (who was known to have epilepsy) containing the message “YOU DESERVE A SEIZURE FOR YOUR POSTS” and a blinding strobe light that triggered a seizure led to the FBI’s arrest of the sender for criminal cyberstalking with the intent to kill or cause bodily harm.29 A teenager was arrested for allegedly terrorizing and interfering with the operation of a school by using a mobile app game to make a video that showed him as a shooter gunning down his classmates, superimposed against real-world images, then uploading that video to YouTube.30 Law enforcement reportedly commented: “We don’t know at what time that game becomes a reality.”31 26   Mike Butcher, The World of Minority Report is About to Be Unlocked, and the Key Is Inside Your Phone, Tech Crunch (Apr. 19, 2017), https://techcrunch.com/2017/04/19/the-world-of-minor​ ity-report-is-about-be-unlocked-and-the-key-is-inside-your-phone/; Steven Feiner et al., MARS— Mobile Augmented Reality Systems, http://monet.cs.columbia.edu/projects/mars/mars.html. 27   Michael S. Sherman, Augmented and Virtual Reality: Emerging Legal Implications of “The Final Platform” (Reed Smith), p.17, www.reedsmith.com/en/perspectives/2017/06/augmented-andvirtual-reality (citing Scott Stein, The Dangers of Virtual Reality, Mar. 29, 2016, www.cnet.com/ news/the-dangers-of-virtual-reality/; Jack Nicas and Deepa Seetharaman, What Does Virtual Reality Do to Your Body and Mind? The Wall Street Journal (Jan. 3, 2016), www.wsj.com/articles/ what-does-virtual-reality-do-to-your-bodyand-mind-1451858778.) Some general tips are also offered at Laura Klepacki, Legal AR/VR Pitfalls Marketers Should Avoid, According to Law Experts, MarketingDive (Jul. 10, 2017), www.marketingdive.com/news/legal-arvr-pitfalls-marketers-shou​ ld-avoid-according-to-law-experts/445963/. 28   Ali Venosa, Twitter Pulls Flashing Vine Ad after Epilepsy Charity Says It May Cause Seizures, Medical Daily (Jul. 15, 2015, 5:25 PM), www.medicaldaily.com/twitter-pulls-flashing-vin​ e-ad-after-epilepsy-charity-says-it-may-cause-seizures-342584. 29   Cecilia Kang, A Tweet to Kurt Eichenwald, a Strobe and a Seizure. Now, an Arrest. The New York Times (Mar 17, 2017), www.nytimes.com/2017/03/17/technology/social-media-attack-thatset-off-a-seizure-leads-to-an-arrest.html. 30   Caroline Moss, Teenager Was Arrested after Using an iPhone App to Go on a Virtual Shooting Spree, Business Insider (Sep. 16. 2013, 11:46 AM), www.businessinsider.com/teenager-arreste​ d-real-strike-app-virtual-shooting-spree-2013-9. 31   Gaby Dunn, Teen Uses Mobile App to Go on Virtual Shooting Spree at School, The Daily Dot (Sep. 14, 2013, 10:23 AM), www.dailydot.com/crime/teen-gun-app-shooting-arrested-the-real-strike/.

BARFIELD PRINT.indd 425

13/11/2018 11:53

426  Research handbook on the law of virtual and augmented reality The products liability/personal injury issue is perhaps more acute for advertisements in augmented reality and mediated reality because of the more direct interaction between the ad and the physical world. For example, there are initiatives to project virtual ads on car windshields: A driver that’s close to running out of gas would see an alert pop up that notes the fuel situation and offers to find a nearby gas station. The car’s virtual assistant will offer a choice, again on the windshield, of two options, including directions to a station where the driver is eligible for a free cup of coffee—an ad placed by the gas company that fits with the driver’s buying patterns, also known by his smart car.32

Similarly, Panasonic has demonstrated technology that allows a driver to “order and pay for fast food through a heads-up display, before pulling off the freeway to collect it.” 33 Given the widespread use of GPS displays in cars, it would seem like a natural extension to allow ads for nearby attractions on a heads-up display during navigation, such as during long distance travel. Like GPS systems and mobile apps generally, augmented reality applications and advertisements could contribute to accidents related to distracted driving.34 Liability issues will likely be treated in a similar manner, with the caveat that advertisements are often given less deference than other types of products and services. Thus, an advertisement may be more susceptible to legal risk than a product that is viewed as more “functional”—such as a low gas alert and locations of the nearest gas stations without an ad display for a coffee coupon. Issues of distracted driving due to texting and use of mobile apps came to the forefront after the success of the Apple iPhone, which launched in 2007. Many states now have laws that prohibit texting and driving.35 In 2014, a six-year-old girl was allegedly struck and killed by an Uber driver when she was crossing the street in San Francisco.36 The family sued and claimed that Uber’s application was unreasonably dangerous because drivers needed to look at their phones while driving, in violation of California vehicle code 23123: “A person shall not drive a vehicle while using a wireless telephone unless that telephone is specifically designed and configured to allow hands-free listening and talking and is used in that manner while driving.” The lawsuit against Uber was settled,37 so there was no definitive judicial ruling on the issue.

32   Alex Webb, Move Over Mobile Phone: The Next Ad Frontier is the Windshield, Bloomberg (Jan. 6, 2017, 7:00 AM), www.bloomberg.com/news/articles/2017-01-06/smart-windshields-seen-as-the-ne​ w-ad-billboards-inside-a-car. 33   Id. 34   Luke Plunkett, Pokemon Go Player Crashes Into Cop Car, Captured on Camera, Kotaku (Jul. 19, 2016, 7:45 PM), https://kotaku.com/pokemon-go-player-hits-cop-car-captured-on-­cam​era1783942861; Sean Lahman, Pokemon Go Player Crashes His Car Into a Tree, USA Today (Jul. 14, 2016, 11:48 PM), www.usatoday.com/story/news/nation/2016/07/14/pokmon-go-player-crashes-hi​ s-car-into-tree/87074762/. 35   Distracted Driving, Governor’s Highway Safety Association, http://www.ghsa.org/statelaws/issues/Distracted-Driving, (last visited Jul. 04, 2017). 36   Dominic Rushe, Uber Sued by Family of Six-Year-Old Killed in San Francisco Crash, The Guardian (Jan. 27, 2014, 2:13 PM), www.theguardian.com/world/2014/jan/27/uber-sued-family-sixyear-old​-killed-crash-san-francisco. 37  www.law360.com/articles/824588/Pokémon-go-s-product-liability-woes.

BARFIELD PRINT.indd 426

13/11/2018 11:53

Advertising legal issues in virtual and augmented reality  427 In a case that also involved risky behavior by app users, Snapchat was sued by an accident victim who was struck by a driver who was allegedly trying to reach 100 miles per hour and use the Snapchat “speed filter,” which tracks a user’s speed and allows them to overlay the speed on top of an image.38 The case was dismissed by a Georgia state court judge in 2017 on the grounds that the Communications Decency Act (“CDA”) gives Snapchat immunity,39 with the judge reportedly relying on the section stating: “No provider or user of an interactive computer service shall be treated as the publisher or speaker of any information provided by another information content provider.”40 This ruling seems like a stretch of the intent and language of the CDA. The victim was not alleging that the publishing of the speech caused the accident, but rather the fact that the Snapchat “speed filter” seemed designed to promote excessive speed and distracted driving from a products liability perspective. In an industry effort to reduce risks and liability exposure, many in-vehicle GPS systems now disable or limit certain features, such as address input, when the vehicle is in motion to reduce the risk of distracted driving incidents.41 Similarly, a lockdown was added on Pokémon Go “hatching” if the user is traveling over 20 mph,42 but “the application does not shut down entirely if the user is exceeding 20 miles per hour. Users who are on the prowl for Pokémon are still able to view maps displaying Pokémon while driving, and are also still able [to] catch Pokémon and interact with other aspects of the application. Moreover, the developers have not modified the aspect of their game that encourages players to be the first on scene to catch the virtual creatures.” 43 The personal injury issues are not just related to driving, however. There have been initiatives regarding texting and walking, although they have not yet been widely adopted.44 A user of Google Maps sued Google when she was injured after the app allegedly instructed her to walk across a rural highway with no stoplights or sidewalks. The US District Court for the District of Utah dismissed the case and refused to impose a legal duty on Google, noting that while it was foreseeable that the plaintiff could get injured as

  Id.   A Win for Snapchat in Crash Lawsuit Tied to Speed Filter, CBS News (Jan. 23, 2017, 3:30 PM), www.cbsnews.com/news/snapchat-speed-filter-accident-lawsuit-claim-dismissed/. 40   47 U.S. Code § 230(c). 41   Anekin007, Comment to Bypass Navigation, Piloteers.org (Feb. 15, 2016, 1:59 AM), www. piloteers.org/forums/114-2016-third-generation-pilot/100385-bypass-navigation.html; www.­jeepp​ atriot.com/forum/5-jeep-patriot-general-discussion/301674-cant-input-address-into-navi-when-mo​ ving.html. 42  https://www.law360.com/articles/824588/Pokémon-go-s-product-liability-woes. 43   Id. 44   Hawaii Law Targets “Smartphone Zombies” with Crosswalk Ban, The Guardian, July 28, 2017 (“A ban on pedestrians looking at mobile phones or texting while crossing the street will take effect in Hawaii’s largest city in late October, as Honolulu becomes the first major US city to pass legislation aimed at reducing injuries and deaths from ‘distracted walking’”); Alyssa Newcomb, Texting while Walking Banned in New Jersey Town, ABC News (May 13, 2012), http://abcnews.go.com/blogs/head​ lines/2012/05/texting-while-walking-banned-in-new-jersey-town/ (“Fort Lee, N.J. police said they will begin issuing $85 jaywalking tickets to pedestrians who are caught texting while walking”); Amy Wang, The Alarming Video that May Persuade You to Never, Ever Text and Walk, The Washington Post (Jun. 09, 2017), www.washingtonpost.com/news/post-nation/wp/2017/06/09/the-alarmingvideo-that-may-convince-you-to-never-ever-text-and-walk/?utm_term=.fccde5e47593#comments. 38 39

BARFIELD PRINT.indd 427

13/11/2018 11:53

428  Research handbook on the law of virtual and augmented reality a result of walking alongside a busy road, users are ultimately responsible for their own safety. The court emphasized the ‘“high social utility’ of applications like Google Maps and found that imposing a duty on Google to provide accurate directions and warn of traffic hazards would ‘clearly be difficult, if not impossible for Google to bear.’”45 The “high social utility” defense that works so effectively for Google in many legal areas may not work as well for applications that a judge might consider to be more “frivolous,”46 such as an advertisement or a video game.47 For the mass market,48 augmented reality was popularized globally by Pokémon Go in 2016,49 and it has been used in games such as Parallel Mafia and Real Strike.50 There have been news reports of various injuries allegedly related to Pokémon Go: “two men fell several stories off a sandstone bluff in California while zealously chasing characters in Pokémon Go . . . Reports and videos have also surfaced showing stampeding mobs of players gathering in public areas causing substantial injuries in an unruly attempt to catch rare Pokémon.” 51 There also have been claims that Pokémon Go has been used for criminal activity: The game’s “beacon feature,” which allows players to attract others to their physical location, has also left players vulnerable to criminals. For example, a teenager in Guatemala allegedly was lured into a side street by the beacon feature and murdered in an apparent robbery. In a similar incident, a dozen people in Missouri allegedly were robbed at gunpoint after being lured into a secluded area by the beacon feature.52

The alleged criminal activity involving Pokémon Go is, in some ways, comparable to luring victims by putting false ads on the classified service Craigslist. Craigslist has generally prevailed in lawsuits on such matters on the ground that it is only a passive host and does not promote such activities. However, under pressure from law enforcement and state attorneys general, who have accused the site of facilitating prostitution and being a

 www.law360.com/articles/824588/Pokémon-go-s-product-liability-woes.  www.eff.org/deeplinks/2015/10/big-win-fair-use-google-books-lawsuit. 47   Corynne McSherry, Big Win for Fair Use in Google Books Lawsuit, Electronic Frontier Foundation (Oct. 16, 2015), https://www.eff.org/deeplinks/2015/10/big-win-fair-use-google-bookslaw​suit. 48   Artyom Dogtiev, Pokemon Go Usage and Revenue Statistics, Business of Apps (Aug. 15, 2016), http://www.businessofapps.com/pokemon-go-usage-revenue-statistics/; Helen Zhao, Pokemon Go Attracts Diverse Crowd of Gamers, Study Suggests, LA Times (Jul. 18, 2016, 4:45 PM), www.latimes.com/business/la-fi-tn-pokemon-demographics-20160718-snap-story.html. 49   R. Darren Price, Rare Pokemon Go Stampede in Central Park, NBC New York (Jul. 18, 2016, 1:40 PM), www.nbcnewyork.com/news/local/Pokemon-Go-Players-Stampede-New-YorkCentral-Park-387303572.html. 50   www.moregameslike.com/parallel-mafia/; www.facebook.com/RealStrike.info/. 51   www.law360.com/articles/824588/Pokémon-go-s-product-liability-woes; see also Hayley Tsukayama, Pokemon Go’s Unexpected Side Effect: Injuries, The Washington Post (Jul. 10, 2016), www.washingtonpost.com/news/the-switch/wp/2016/07/08/pokemon-gos-unexpected-sideeffect-injuries/?utm_term=.2560fa11213c. There have also been anecdotal reports of injuries from Pokémon Go, such as on Facebook social media posts: www.rt.com/business/351813-Pokémon-gosberbank-insurance/ (“Eight employees in VICE’s Toronto office were injured today after walking into each other while trying to catch Pokémon”). 52   Id. 45 46

BARFIELD PRINT.indd 428

13/11/2018 11:53

Advertising legal issues in virtual and augmented reality  429 “public nuisance,” it has taken steps to reduce the risk of liability: “the website eliminated its erotic services section, replacing it with ‘adult services’ and pledging to review every ad posted there to prevent flagrant prostitution and pornography.”53 Backpage.com has been less successful. In early 2017, the US Senate investigative subcommittee rejected the company’s claim of immunity as a host of content under the Communications Decency Act (“CDA”).54 The subcommittee concluded: “Backpage’s public defense is a fiction. Backpage has maintained a practice of altering ads before publication by deleting words, phrases, and images indicative of criminality, including child sex trafficking.”55 One of the lessons from Backpage.com and Craigslist is that virtual and augmented reality advertisements that actively encourage or otherwise take steps to promote risky behavior, particularly involving children, are at higher risk for liability. When Pokémon Go launched, there were predictions that “[g]iven the mounting reports of injuries, lawsuits are likely to follow.”56 Many personal injury law firms sought to attract such plaintiffs with website articles about the dangers of Pokémon Go.57 For example, the website mobilelawyers.com placed a large banner at the top of the page inviting anyone “Injured Because of Pokémon Go” to call a toll-free number for a free legal consultation. Nevertheless, the authors could not, as of the summer of 2017, find on Lexis Courtlink any reported personal injury lawsuit filed against Niantic Inc. Many of such injuries may have been handled via non-public settlements or insurance claims. And the players of the game, particularly adult players,58 bear at least some responsibility for their actions. In addition, stronger safety warnings were added to the Pokémon Go game in July of 2016: The popular augmented reality game doesn’t just urge you to “remember to be alert at all times” and “stay aware of your surroundings” anymore. It’s gotten much more specific. “Do not trespass while playing Pokémon Go,” one warns. “Do not play Pokémon Go while driving,” reads another. Lastly, “Do not enter dangerous areas while playing Pokémon Go” joined the game’s concerns.59

It seems likely that the personal injury claims regarding augmented reality advertising will be influenced by the guidance in the distracted driving and walking-related cases generally. Actions that tend to encourage dangerous behavior, such as the Snapchat speed filter or Tesla’s Autopilot feature, are likely to increase exposure. Voluntary steps to reduce 53   DeeDee Correll, Former Boyfriend Used Craigslist to Arrange Woman’s Rape, Police Say, Los Angeles Times (Jan. 11, 2010), http://articles.latimes.com/2010/jan/11/nation/la-na-ra​pe-crai​ gslist11-2010jan11. 54  www.mccaskill.senate.gov/imo/media/doc/2017.01.10%20Backpage%20Report.pdf 55   Senate Report, p. 4. 56  www.law360.com/articles/824588/Pokémon-go-s-product-liability-woes. 57   See, e.g., Determining Who’s at Fault in Pokemon Go Injuries, Clements, Taylor, Butkovich & Cohen, LPA, Co. (Nov. 22, 2016), https://ctbclawyers.com/2016/11/22/determining-whos-faultpokem​on-go-injuries/. 58   The men in the California cliff incident were reportedly “in their early 20s” and reportedly crossed a fenced area to get to the bluff. Veronica Rocha, 2 California Men Fall Off Edge of Ocean Bluff While Playing Pokemon Go, Los Angeles Times (Jul. 14, 2016, 3:45 PM), www.latimes.com/ local/lanow/la-me-ln-pokemon-go-players-stabbed-fall-off-cliff-20160714-snap-story.html. 59   Brianne Heldman, Pokemon Go Adds New Safety Warnings, Entertainment Weekly (Jul. 31, 2016, 12:00 AM), http://ew.com/article/2016/07/31/pokemon-go-new-safety-warnings/.

BARFIELD PRINT.indd 429

13/11/2018 11:53

430  Research handbook on the law of virtual and augmented reality legal risk, such as disabling certain features or functionality, allowing for prompt online processing of complaints, and strong user warnings will likely continue to be the first line of defense for augmented and virtual reality advertisements.

3.  TRESPASS AND NUISANCE Trespass and nuisance claims may relate to a virtual ad projected onto a virtual version of private property, or to the use of augmented reality to project ads onto physical property or the airspace above (or space below) land or private structures. In addition, trespass and nuisance claims may come into play when humans interacting with an augmented reality advertisement enter or congregate near private property without permission (or exceed permitted use) due to that interaction. In this section, we will consider the appropriateness of trespass to land, trespass to chattels, and nuisance claims involving advertising using virtual reality and augmented reality.60 3.1  Trespass—to Land and Chattel The law of trespass distinguishes between trespass to land and trespass to chattels.61 A trespass to land occurs when a person enters land of another without permission or privilege to do so.62 Trespass to chattels occurs when there is an intentional act to dispossess another of chattel or use or intermeddle with a chattel in possession of another.63 Further, the Restatement of Torts defines physical intermeddling as intentionally bringing about physical contact with chattel.64 Additionally, there is an extension of the trespass to land doctrine known as trespass to airspace, which occurs if the defendant invades the airspace of another party. Normally, a trespass to airspace is committed only if the defendant’s conduct occurs within a certain distance above the land.65 The Restatement of Torts addresses trespass to airspace by stating that a trespass may be committed on, beneath, or above the surface of the earth.66 Flight by aircraft in the air space above the land of another is a trespass if, but only if, (1) it enters into the immediate reaches of the air space

60   Andrew L. Rossow, Gotta Catch . . . A Lawsuit? A Legal Insight into the Intellectual, Civil, and Criminal Battlefield Pokémon Go has Downloaded onto Smartphones and Properties around the World, 16 J. Marshall Rev. Intell. Prop. L. 329, 338–9 (2017) (discussing Ohio’s “attractive nuisance” doctrine and whether property owners would have a duty to warn) and Note: Augmenting Property Law: Applying the Right to Exclude in the Augmented Reality Universe, 19 Vand. J. Ent. & Tech. L. 1057 (2017). 61   I. Trotter Hardy, The Ancient Doctrine of Trespass to Web Sites, 1996 J. ONLINE L. 7 (1996). 62   RESTATEMENT (SECOND) OF TORTS, § 163 (2nd 1979). 63   RESTATEMENT (SECOND) OF TORTS § 217(b) (1965). The tort of trespass to chattels contains very similar fault requirements. See RESTATEMENT (SECOND) OF TORTS § 217: “A trespass to a chattel may be committed by intentionally (a) dispossessing another of the chattel, or (b) using or intermeddling with a chattel in the possession of another.” 64   Id. 65   See Colin Cahoon, Low Altitude Airspace: A Property Rights No-Man’s Land, 56 J. AIR. L. & CoM. 157 (1990). 66   RESTATEMENT (SECOND) OF TORTS, § 159 (2nd 1979).

BARFIELD PRINT.indd 430

13/11/2018 11:53

Advertising legal issues in virtual and augmented reality  431 next to the land, and (2) it interferes substantially with the other’s use and enjoyment of his land.67 To bring an action for trespass into another’s airspace, a land owner is required to show either actual physical damage to the land, diminution in property value, infringement of the owner’s occupation or livelihood, or substantial interference with the plaintiff’s use and enjoyment of their real property.68 Because virtual advertisements and augmented reality advertisements never physically enter the airspace above or below land in the real world, it seems unlikely that a trespass to land (or airspace) claim would make sense for such advertising. However, augmented reality advertising could entice or lure real people to enter the private property of another, or exceed authorized use of private property. Also, a trespass to chattels claim may be available even if there is no physical trespass. Unlike trespass to real property, for which proof of injury is not required,69 trespass to chattels does require proof of demonstrable injury to the chattel, or to the owner’s ability to use the chattel.70 Proof of injury has been required even for situations where there was no physical contact or in which the contact was not detectable by human senses. For example, in Thrifty-Tel Inc. v. Bezenek,71 the defendant used a computer to perform an automated search of a telephone carrier’s system for authorized telephone code. A California court held that injury to the chattel “lies where an intentional interference with the possession of personal property has proximately caused injury.”72 The California Supreme Court, in Intel Corp. v. Hamidi,73 found unauthorized email messages sent by the defendant to the plaintiff’s proprietary computer system to be insufficient to constitute trespass to chattel. The Hamidi court concluded: “The tort [of trespass to chattels] does not encompass, and should not be extended to encompass, an electronic communication that neither damages the recipient computer system nor impairs its functioning.”74 Finally, in Bradley v. American Smelting,75 the defendant’s micrometallic particles present on plaintiff’s property were undetectable by the human senses. The California Supreme Court required proof of actual and substantial damages to support a trespass action.76 The Court suggested that the invasion of an intangible such as noise, without damaging property, will not support a tort action for trespass.77   RESTATEMENT (SECOND) OF TORTS, § 158 cmt. I., 159 (2nd 1979).   See Cahoon, supra note 65, at 197. 69   RESTATEMENT (SECOND) OF TORTS, § 163 (“One who intentionally enters land in the possession of another is subject to liability to the possessor for a trespass, although his presence on the land causes no harm to the land, its possessor, or to any thing or person in whose security the possessor has a legally protected interest”). 70   Intel Corp. v. Hamidi, 71 P.3d 296 (Cal. 2003) (limiting trespass to chattels under California law to acts physically damaging or functionally interfering with property). 71   Thrifty-Tel, Inc. v. Bezenek, 54 Cal. Rptr. 2d 468, 473 (Cal. Ct. App. 1996) (affirming judgment that unauthorized access to telephone system constituted a trespass to chattels). 72   Thrifty-Tel, Inc. v. Bezenek, 54 Cal. Rptr. 2d 468, 473 (Cal. Ct. App. 1996) (“Trespass to chattel, although seldom employed as a tort theory in California (indeed, there is nary a mention of the tort in Witkin’s Summary of California Law), lies where an intentional interference with the possession of personal property has proximately caused injury”). 73   Hamidi, 71 P.3d at 300. 74   Id. 75   Bradley v. American Smelting and Refining Co., 709 P.2d 782 (Wash. Sup. Ct. 1985). 76   Id. 77   Wilson v. Interlake Steel Co., 32 Cal. 3d 229, 233–4 (1982). 67 68

BARFIELD PRINT.indd 431

13/11/2018 11:53

432  Research handbook on the law of virtual and augmented reality 3.1.1  Trespass and virtual advertising One example of a situation where the value of certain physical property—land and buildings—for physical advertising was allegedly diminished due to virtual advertising is Sherwood 48 Associates v. Sony Corp. of America.78 In the movie Spider-Man, instead of using the photos of actual ads on the buildings in the physical world, Sony generated additional product placement/sponsorship revenue by allowing companies to pay to have their own advertising superimposed when images of the buildings appeared.79 The Second Circuit Court of Appeals addressed the issue of whether a trespass is committed under New York law when one party’s physical contact with another party’s personal property diminishes the value of that property without actual physical damage to the property.80 Although a previous federal district court applying New York law had suggested that a “trespasser is liable when the trespass diminishes the . . . value of personal property” even in the absence of physical damage to the property,81 the Second Circuit relied on the Hamidi decision,82 and ruled that without proof of damages, no trespass was committed.83 The court reasoned that bouncing a laser beam off a building to create a digital photograph did not constitute a trespass as light beams naturally bounce off of buildings all the time. In augmented reality advertising, the apparent projection of a virtual image does not physically touch the surface of a tangible object. Thus, under Sherwood,84 the apparent projection of a virtual image into another’s space should not be sufficient to constitute trespass to chattels. Holding otherwise would mean that the courts could find that a trespass occurs for other intangibles, such as radio or television signals, that enter a plaintiff’s land but cause no harm. Augmented reality advertising, however, may be treated differently in a trespass analysis from the ads in the virtual world of the Spider-Man movie because augmented reality ads superimpose virtual images in the tangible world. 3.1.2  Trespass and augmented reality advertising In the context of augmented reality advertising, it is unlikely that a plaintiff will be able to show that the apparent projection of a virtual image into their airspace resulted in physical damage to real property since no physical contact with the plaintiff’s land occurs. However, a property owner may try to show that the apparent projection of a virtual image into their airspace resulted in a diminution of their property value, such as may have been the case in the Sony Spider-Man movie advertisement situation if, hypothetically, Sony had projected the new ads on top of the existing ads in the physical world. This could occur if the virtual image obscured an ad on the plaintiff’s property, thus lowering the rent for billboards in their space. If the property owner was in the business of renting advertising space, the projection of a virtual image into such space could also infringe on

78   Sherwood 48 Assocs. 76 Fed. Appx. at 390 (affirmed dismissal of federal trade dress claim on Sony’s use of Times Square buildings in Spider-Man film). 79   Id. 80   Id. 81   Register.com, Inc. v. Verio, Inc., 126 F. Supp. 2d 238, 250 (S.D.N.Y. 2000). 82   Hamidi, 71 P.3d at 296. 83   Id. 84   Sherwood 48 Assocs., 76 Fed. Appx. at 389.

BARFIELD PRINT.indd 432

13/11/2018 11:53

Advertising legal issues in virtual and augmented reality  433 the land owner’s occupation or livelihood. Thus, it is worth considering whether a trespass to chattels claim might be available in situations involving augmented reality ads. Generally, trespass to chattels requires a substantial interference with the use and enjoyment of the chattel.85 Trespass to chattels “lies where an intentional interference with the possession of personal property has proximately caused injury.”86 In the context of augmented reality advertising, a court would likely analyze whether the apparent projection of a virtual image into the space of another party is sufficient to cause harm to the chattel. Direct physical contact may not be necessary because even an indirect touching, or indirect entry,87 may give rise to trespass.88 Indeed, the requirement of a tangible contact has been relaxed almost to the point of being discarded.89 When analyzing whether there is damage to chattel resulting from contact with an intangible, the courts look to whether the contact with the chattel is sufficient to impair the chattel.90 A number of cases have addressed the issue of how much contact is necessary to constitute trespass to chattel. Trivial contacts with chattel are not sufficient to rise to the level of trespass.91 The case law dealing with trespass in relation to computer networks reveals a fairly wide standard on what constitutes harm to chattel. For example, in Thrifty-Tel v. Bezenek,92 the defendant used electronic signals sent by a computer to perform an automated search of a telephone carrier’s system for authorized telephone calling codes. In Bezenek, the court held that because such interference overburdened the system, denying some subscribers access to phone lines, there was sufficient harm to maintain a trespass action present.93 In CompuServe, Inc., v. Cyber Promotions, Inc.94 an Ohio court found that the electronic signals received by the CompuServe system for spam advertisements sent by the defendants were sufficient to support a trespass claim.95 Although the court found that   Dan Burk, The Trouble with Trespass, 4 J. Small & Emerging Bus. L. 27 (2000).   Thrifty-Tel, Inc. v. Bezenek, 46 Cal. App. 4th 1559 (1996); see also America Online, Inc. v. IMS, 24 F.Supp.2d 548, 550 (E.D.Va.1998) (“A trespass to chattels occurs when one party intentionally uses or intermeddles with personal property in rightful possession of another without authorization,” citing Restatement (Second) of Torts § 217(b)). “One who commits a trespass to a chattel is liable to the possessor of the chattel if the chattel is impaired as to its condition, quality, or value.” Id (quoting Restatement (Second) of Torts § 218(b)). 87   Dust particles which migrate onto another’s real and personal property may be sufficient to cause damage. 88  Wilson, supra note 77; Roberts v. Permanente Corp., 188 Cal. App. 2d 526 (1961). 89   See Burk, supra note 85. 90   Section 218 of the Restatement (Second) of Torts describes the circumstances under which a trespass to chattels may be actionable: “One who commits a trespass to a chattel is subject to liability to the possessor of the chattel if, but only if, (a) he dispossesses the other of the chattel, or (b) the chattel is impaired as to its condition, quality, or value, or (c) the possessor is deprived of the use of the chattel for a substantial time, or (d) bodily harm is caused to the possessor, or harm is caused to some person or thing in which the possessor has a legally protected interest.” RESTATEMENT (SECOND) OF TORTS § 218. 91   Id. 92   Thrifty-Tel, 54 Cal. Rptr. 2d at 472. 93   Id. 94   CompuServe, Inc. v. Cyber Promotions, Inc., 962 F. Supp. 1015 (S.D. Ohio 1997) (granting preliminary injunction on grounds that unsolicited email constituted a trespass to chattels). 95   Cyber Promotions, Inc. was a commercial service that transmitted unsolicited bulk email, or spam, to thousands of user addresses on the CompuServe network. 85 86

BARFIELD PRINT.indd 433

13/11/2018 11:53

434  Research handbook on the law of virtual and augmented reality transmission of the ads over the CompuServe system did not amount to a dispossession of the system, it held that plaintiffs did not need to show physical dispossession of the system.96 The plaintiff had to show an interference that impaired the value of the chattel, but not necessarily impairment of the chattel’s physical condition. In a related case, Hotmail Corp. v. Van$ Money Pie, Inc.,97 a company’s mailing storage space was filled up by unwanted advertising spam emails (some of which allegedly had pornographic materials). Complaints about the emails, which went to the hotmail.com return email addresses provided by the defendants, “threatened to damage” the company’s “ability to service its . . . customers.” The court found sufficient damage and concluded that a trespass to chattels had occurred.98 Finally, in America Online, Inc. v. IMS, which also involved unsolicited email advertisements, the court held: “A trespass to chattels occurs when one party intentionally uses or intermeddles with personal property in rightful possession of another without authorization.”99 Not all cases have found electronic interference with a computer network sufficient contact to constitute trespass to chattels. Based on the courts’ reasoning in the above cases, a key issue for augmented reality advertising will be determining the standard for damages when a virtual ad replaces or changes the meaning of a physical ad. Since the virtual ad will not damage the physical structure of a real-world billboard or other advertising physical infrastructure, the plaintiff will have to focus on economic harm or possible damage to the brand to support a cause of action for trespass to chattels. Interestingly, the pre-Hamidi cases correspond to a time of heightened government concern about hackers accessing military and financial institutions’ databases and networks. In response to these concerns Congress enacted several federal statues, such as the Computer Fraud and Abuse Act,100 and the Economic Espionage Act.101 However, these Acts do not directly address the issue of trespass for virtual images projected into the space of another party. 3.2 Nuisance If trespass to land and trespass to chattel are unavailable as a cause of action, are any other legal theories relevant for a plaintiff seeking redress for the unwanted projection of a virtual image onto their property? Plaintiffs seeking redress for invasion of their property by intangibles have often turned to the law of nuisance.102 A “nuisance” can be defined generally as an activity that interferes—intentionally103—with the use and enjoyment of   CompuServe, 962 F. Supp. at 1022.   Hotmail Corp. v. Van$ Money Pie, Inc., No. C-98 JW PVT ENE, 1998 WL 388389 (N.D. Cal. 1998) (granting a preliminary injunction on defendant’s sending spam via plaintiff’s free email service on grounds, in relevant part, that the defendant thereby trespassed on the plaintiff’s chattels).  98   Id.  99   America Online, Inc. v. IMS, 24 F.Supp.2d 548, 550 (E.D.Va.1998) (citing Restatement (Second) of Torts § 217(b)). 100   Computer Fraud and Abuse Act, 18 U.S.C. § 1030 (2001). 101   Economic Espionage Act, 18 U.S.C. § 1831 (1996). 102   See generally City of Modesto Redevelopment Agency v. Superior Court, 119 Cal. App. 4th 28 (Cal. Ct. App. 2004). 103   The intent requirement is usually satisfied on the basis of knowledge or purpose—if the  96

 97

BARFIELD PRINT.indd 434

13/11/2018 11:53

Advertising legal issues in virtual and augmented reality  435 land through unreasonable intangible invasions.104 Nuisance law has proved to be a very flexible legal doctrine, making adjustments to cover sounds, sights, and odors.105 Further, private nuisance law primarily addresses the right to use and enjoy one’s land free of intangible invasions.106 Examples of nuisance include disturbances through dust, smoke, gas, chemicals, or excessive light.107 Nuisances generally can be enjoined only if the harm they cause the property owner exceeds the benefits associated with the conduct. Under nuisance law, the invasion of the plaintiff’s interest in the use and enjoyment of their land must be substantial.108 However, this requirement has not been interpreted to mean that the invasion must have a physical or tangible impact. In determining harm as a result of nuisance, the court may look to whether there is depreciation in the market or rental value of the plaintiff’s land. In Sherwood 48 Associates,109 the New York court declined to find an economic effect for the projection of light onto the surface of the plaintiff’s advertisements located on buildings in Times Square. To determine whether an intangible may serve as a nuisance, courts often use a balancing test.110 An example is The Shelburne v. Crossan Corp., an early case which dealt with light from a large electronic sign as an invading intangible into the rooms of a hotel on an Atlantic City boardwalk.111 In Shelburne, the court found a nuisance (for the operation of the sign after the hotel’s dance orchestra stopped playing at midnight) after weighing the plaintiff’s complaint against the defendant’s activities. By contrast, the court in Hildebrand v. Watts,112 using a balancing test, found that the defendant’s security lights that illuminated the plaintiff’s property did not constitute a nuisance because the intrusion of light into the plaintiff’s property did not outweigh the defendant’s legitimate interest in casting light for security purposes. For augmented reality advertising, the success of a nuisance action will depend on whether the court finds that the apparent projection of a virtual image into another’s space substantially interferes with the plaintiff’s use and enjoyment of their land and outweighs the defendant’s legitimate use of their land, including the airspace above the land. This is a high bar for such advertising to meet, and more recent cases involving augmented reality games show that aggrieved parties may be willing to take legal action. The Netherlands threatened action against Niantic in The Hague in order to force the removal of Pokémon virtual characters from Kijkduin, a protected beach. Niantic

actor acts with the purpose of causing an invasion or knows that the invasion is occurring or is substantially certain to result. 104   Pestey v. Cushman, 788 A.2d 496, 502 (Conn. 2002). 105   Cf Horn v. City of Birmingham, 648 So.2d 607 (Ala. Civ. App. 1994). 106   Angerman v. Burick, 791 N.E.2d 984 (Ohio 2003). 107   Id. Cf Horn, 648 So.2d at 607. 108   See generally Henry E. Smith, Exclusion and Property Rules in the Law of Nuisance, 90 Va. L. Rev. 965 (2004). 109   Sherwood 48 Assocs., 76 Fed. Appx. at 389. 110   See, e.g., The Shelburne v. Crossan Corp., 95 N.J. Eq. 188 (N.J. Ch. 1923); Kristen M. Ploetz, Light Pollution in the United States: An Overview of the Inadequacies of the Common Law and State and Local Regulation, 36 New England Law Rev. 987 (2002). 111   Some of the plaintiff’s hotel rooms were illuminated by the defendant’s lights emanating from a bright billboard. Shelburne, 95 N.J. Eq. at 189. 112   Hildebrand v. Watts, 1997 Del. Ch. Lexis 32 (Del. Ch. 1997).

BARFIELD PRINT.indd 435

13/11/2018 11:53

436  Research handbook on the law of virtual and augmented reality has since settled.113 In California, a sober-living facility where sex offenders resided had reportedly been made a designated Pokémon Go stop. The facility manager reportedly stated: “I have no idea what Pokémon is. I have no idea who put the stop—if it was sabotage—because we don’t want kids showing up here.”114 In the US, a class action suit was filed consolidating various separate actions, including Jeffrey Marder v. Niantic, Inc. in the Northern District of California, claiming that the Pokémon Go game encourages nuisance, trespass, and unjust enrichment.115 The property owners contended that Niantic “encouraged and rewarded” players for congregating near their property and for “unauthorized entry”116 because the augmented reality game designated those areas as Pokéstops and Pokémon gyms and directed players there. The game developer contended that the plaintiffs’ claims for trespass and nuisance should fail for several reasons, including that it had required players to agree that they would not trespass and that placing virtual items on a virtual map for game players to find did not qualify as “unauthorized entry” onto private property. Plaintiffs countered by claiming that Niantic “had every ability to abate the nuisance” and stop players from trespassing by responding to their repeated pleas to relocate PokéStops and Pokémon gyms away from their properties, which the company failed to do.117 The plaintiffs further argued that after having placed Pokémon and other virtual objects on private property without seeking permission, Niantic continued to maintain and promote and profit from the Pokémon Go game, which inherently encouraged and rewarded players for traveling to those objects whether they were on private property or not; it was alleged that by taking these actions, Niantic not only foresaw but actively encouraged nuisance and trespass. In 2017, the district court denied class certification on the ground that the proposed class was too broad and that the plaintiffs had failed to show that the alleged damages exceeded the minimum requirement for federal jurisdiction.118 The judge raised the jurisdiction issue on his own initiative. Niantic’s attorneys had sought to dismiss the case on the grounds that the game developer was not liable for the actions of third parties it did not control, and that it had provided sufficient instructions to users not to trespass. Plaintiffs were given the 113   Associated Free Press, Dutch Freeze Court Action against Pokémon Go Makers (2016), https:// phys.org/news/2016-10-dutch-court-action-Pokémon-makers.html (last visited May 2, 2017). 114   Veronica Rocha, 2 California Men Fall Off Edge of Ocean Bluff While Playing Pokemon Go, Los Angeles Times (Jul. 14, 2016, 3:45 PM), www.latimes.com/local/lanow/la-me-ln-pokemongo-players-stabbed-fall-off-cliff-20160714-snap-story.html. 115   Allison Grande, “Pokémon Go” Maker Encouraged Trespassing, Court Told (March 3, 2017), www.law360.com/articles/898098. 116   Id. See also In re: Pokémon Go Nuisance Litigation, case number 3:16-cv-04300, in the U.S. District Court for the Northern District of California. 117   Allison Grande, “Pokémon Go” Maker Encouraged Trespassing, Court Told (March 3, 2017), www.law360.com/articles/898098. 118   Cara Bayles, “Pokémon Go” Maker Escapes Trespass Class Action, For Now,” Law360, July 27, 2017 (case is In re Pokémon Go Nuisance Litigation, case number 3:16-cv-04300 N.D. Calif.). The plaintiffs filed a stipulation of voluntary dismissal with prejudice against the defendant The Pokémon Company (“TPC”) on July 19, 2017. TPC’s lawyers had argued that “unjust enrichment” is not an independent cause of action under California law, and that TPC, as licensee, only received revenue from the game, but plaintiffs had not alleged that TPC had anything to do with the selection of PokéStops. Def. Motion to Dismiss, pp.3–6, filed Jan. 27, 2017, In re Pokémon Go Nuisance Litigation, case number 3:16-cv-04300 N.D. Calif.

BARFIELD PRINT.indd 436

13/11/2018 11:53

Advertising legal issues in virtual and augmented reality  437 opportunity to file a new complaint and as of this writing the case was proceeding with a trial date scheduled for late 2019. In response to these types of complaints, in the summer of 2016 Niantic added a support form online so people could request that PokéStops and Gyms be added or removed from the game.119 Users can fill out a form to “Request removal of a PokéStop or Gym” because of “Safety Concerns,” “Private Property,” or “Other.”120 From a practical perspective, if Niantic responds promptly to such requests for removal, it can reduce the likelihood of a dispute rising to the level of a lawsuit. The legal issues raised in the Pokémon Go class action may be applied directly to augmented reality advertising and promotions if such campaigns also encourage consumers to go to particular physical-world locations in order to be rewarded with product discounts or prizes. For example, Nissan ran a social media scavenger hunt to promote the launch of the Altima brand in 2015,121 and Delta ran a scavenger hunt with Snapchat in 2016.122 It is easy to see how an augmented reality scavenger hunt could be used as a brand promotion or advertising campaign. Based on the lessons from the Pokémon Go disputes, a risk reduction strategy should include, at a minimum: (1) terms of use and instructions to the participants to be aware of their surroundings and not enter private property, and (2) the opportunity for private property owners to request that their property be removed from the activity. Having outlined trespass and nuisance issues, we now turn to intellectual property issues in virtual and augmented reality advertising. Such issues can involve trademark/ trade dress, copyright, patent and trade secret.

4. INTELLECTUAL PROPERTY: TRADEMARK, COPYRIGHT, PATENT, AND TRADE SECRETS 4.1  Trademark/Trade Dress Brands and logos are commonly used in traditional advertising, and therefore are likely to be used in virtual and augmented reality environments as well. Such use can raise novel questions. For example, one issue may be whether the appearance of an augmented reality image or logo when interacting with a real-world image or advertisement constitutes trademark infringement. Another issue might be whether use of buildings or physicalworld objects and images in a virtual environment violates the trademark rights of the owners of such buildings, objects, or images.

119   Jeff Grubb, Pokémon Go: How to Request or Remove a PokéStop or Gym, VentureBeat (Jul. 14, 2016, 11:30 AM), https://venturebeat.com/2016/07/14/pokemon-go-how-to-request-or-rem​ ove-a-pokestop-or-gym/. 120  https://support.Pokémongo.nianticlabs.com/hc/en-us/requests/new?ticket_form_id=341148. 121   Lindsey Stein, Nissan Launches Scavenger Hunt to Promote Altima Brand, PR Week (May 13, 2015), www.prweek.com/article/1346887/nissan-launches-scavenger-hunt-promote-altima-brand. 122   Alesandra Dubin, Delta’s New Brand Promotion Is a Social Media First, Bizbash (Sep. 14, 2016, 7:00 AM), www.bizbash.com/deltas-new-brand-promotion-a-social-media-first/los-angeles/ story/32789/#.WXI6isbMw_V.

BARFIELD PRINT.indd 437

13/11/2018 11:53

438  Research handbook on the law of virtual and augmented reality Trademark protection is available under federal and state registration systems, as well as the common law.123 A trademark is a word, phrase, symbol, or design (or a combination thereof) that is used to identify and distinguish the source of goods or services.124 “Trade dress” may be the packaging, the attendant displays, and even the configuration of the product itself. For a service, it may be the decor or environment in which a service is provided—for example, the distinctive decor of the Hard Rock Cafe restaurant chain, or a particular stun gun used in a video game.125 So-called “nontraditional” marks, such as sounds, colors, and scents,126 can be protected as trademarks under existing US law (although obtaining such protection in practice can be challenging).127 Virtual or augmented reality marks themselves (including hologram images and marks that are in motion)128 should be entitled to trademark protection provided they meet the other trademark requirements, including, for example, distinctiveness and lack of functionality, and serve as a source identifier. Trademark infringement happens when there is a “likelihood of confusion” as to source or sponsorship, affiliation, or endorsement, considering a variety of factors. Such factors include, for example, similarity of the original mark and the allegedly infringing mark in sound, appearance, and meaning; similarity of the goods and services; and whether there has been any evidence of actual confusion.129 Brands recognized as “famous” can also receive additional protection from “tarnishment” or “dilution,” which occurs when a similar mark is used in a manner that will cause a negative association with, or otherwise weaken, the original mark. Trademark rights can last indefinitely if the mark continues to be in use. 123   For the purposes of this discussion, we will use the term “trademark” interchangeably to refer not only to trademarks for goods, but also to service marks for services and to trade dress. 124   15 U.S.C. Sec. 1127. 125   Are Augmented Reality Games Liable for Depiction of Buildings, Trademarks or Artwork? Frankfurt Kurnit Klein + Selz PC (Jul. 27, 2016), http://fkks.com/news/are-augmentedreality-games-liable-for-depictions-of-buildings-trademarks-o; David K. Stark, Note, Grand Theft Architecture: Architectural Works in Video Games After E.S.S. Entertainment v. Rockstar Games, 25 Berkeley Tech. L.J. 429; Brittany Frandsen, Comment: Is Using Call of Duty in This Comment Infringement? 2016 B.Y.U.L. Rev. 295 (2016) (discussing trademark issues for products used in video games). 126   Erin M. Reimer, Note, A Semiotic Analysis: Developing a New Standard for Scent Marks, 14 Vand. J. Ent. & Tech. L. 693 (Spring, 2012). 127   For example, General Mills has thus far (at least as of this writing) been unsuccessful in its efforts to register the yellow color for its Cheerios cereal box. See U.S. Trademark Application No 86757390. 128   Lesley Matty, NOTE, Rock, Paper, Scissors, Trademark? A Comparative Analysis of Motion as a Feature of Trademarks in the United States and Europe, 14 Cardozo J. Int’l & Comp. L. 557 (Fall 2006). 129   See, e.g., AMF, Inc. v. Sleekcraft Boats, 599 F.2d 341 (9th Cir. 1979) (relevant factors include:

               

1. 2. 3. 4. 5. 6. 7. 8.

strength of the mark; proximity of the goods; similarity of the marks; evidence of actual confusion; marketing channels used; type of goods and the degree of care likely to be exercised by the purchaser; defendant’s intent in selecting the mark; and likelihood of expansion of the product lines.)

BARFIELD PRINT.indd 438

13/11/2018 11:53

Advertising legal issues in virtual and augmented reality  439 Generally, to prevail on a direct trademark infringement action under the Lanham Act, a plaintiff must prove that: (a) they own the valid mark; (b) the defendant used the mark in commerce130 without the plaintiff’s authorization; (c) the defendant used the mark (or an imitation of it) “in connection with the sale, distribution, or advertising” of goods or services; and (d) the defendant’s use of the mark is likely to confuse consumers. Imitation of the mark using virtual images should be actionable under the Lanham Act. Businesses may seek to display extra information or content when consumers are in or near their stores, or when product packaging is scanned, as has already been happening with beacons and QR codes,131 or in other locations where they want to target potential sales, as might be the case for a running shoe company at a marathon. One possible area of contention might be businesses seeking to stop competitors from running virtual reality or augmented reality ads near locations, events or products associated with the original business’ brands. Such activities are called “ambush marketing” in the physical world when, for example, a competitor runs a promotion related to a sports event that is officially sponsored by another entity,132 or runs Olympics-themed campaigns in connection with the Olympics but is not a sponsor. A nonprofit or political organization may seek to stop negative augmented reality advertising near their offices or ads, such as the prohibition of “geofencing” around Massachusetts healthcare facilities.133 Such disputes would be roughly analogous to the existing ambush marketing cases, and to cases involving the use of keywords triggering competing web search results. Google’s keyword advertising provides a useful example. Google places paid ads at the top and sides of its search results page as “sponsored links.” In 2004, Google began allowing advertisers to bid on trademarked words. In 2009, Google changed its AdWord policy to allow use of trademarks both in advertising text and as advertisement keywords. Google allowed competitors to pay to “purchase” the right to appear as search results when a user typed in a search for the competing brand. For example, Pepsi could pay to appear first in the Google results for users searching for “Coke”. In response to the changed AdWord policy, Rosetta Stone filed suit against Google for violating the Lanham Act, claiming Google was “helping third parties to mislead consumers and misappropriate the Rosetta Stone marks by using them as ‘keyword’ triggers for paid advertisements

130   Use of the mark “in commerce” and likelihood of consumer confusion are key factors that would mitigate trademark infringement issues if an individual user-generated personal content on their own initiative versus using preloaded content. For a discussion of intellectual property issues involving user generated content, see Deborah Halbert, Mass Culture and the Culture of the Masses: A Manifesto for User-Generated Rights, 11 Vand. J. Ent. & Tech. L. 921 (2009); Carl Michael Szabo, Thwack!! Take That, User-Generated Content!: Marvel Enterprises v. NCSoft, 62 Fed. Comm. L.J. 541 (2009–10) (discussing copyright liability issues). 131   Kate Kaye, Ad Dollars Help Fuel Proliferation of Beacons and Proximity Sensors, Advertising Age (Feb. 10, 2017), http://adage.com/article/datadriven-marketing/ad-dollars-fuel-proliferation-bea​ cons-proximity/307951/. 132   Todd Wasserman, Hacking the Super Bowl: Get Your Brand Noticed for Less Than $10m a Minute, PR Week (Jan. 28, 2016), www.prweek.com/article/1381332/hacking-super-bowl-brandnot​iced-less-10m-minute. 133   Maura Healey, AG Reaches Settlement with Advertising Company Prohibiting “Geofencing” around Massachusetts Healthcare Facilities, Mass.gov (Apr. 04, 2017), www.mass.gov/ago/newsand-updates/press-releases/2017/2017-04-04-copley-advertising-geofencing.html.

BARFIELD PRINT.indd 439

13/11/2018 11:53

440  Research handbook on the law of virtual and augmented reality and by using them within the text or title of paid advertisements.” Rosetta Stone alleged direct trademark infringement, contributory trademark infringement, vicarious trademark infringement, and trademark dilution. The Fourth Circuit affirmed the grant of summary judgment for the vicarious trademark infringement claim and upheld the motion to dismiss for unjust enrichment. On appeal, only the “likelihood of confusion” prong was in dispute. In the Fourth Circuit, the likelihood of confusion is determined by looking at nine factors: (1) the strength or distinctiveness of the plaintiff’s mark as actually used in the marketplace; (2) the similarity of the two marks to consumers; (3) the similarity of the goods or services that the marks identify; (4) the similarity of the facilities used by the mark holders; (5) the similarity of advertising used by the mark holders; (6) the defendant’s intent; (7) actual confusion; (8) the quality of the defendant’s product; and (9) the sophistication of the consuming public. Each of these factors could, of course, be used by the Fourth Circuit in determining the likelihood of confusion resulting from the use of virtual images in an advertising context. The appellate court overturned a grant of summary judgment for Google. The appellate court indicated that Google’s AdWords keyword search program could amount to direct infringement, contributory infringement, or trademark dilution.134 The Google AdWords program has also been challenged under a theory of “initial interest confusion,”135 which may have direct relevance to virtual reality and augmented reality advertising. In situations involving initial interest confusion, a consumer’s initial attention is captured by the use of a trademark without permission, but by the time the consumer actually is ready to make the purchase, they are aware of the source of the goods or services. For example, a virtual reality or augmented reality ad that uses a particular brand name or product image appears (for example, a Chick-Fil-A restaurant logo) to attract the viewer of the ad to enter the virtual or physical space to purchase a competitor’s product or service (for example, Kentucky Fried Chicken). One of the leading cases on the topic of initial interest confusion was Brookfield v. West Coast Entertainment, Inc.,136 in which a West Coast video store registered the domain name “moviebuff.com,” which used its competitor Brookfield’s trademark “MovieBuff.” The court analogized the use of the brand of a competitor in a domain name or metatags to the physical-world equivalent of a false billboard on the highway: Using another’s trademark in one’s metatags is much like posting a sign with another’s trademark in front of one’s store. Suppose West Coast’s competitor (let’s call it “Blockbuster”) puts up a billboard on a highway reading—“West Coast Video: 2 miles ahead at Exit 7”—where West Coast is really located at Exit 8 but Blockbuster is located at Exit 7. Customers looking for West Coast’s store will pull off at Exit 7 and drive around looking for it. Unable to locate West Coast, but seeing the Blockbuster store right by the highway entrance, they may simply rent there. Even consumers who prefer West Coast may find it not worth the trouble to continue searching for

  Rosetta Stone v. Google, 676 F.3d 144 (4th Cir. 2012).   Google Inc. v. Am. Blind & Wallpaper Factory, Inc., 2005 U.S. Dist. LEXIS 6228, 74 U.S.P.Q.2nd (BNA) 1385 (N.D. Cal. Mar. 30, 2005) (see footnotes 21 and 24) (denying motion to dismiss claims of contributory trademark infringement and contributory dilution but dismissing claim of tortious interference with prospective business advantage). 136   Brookfield Communications, Inc. v. West Coast Entm’t Corp., 174 F.3d 1036, 1063-65 (9th Cir. 1999). 134 135

BARFIELD PRINT.indd 440

13/11/2018 11:53

Advertising legal issues in virtual and augmented reality  441 West Coast since there is a Blockbuster right there. Customers are not confused in the narrow sense: they are fully aware that they are purchasing from Blockbuster and they have no reason to believe that Blockbuster is related to, or in any way sponsored by, West Coast. Nevertheless, the fact that there is only initial consumer confusion does not alter the fact that Blockbuster would be misappropriating West Coast’s acquired goodwill.137

The court cited Blockbuster Entertainment Group v. Laylco, Inc.,138 in which a video rental store (Video Busters) attracted customers by using a sign confusingly similar to that of its competitor Blockbuster sign. In granting a preliminary injunction to Blockbuster, the court stated: That a customer would recognize that Video Busters is not connected to Blockbuster after entry into a Video Busters store and viewing the Video Busters membership application, brochure, video cassette jacket, and store layout is unimportant. The critical issue is the degree to which Video Busters might attract potential customers based on the similarity to the Blockbuster name. The court finds that Video Busters might attract some potential customers based on the similarity to the Blockbuster name. Because the names are so similar and the products sold are identical, some unwitting customers might enter a Video Busters store thinking it is somehow connected to Blockbuster. Those customers probably will realize shortly that Video Busters is not related to Blockbuster, but . . . that is irrelevant.139

The court further noted: “because it would be inconvenient to leave one video store to find another, those customers lured to a Video Busters store may rent video cassettes from that store despite having realized that it bears no connection to Blockbuster.”140 Rather than treating initial interest confusion as a separate doctrine, it may be better to incorporate considerations of initial interest confusion into one or more of the traditional trademark factor analyses for trademark infringement.141 Moreover, if a mark is “famous” under the Federal Trademark Dilution Act of 1995,142 initial interest confusion may be relevant to a claim of trademark dilution by blurring, which does not require a showing of a likelihood of consumer confusion as to source. Either way, the theory of initial interest confusion could certainly be relevant to situations in which a competitor’s trademark or trade dress was used in a virtual reality or augmented reality advertisement to attract consumer interest. Another example of trademark issues involves companies that want to stop customers or the public from associating bad ratings or reviews with their products. For example,   Brookfield, 174 F.3d at 1064.   Blockbuster, 869 F. Supp. 505 (E.D. Mich. 1994). 139   Blockbuster, 869 F. Supp. at 513 (citations omitted). 140   Blockbuster, 869 F. Supp. 505 at footnote 2. 141   Paul F. Rush, Note, Clearing Up the Confusion: Aiding Court Decisions Regarding Trademark Infringement and the Internet, 43 Stetson L. Rev. 637, 677 (2014) (provides a helpful summary of the evolution of initial interest doctrine and internet law). 142   15 U.S.C. Sec. 1125(c) (“the owner of a famous mark that is distinctive, inherently or through acquired distinctiveness, shall be entitled to an injunction against another person who, at any time after the owner’s mark has become famous, commences use of a mark or trade name in commerce that is likely to cause dilution by blurring or dilution by tarnishment of the famous mark, regardless of the presence or absence of actual or likely confusion, of competition, or of actual economic injury”). The Blockbuster case was decided before the Federal Trademark Dilution Act took effect. 137 138

BARFIELD PRINT.indd 441

13/11/2018 11:53

442  Research handbook on the law of virtual and augmented reality a Canadian court granted United Airlines an injunction against the website untied.com, which cataloged criticisms and complaints against the airline.143 In the United States, the website operator would probably have had a colorable First Amendment defense, but it is difficult to predict when such a defense will succeed.144 An augmented reality experience that was critical of a brand was an art installation called “The Leak in Your Home Town.” In the art installation, augmented reality software was used so that when visitors pointed an iPhone at a British Petroleum logo installed on the floor, the viewscreen ignited with the image of a swirling toxic oil plume in Flash animation, which leapt out of the corporation’s once benign-looking green-and-yellow sunburst. Leak . . . will work on any BP logo, anywhere, and [the artists] were deciding whether to make the smartphone application that runs it available to the general public for “next-to-free.”145

The case law for Lanham Act disputes resulting from the digital alteration of images is sparse, and the authors could not locate any decisions directly with Lanham Act claims involving augmented reality advertisements. Nevertheless, Sherwood 48 Associates, also discussed above regarding trespass claims, may provide some insight into how the courts may decide Lanham Act disputes involving augmented reality advertising. In Sherwood 48 Associates,146 a Lanham Act, Section 43(a) claim was brought in a case involving digitally replaced billboard advertisements viewed as part of the Spider-Man movie.147 The Sherwood 48 Associates plaintiffs also asserted trademark and trade dress violations,148 unfair competition and false endorsements. The plaintiffs contended that the digital images confused film viewers as to their several buildings’ associations with the digitally inserted advertisements.149 The Second Circuit Court of Appeals held that the digitally modified billboard advertisements did not infringe the building owners’ trademark rights, absent any showing of relevant consumer confusion.150 In addition, since the plaintiffs were constantly changing the content of their physical billboards, the court found that the digitally modified billboard advertisements on buildings did not infringe the building owners’ trade dress.151 In determining whether there was a cause of action for trade dress 143   Helen Coffey, United Airlines: Meet the Man Waging a 20-Year War Against Company that Ignored his Complaint, The Independent, June 28, 2017 (“United filed a lawsuit five years ago for trademark infringement (Untied.com featured the United globe logo with a frowning face and the tagline ‘Evil Alliance member’ instead of ‘Star Alliance member’), adding that the site could fool customers into thinking they are on United’s official website”), available at www.independent. co.uk/travel/news-and-advice/united-airlines-flights-20-years-untied-parody-complaint-website-jer​ emy-cooperstock-court-injunction-a7811921.html. 144  Compare Dr. Seuss Enters., LP v. Penguin Books USA, Inc., 109 F.3d 1394 (9th Cir. 1997) (holding “The Cat NOT in the Hat! A Parody by Dr. Juice” book criticizing the O.J. Simpson trial was infringing and granting an injunction) with Lombardo v. Dr. Seuss Enters., LP (Case No. 17-2952) (2nd Cir. 2018) (holding a play “Who’s Holiday” involving a raunchy version of “The Grinch” was fair use). 145   Review of “Tunneling” group art show, Nov. 16, 2010, https://theleakinyourhometown. wordpress.com/press/. 146   Sherwood 48 Assocs., 76 Fed. Appx. 147   Id. 148   Id at 391. 149   Id. 150   Id at 392. 151   The issue in Sherwood was determining the elements of the trade dress.

BARFIELD PRINT.indd 442

13/11/2018 11:53

Advertising legal issues in virtual and augmented reality  443 infringement, the court relied upon First Amendment law and concluded that the digitally altered ads were central to a major scene in the film existing for theatrically relevant and artistic purposes.152 The court found that Sony had the right to alter the setting of the scene for the primary objective of advancing the filmmaker’s creative purpose.153 In the court’s opinion, the First Amendment rights protecting fictional artistic expression outweighed any arguments based on real-world commercial interests.154 Under the court’s reasoning, non-permanent advertisements can be digitally altered in new works, provided that such alteration primarily serves the creative endeavors of their work. However, even though the Spider-Man movie was “for profit,” works such as books, movies, and TV shows are generally given stronger First Amendment protection than so-called commercial speech in advertisements. Thus, a virtual or augmented reality advertiser who alters buildings, objects, or other content (including other ads) may have less success in the courts than Sony. In virtual and augmented reality advertising, competitors may use variants of ambush marketing approaches to try to attract business or to denigrate the other business’s products, with implications for both trademark infringement and dilution. Courts will need to decide whether or not use of trademarked words, packaging, or designs, or geolocation to determine presence near a competitor store or at sponsored events, to trigger augmented reality popups is likely to cause confusion, or dilutes or tarnishes a trademark holder’s brands. If the court’s decision in Sherwood 48 Associates is followed in a dispute directly involving augmented reality advertising,155 the likelihood of consumer confusion will be a key issue, as will the court’s assessment of whether the ad is “commercial speech” from a First Amendment perspective. 4.2 Copyright Just like print or other audiovisual ads, virtual and augmented reality ads can be accused of infringing a copyright holder’s rights if such ads incorporate copyrighted images,156 video, music,157 or other protected works without permission. Copyright owners have the exclusive right to “reproduce the copyrighted work in copies” and to “prepare derivative works based upon the copyrighted work,” among other rights.158 These rights have already been invoked in disputes involving fictional virtual words in online simulation games such as in the lawsuit filed by Electronic Arts (“EA”), who accused Zynga’s game   Id at 377.   Id. 154   Id. 155   Id at 389. 156   Kimberly Culp and Shannon Yavorsky, Are You Prepared for the Legal Issues of Augmented Reality? (2017), www.allaboutadvertisinglaw.com/2017/08/are-you-prepared-for-the-legal-issuesof-augmented-reality.html. 157   Jon Blistein, Beastie Boys Settle Lawsuit over “Girls” Toy Commercial, Rolling Stone (Mar. 18, 2014), www.rollingstone.com/music/news/beastie-boys-settle-lawsuit-over-girls-toy-­commer​cial20140318; Jason Newman, Beastie Boys Win $1.7 Million in Monster Energy Copyright Lawsuit, Rolling Stone (June 5, 2014), www.rollingstone.com/music/news/beastie-boys-win-1-7-million-inmons​ter-energy-copyright-lawsuit-20140605. 158   17 U.S.C. § 106(1) and (2). 152 153

BARFIELD PRINT.indd 443

13/11/2018 11:53

444  Research handbook on the law of virtual and augmented reality The Ville game of infringing EA’s copyrighted online interactive social media game The Sims Social,159 or a copyright lawsuit related to the green crystal icon that appears over a character’s head in the game The Sims.160 There are even ad-versus-ad copyright lawsuits, such as the Nike case against Sega for Sega’s NBA 2k2 video game ad reenactment of Nike’s Frozen Moment ad with Michael Jordan.161 Similar copyright issues would probably apply to a virtual reality advertisement that was based on or incorporated another copyrighted work, such as a virtual reality version of the Sega NBA 2k2 ad or an image of a copyrighted sculpture.162 But perhaps the situation might be different for augmented reality ads? The answer might turn on the type of augmented reality technology used. Two display types are typically used in augmented reality: one that involves a see-through display that allows the person to view the world directly with the virtual ad projected in the space around the viewer, and another that involves real-time video of the world combined with computer-generated images. When a see-through display is used, the virtual image and physical object never physically interact; the virtual image only appears to be projected onto the surface of the physical object. Therefore, the virtual ad neither makes any permanent changes to the physical ad nor makes a separate copy of the physical ad, and once the augmented reality display is turned off, or the viewer looks in a different direction, the physical image disappears. Due to the intangible nature of augmented reality ads, a threshold question in any infringement analysis might be whether an augmented reality ad is a “copy” or a “derivative work.” The Copyright Act defines “copies” as “material objects . . . in which a work is fixed by any method now known or later developed, and from which the work can be perceived, reproduced, or otherwise communicated, either directly or with the aid of a machine or device” (emphasis added). Thus, in order for a virtual or augmented reality ad to be a “copy” of a copyrighted work, there must be some sort of material object in which the copyrighted work (or sufficient portion thereof) is fixed. In theory, if the virtual or augmented reality ad used source code, the drives or server(s) where the source code was stored could be the material objects where the ad was “fixed” by a method later 159   Jonathan Stempel, Electronic Arts, Zynga Settle Competing Lawsuits, Reuters (Feb. 15, 2013, 1:30 PM), www.reuters.com/article/us-electronicarts-zynga-lawsuit-idUSBRE91E0YG20130215; https://en.wikipedia.org/wiki/File:Ea_vs_zynga_lawsuit_screenshot_comparison.jpg (shows image of screens). 160   Hannah Madans, EA Video Game Maker Loses Appeal in Copyright Lawsuit v. O.C. Company, The Orange County Register (Sep 16, 2016, 1:49 PM), www.ocregister.com/2016/09/16/ ea-video-game-maker-loses-appeal-in-copyright-lawsuit-vs-oc-company/. 161   Sega, Nike Settle Lawsuit, USA Today (Apr 11, 2002), http://usatoday30.usatoday.com/ tech/news/2002/04/11/nike-sega.htm. 162   Lisa Rein, Sculptor Sues Postal Service over Mistaken Lady Liberty Stamp, Washington Post Dec. 3, 2013, available at www.washingtonpost.com/news/federal-eye/wp/2013/12/03/sculptorsues-u-s-postal-service-says-lady-liberty-stamp-of-his-statue-issued-without-his-permission/?utm_ term=.5cee743cc962 (“The Las Vegas sculptor who built the Western Lady Liberty is now suing the agency in U.S. Federal Court for copyright infringement. The Postal Service chose the image from an online photography service, Getty Images. But it did not acquire the rights to duplicate it from the artist, Robert S. Davidson, probably because postal officials assumed they were issuing a stamp with a famous image [in New York] that already was in the public domain . . . The lawsuit, filed last week, claims that Davidson’s sculpture ‘brought a new face to the iconic statue—a face which audiences found appeared more ‘fresh-faced,’ ‘sultry’ and ‘even sexier’ than the original”).

BARFIELD PRINT.indd 444

13/11/2018 11:53

Advertising legal issues in virtual and augmented reality  445 developed. This would seem to require a relatively sophisticated technical analysis and turn on the details of the storage of the ad; for example, it might not be treated as “fixed” storage if the ad was created on the fly based on motions of the viewer or using fuzzy logic with code stored in the cloud for a mobile device. For example, certain augmented reality systems have the capability to edit the video of the world captured in real time by a video camera.163 Ads that incorporated copyrighted works might not be deemed to be infringing “copies” in such situations, but they might instead be infringing “derivative works.” Thus, it would be a situation where the advertiser jumped out of the frying pan into the fire. Unlike the definition of “copies,” the definition of “derivative work” does not, at least on its face, require any material object or fixation. A “derivative work” is defined under the Copyright Act as “a work based upon one or more preexisting works, such as a translation, musical arrangement, dramatization, fictionalization, motion picture version, sound recording, art reproduction, abridgement, condensation, or any other form in which a work may be recast, transformed, or adopted.”164 However, in a case that may be something of an outlier, the Ninth Circuit has indicated that a derivative work must have “form” or “permanence” in order to infringe copyrights. Lewis Galoob Toys, Inc. v. Nintendo of America Inc.165 involved a physical device (the Game Genie) with incorporated software that plugged into Nintendo’s video game cartridges and game control deck. The Game Genie allowed a player to change the rules of the copyrighted Nintendo game being played—for example, to speed it up or slow it down.166 The Game Genie did not create a separate copy of the Nintendo game and did not make permanent changes to the original game. The effect of the Game Genie lasted only until the player unplugged the game or reset to start a new game. The issue in the case was whether use of the Game Genie infringed Nintendo’s copyrighted games by permitting users to make unauthorized derivative works.167 The Ninth Circuit found that it did not,168 reasoning that in order to infringe, the alleged derivative work must incorporate the underlying work in a concrete or permanent form. The Ninth Circuit affirmed the district court’s finding that the enhancement of Nintendo’s game display was not a derivative work because it was neither fixed nor embodied.169 The court reasoned that a derivative work must have “form” or permanence.170 The court referenced the fact that the examples of derivative works in Section 101 of the Copyright Act all physically incorporate the underlying work or works.171 If the court’s reasoning in Galoob is applied to virtual reality advertising, a see-through display used to view the combined virtual and physical ad might not be deemed to be

 Mann, supra note 23.   Id at § 101 (definition of derivative work). 165   Lewis Galoob Toys, Inc. v. Nintendo of America, Inc., 964 F.2d 965, 967 (9th Cir. 1992). 166   Id. 167   Id. 168   Id at 989. 169   Id. 170   Id. 171   These include translations, musical arrangements, abridgements, condensations, or any other form in which a work may be recast, transformed, or adopted: 17 U.S.C. § 101 (2006). 163 164

BARFIELD PRINT.indd 445

13/11/2018 11:53

446  Research handbook on the law of virtual and augmented reality a “derivative work” due to the lack of permanence whereas an augmented reality ad that combines live video of a physical ad with a virtual image might be deemed to be a “derivative work.” In both cases, the virtual ad never physically interacts with the physical ad, but in the live video situation the video of the physical ad is combined with the virtual ad and then viewed as one combined advertisement.172 The Galoob decision has not been applied broadly in the two decades since it was issued. It does not seem likely to offer a safe harbor for virtual and augmented reality ads when it comes to copyright infringement allegations. One other interesting aspect of copyright law to consider is that virtual and augmented reality ads may, at least in certain jurisdictions,173 be free to use otherwise copyrightable architectural works—public buildings—under an exception to US copyright law. Specifically, “the copyright in an architectural work that has been constructed does not include the right to prevent the making, distributing, or public display of pictures, paintings, photographs, or other pictorial representations of the work, if the building in which the work is embodied is located in or ordinarily visible from a public place.”174 The term “building” is defined by the Copyright Office as “humanly habitable structures that are intended to be both permanent and stationary, such as houses and office buildings, and other permanent and stationary structures designed for human occupancy, including but not limited to churches, museums, gazebos, and garden pavilions.”175 In the United States, this so-called “freedom of panorama” for buildings does not apply to art, however. Thus, a virtual reality or augmented reality image of a public building in an advertisement may be allowed, but not any copyrighted art appearing nearby. 4.3  Patent and Trade Secrets In addition to patent and trade secret issues relating to the underlying technology used in virtual reality and augmented reality advertising, there may also be issues related to the advertising itself. Because advertising is generally public in nature, as opposed to secret, trade secret issues are more likely to relate to behind the scenes strategies or algorithms, especially for targeting based on use of large data sets.176 The types of extreme personalization that “Big Data” allows are well suited for virtual and augmented reality targeted advertisements.177

  The combined image is also “fixed” in that it can be stored on the user’s wearable computer.   Rory Stott, Freedom of Panorama: The Internet Copyright Law that Should Have Architects Up in Arms, Arch Daily (Apr. 07, 2016), www.archdaily.com/785138/freedom-of-panorama-the-inter​ net-copyright-law-that-should-have-architects-up-in-arms. 174   17 U.S.C. § 120(a). 175   37 C.F.R. Sec. 202.11. 176   Frost & Sullivan, Big Data & Analytics, Virtual and Augmented Reality, Artificial Intelligence and Cloud Are Driving Universities to Innovate, Finds Frost & Sullivan, PR News Wire (Mar. 21, 2017, 2:26 PM), www.prnewswire.com/news-releases/big-data--analytics-virtual-and-augmentedreality-artificial-intelligence-and-cloud-are-driving-universities-to-innovate-finds-frost--sullivan-30​ 0426755.html. 177   Michael Wu, Personalization Comes to Retail with Big Data, IoT and Augmented Reality, CMS Wire (Nov. 10, 2016), www.cmswire.com/digital-experience/personalization-comes-to-retailwith-big-data-iot-and-augmented-reality/; Patrick Seitz, Advertising, Marketing to Invade Virtual, 172 173

BARFIELD PRINT.indd 446

13/11/2018 11:53

Advertising legal issues in virtual and augmented reality  447 An example of a virtual reality trade secret case is the dispute between the ZeniMax video game developer and the Facebook-owned Oculus VR LLC. The Oculus CTO was previously a software developer at ZeniMax and he allowed Oculus to use a virtual reality demonstration program he had written. A jury found Oculus liable for breach of a nondisclosure agreement, as well as copyright and trademark infringement, and ordered damages of $200 million for breach of the nondisclosure agreement alone.178 It is not surprising that confidential information regarding cutting-edge technologies such as virtual reality and augmented reality would be considered to be valuable trade secrets. They may also be the subject of patent protection, which would mean that trade secret protection would not exist once the patent application was published, at least for what was disclosed in the patent application itself. The US Court of Appeals for the Federal Circuit may be resistant to allowing design patent protection to three-dimensional digital models such as the models that would be used for virtual world or augmented reality objects.179 However, utility patents continue to be available. Niantic has several utility patents that are potentially relevant to technologies used in virtual and augmented reality advertising.180 Niantic’s US Patent No 8,968,099 entitled “System and Method for Transporting Virtual Objects in a Parallel Reality Game” has claims specific to gameplay but the written description in the patent includes potential advertising-related uses of the technology in such games, such as “an entity, such as a business, sponsor, advertiser, or other entity, can request the distribution of virtual objects based upon data associated with the location of a real-world carrier.” The description further states: “Other exemplary forms of verification can require that a player photograph the interior or exterior of the Philadelphia Train Station, scan a QR-Code located at a given location, or otherwise interact with a real-world vendor, advertisement or other suitable real world object.” The application offers examples involving a soda company, which might “find it desirable to distribute virtual coupons throughout the virtual world of the parallel reality game.” Alternatively, “a limited or unlimited number of virtual coupons can be generated and then transferred to or associated with the virtual counterpart of a soda delivery truck.” Although Niantic’s US Patent No 9,226,106 entitled “Systems and Methods for Filtering Communication Within a Location-Based Game” (the “’106 Patent”) has claims related to a game, the use of ads in such game may also be covered by the patent.

Augmented Reality in 2017, Investor’s Business Daily (Nov. 02, 2016), www.investors.com/news/ technology/click/advertising-marketing-to-invade-virtual-augmented-reality-in-2017/. 178   Jess Krochtengel, 5 Issues To Watch As Oculus Fights $500M ZeniMax Verdict, Law360, June 16, 2017. 179   See John R. Boule III, Comment, Redefining Reality: Why Design Patent Protection Should Expand to the World, 66 Am. U.L. Rev. 1113, 1157 (April 2017) (author contends: “Just as the USPTO expanded design patent subject matter to include software icons, the court should interpret articles of manufacture to include three-dimensional digital models. The inclusion of digital models will allow designers who are developing items for use in VR and AR worlds, as well as those designers whose items already exist in the physical world, to fully protect their intellectual property rights thus ensuring that statutory structure keeps pace with the boom in VR and AR technologies”). 180   See Andrew L. Rossow, Gotta Catch . . . A Lawsuit? A Legal Insight into the Intellectual, Civil, and Criminal Battlefield Pokémon Go Has Downloaded onto Smartphones and Properties around the World, 16 J. Marshall Rev. Intell. Prop. L. 329, 331–2 (2017).

BARFIELD PRINT.indd 447

13/11/2018 11:53

448  Research handbook on the law of virtual and augmented reality Finally, the summary section of Niantic’s US Patent No 9,128,789 entitled “Executing Cross-Cutting Concerns for Client-Server Remote Procedure Calls” (the “’789 Patent”) notes that the technology relates to “executing cross-cutting actions across a plurality of remote procedure calls in an interactive application, such as a location-based parallel reality game” but, unlike the ’106 Patent, the independent claims in the ’789 Patent are not limited to game applications. Niantic has itself been sued for allegedly infringing a utility patent related to “location interaction rules” using locations associated with a map display. In Location Based Services, LLC v. Niantic, Inc., filed on May 9, 2017, Niantic was accused of infringing US Patent No. 7,522,996 “Map Display System and Method,” with priority to April 30, 2005. Claim 1 of the patent is directed to: A method for providing map-related data, the method comprising: receiving a request for a map display illustrating information relative to one or more locations in a predetermined area; determining a status associated with at least one of the locations on the map display, the status being a function of one or more location interaction rules associated with at least one of the locations on the map display; and generating a signal indicating on the map display the status regarding a permitted traverse or visit that is allowed under an applicable location interaction rule associated with the at least one of the locations on the map display.

The patent holder, Location Based Services, LLC, alleged that examples of the ­“location interaction rules” used in Pokémon Go included, for example, determining whether a location was a PokéStop or checking whether a lure had been used at the PokéStop or whether the user last visited the location within the past five minutes.181 If Pokémon Go was found to infringe this patent, it is highly likely that any map-based use of a ­ugmented or virtual reality advertising or marketing would also infringe. Fortunately for Niantic, the Location Based Services patents were found to be invalid as being directed to an abstract idea with no inventive concept, and granted Niantic’s motion to dismiss with prejudice under the Alice Corp. v. CLS Bank guidance from the Supreme Court.182 In addition to patents on the underlying technology for augmented and virtual reality, there may also be patents related to specific applications of such technology for advertising and marketing. For example, the pending published utility patent application from Magic Leap Inc. entitled “Methods and Systems for Creating Virtual and Augmented Reality,” Application Number 14/738,877, Publication No US20160026253, describes, but does not yet (as of the time of this writing) claim, potential augmented reality advertising applications. For example, Figure 89F in the application shows an augmented reality display that appears to a mother and daughter shopping in a grocery store, optionally using an augmented reality device attached to the shopping cart or a wearable gesture-based recognition system (see Figure 14.1). 181   Complaint Para. 12, Location Based Services, LLC. v. Niantic, Inc., Civil Action No. 2:17cv-411, E.D. Texas—Marshall Division, filed May 9, 2017. 182   Order Granting Niantic’s Motion to Dismiss with Prejudice Under Alice, Case No. 17-cv04413 NC, N.D. Calif. Dec. 19, 2017 (applying Alice Corp. v. CLS Bank, 134 S.Ct. 2347 (2014)).

BARFIELD PRINT.indd 448

13/11/2018 11:53

Advertising legal issues in virtual and augmented reality  449

Source:  Available from the US Patent Office, USPTO.gov (Pub. No US20160026253—Figure 89F).

Figure 14.1  A drawing from Magic Leap’s published patent application Here is the Magic Leap patent application’s description of some potential advertising and marketing-related uses of the augmented reality advertising technology: The AR system may render virtual content which replicates a celebrity appearance. For example, the AR system may render a virtual appearance of a celebrity chef at a supermarket. The AR system may render virtual content which assists in cross-selling of products. For example, one or more virtual affects may cause a bottle of wine to recommend a cheese that goes well with the wine. The AR system may render visual and/or aural affects which appear to be proximate the cheese, in order to attract a shopper’s attention. The AR system may render one or more virtual affects in the field of the user that cause the user to perceive the cheese recommending certain crackers. The AR system may render virtual friends who may provide opinions or comments regarding the various produces (e.g., wine, cheese, crackers). The AR system may render virtual affects within the user’s field of view which are related to a diet the user is following. For example, the affects may include an image of a skinny version of the user, which is rendered in response to the user looking at a high calorie product. This may include an aural oral reminder regarding the diet.183

Further, Chinese Application CN 201610575220 entitled “Advertisement pushing method and advertisement pushing system based on augmented reality technology” is specifically 183   “Methods and Systems for Creating Virtual and Augmented Reality,” U.S. Patent Application No 14/738,877, PCT/US2015/035718, Publication No. US20160026253, application filed June 13, 2015.

BARFIELD PRINT.indd 449

13/11/2018 11:53

450  Research handbook on the law of virtual and augmented reality directed to augmented reality advertising. The first independent claim includes “obtaining a realistic image of the user profile acquisition; recognition of the reality of the object scene image appearing whether to include characteristics of the target and the target for the call ads feature; based on augmented reality technology to present the user with the advertisement.” Due to the relatively late priority date (July 20, 2016), if the application is allowed the final claims may well be narrower than the published claims, but the application still illustrates the fact that patent protection may be sought on augmented reality advertising itself. Even if potential consumers do not have an augmented or virtual reality device or have not enabled such functionality on their smartphones, facial recognition could be used to target virtual or augmented reality advertising,184 as envisioned in the dystopian movie Minority Report when Tom Cruise’s character is recognized (mistakenly as another person) when walking into a Gap store and an augmented reality hologram says “Hello, Mr. Yakamoto. Welcome back to the Gap. How did those assorted tank tops work out for you?”185 In a study released in the fall of 2016, Georgetown University found that “one in two American adults is in a law enforcement face recognition network”186 and it seems likely that commercial companies such as Facebook, with more than 1.65 billion users, have a vastly larger database of images of faces that are able to be recognized.187 According to Facebook, it can accurately identify a person 98 percent of the time.188 Patent applications have been filed related to use of facial recognition (or eye movements) and augmented or virtual reality advertising, such as US Patent Application No 13/342,949 filed by Osterhout Group Inc. and currently assigned to Microsoft Technology Licensing LLC, which is entitled “AR glasses with state triggered eye control interaction with advertising facility” and notes that an embodiment of the technology may include “an an embodiment, an apparatus for biometric data capture is provided. Biometric data may be visual biometric data, such as facial biometric data or iris biometric data, or may be audio biometric data.”189 Similarly, issued US Patent No 8,965,460 owned by IP Holdings Inc. entitled “Image and augmented reality based networks using mobile devices and intelligent electronic glasses” describes the following possible applications of the technology: “various microexpressions and emotions may be detected by the facial expressions of a person” and “a camera may perform lip reading to understand spoken words when audio is not available. This may enable a person wearing intelligent electronic glasses to read the lips of a person far across a room.”

184   Kalev Leetaru, From Cellphone Tracking to Facial Recognition: Minority Report Is Here, Forbes (Oct. 20, 2016, 1:38 PM), www.forbes.com/sites/kalevleetaru/2016/10/20/from-cellphon​ e-tracking-to-facial-recognition-minority-report-is-here/#1aa98dbb115a. 185   Clips are available on YouTube, for example at www.youtube.com/watch?v=ITjsb22-EwQ . 186  www.perpetuallineup.org/. 187   Naomi Lachance, Facebook’s Facial Recognition Software Is Different from the FBI’s. Here’s Why, NPR (May 18, 2016, 9:30 AM), www.npr.org/sections/alltechconsidered/2016/05/18/477819617/ facebooks-facial-recognition-software-is-different-from-the-fbis-heres-why. 188   Id. 189   Paragraph 29 of U.S. Publication No. US20120200601.

BARFIELD PRINT.indd 450

13/11/2018 11:53

Advertising legal issues in virtual and augmented reality  451 As indicated in the Location Based Services case, there is a potential weakness in any patents covering specific advertising methods, based on the US Supreme Court case Alice Corp. v. CLS Bank,190 which held that “abstract ideas” implemented using a computer were not eligible for patentability under Section 101 of the Patent Act. The Alice Corp. decision involved a patent directed to financial trading systems in which trades between two parties are settled by a third party in ways that reduce the risk that one party will perform, that is, make a payment, and the other will not perform. The decision was applied by the US Patent Office Patent Trial and Appeal Board to invalidate a web advertising patent relating to a system to detect invalid clicks in pay per click ads that was challenged by Google.191 Nintendo also successfully raised an Alice Corp. invalidity defense against a facial image recognition patent. Nintendo’s use of facial recognition to customize “Mii” avatars for the Wii game console was accused of infringing RecogniCorp’s US Patent No 8,005,303 entitled “Method and Apparatus for Encoding/Decoding Image Data.” In RecogniCorp, LLC. v. Nintendo Co., Ltd. and Nintendo of America Inc.,192 the Court of Appeals for the Federal Circuit affirmed the district court’s ruling that the patent was “directed to the abstract idea of encoding and decoding image data,” and did not “contain an inventive concept sufficient to render the patent eligible.” However, it is not safe for advertisers to assume that all relevant patents will be invalidated based on the Alice Corp. decision because it can be difficult to predict when the Alice case will be the basis for a rejection in the US Patent Office or in the courts. For example, the Magic Leap patent application described above was rejected by the US Patent Office based on prior art but not for a lack of subject matter patentability under the Alice Corp. precedent.193 Thus, virtual and augmented reality advertisers and marketers may be subject to patent infringement claims, as well as trade secret theft allegations.

5.  RIGHT OF PUBLICITY/PRIVACY A common legal issue in traditional advertising involves the use of celebrities in order to promote products or services, whether by use of their names, photos, or voices,194 or even indirect references such as an athlete’s sports jersey number, home state, height,

  Alice Corp. v. CLS Bank International, 573 U.S. __, 134 S. Ct. 2347 (2014).   Brite Smart U.S. Patent No 8,671,057; see Google Inc. v. Zuili, case number CBM201600008, before the Patent Trial and Appeal Board, discussed in Matthew Bultman, PTAB Nixes Ad Patent Claims under Alice in Google Challenge, Law360 (April 25, 2017). 192   RecogniCorp, LLC. v. Nintendo Co., Ltd. and Nintendo of America Inc., Civil Action 20161499, U.S. Court of Appeals for the Federal Circuit, Opinion Issued April 28, 2017. 193   See USPTO.gov file history for U.S. Patent Application No 14/738,877 Final Rejection issued on March 29, 2017; Request for Continued Examination filed on June 28, 2017. 194   Midler v. Ford Motor Co., 849 F.2d 460 (9th Cir. 1988) (after Midler declined Ford’s request to sing in a TV commercial, and Ford used a singer with a voice remarkably similar to Midler’s); 2–29 The Law of Advertising § 29.02 (2017) (“An economic value can be attached to many aspects of an individual’s (or sometimes a group’s) persona. It is irrelevant whether that value is derived from the name, physical likeness, or voice of a party; a value may still attach to it . . . Thus, use of a persona in an advertisement should be compensable, assuming, of course, that the persona is recognizable”) (footnotes omitted). 190 191

BARFIELD PRINT.indd 451

13/11/2018 11:53

452  Research handbook on the law of virtual and augmented reality weight, skin color, and skill set.195 Advertisers using virtual reality and augmented reality are likely to want to use celebrities also. For example, Magic Leap expressly contemplated use of celebrities in augmented reality advertising. Its patent application notes: “The AR system may render virtual content which replicates a celebrity appearance. For example, the AR system may render a virtual appearance of a celebrity chef at a supermarket.”196 The common law right of publicity is an additional right that a celebrity may establish in their name, in addition to their trademark rights. For example, the late actor Paul Newman had a right of publicity in his name and likeness, a federally registered trademark (technically a “service mark”) for providing “entertainment services comprising dramatic performances by an actor in movies and on television,” and a federally registered trademark “Newman’s Own” for his food products.197 The right of publicity is a common recourse for celebrities who seek either compensation or to prohibit use of their identity altogether. For example, Michael Jordan successfully sued the now-defunct grocery chain Dominick’s and was awarded $8.9 million. The chain had used Jordan’s name in a 2009 Sports Illustrated advertisement (on the inside cover) congratulating him for his induction in the Hall of Fame and including a clip-off coupon for a $2 discount on steak. Ironically, only two of the coupons were used because people were reluctant to damage the cover of the special-edition $10 version of the magazine.198 Jordan later pledged to donate the proceeds of the lawsuit to local Chicago charities,199 reportedly saying the lawsuit “was never about the money.”200 The right of publicity is protected at the state level by statute and/or common law. There is no national uniformity. California’s right of publicity statute provides:201 (a)  Any person who knowingly uses another’s name, voice, signature, photograph, or likeness, in any manner, on or in products, merchandise, or goods, or for purposes of advertising or selling, or soliciting purchases of, products, merchandise, goods or services, without such person’s prior consent . . . shall be liable for any damages sustained by the person or persons injured as a result thereof ” (emphasis added).

195   Pablo Mena, EA Sports and Ed O’Bannon Reach $60 Million Settlement in College Athletes Lawsuit (2016), https://usports.org/ea-sports-and-ed-obannon-reach-60m-settlement-in-college-athl​ etes-lawsuit/. 196   “Methods and Systems for Creating Virtual and Augmented Reality,” Application No 14/738,877. 197   For example, see U.S. Trademark Registration 3257796 (“PAUL NEWMAN” for “Entertainment services. . .”) and U.S. Trademark Registration No. 4889352 (“NEWMAN’S OWN” for “tea; candy”). 198   Kim Janssen, Michael Jordan’s $8.9M Verdict Offers Valuable Lesson for Firms, Athletes, Chicago Tribune (Aug. 22, 2015, 8:20 PM), www.chicagotribune.com/business/ct-michael-jordandominicks-analysis-0823-biz-20150822-story.html. 199   Kim Janssen, Michael Jordan Hands Court Settlement to 23 Chicago Nonprofits, Chicago Tribune (Dec. 15, 2015, 2:28 PM), www.chicagotribune.com/business/ct-michael-jordan-charity1216-biz-20151215-story.html. 200   Brenna Houck, Michael Jordan Scores $8.9M in Steak Coupon Lawsuit, Eater (Aug. 22, 2015, 5:02 PM), www.eater.com/2015/8/22/9192115/michael-jordan-scores-8-million-steak-cou​ pon-lawsuit. 201   California Code, Civil Code—CIV § 3344.

BARFIELD PRINT.indd 452

13/11/2018 11:53

Advertising legal issues in virtual and augmented reality  453 The statute defines “photograph” broadly, to include any “photogenic reproduction” (Sec. 3344(b)). Although “likeness” is not defined in the statute, the California courts have held that the image of a robot used in a Samsung ad campaign was not a “likeness” of Wheel of Fortune star Vanna White (see Figure 14.2; the robot is on the left and a photo of White is shown on the right) within the meaning of Section 3344(a).202 However, White was successful in her California common law right of publicity (which has roots in a right of privacy) claim that her “identity” had been misappropriated. The Ninth Circuit held that the California common law right of celebrities to protect their identity from being commercially exploited was broader than the statutory protection for their “likeness.”203 California also protects the right of publicity even after a celebrity’s death.204 While the damages awards are typically less than those awarded for celebrities (absent aggravating circumstances such as extreme invasions), ordinary people also have an equivalent to a right to publicity in the form of a right to privacy. Some examples include a lawsuit involving a photo of a high school student that was used without her permission in a Virgin Mobile ad campaign,205 a Kia ad using a photo of a tourist in San Francisco,206 and a woman’s lawsuit against Anheuser-Busch for using a Facebook photo she posted in an ad for Natural Light beer.207 Lawsuits also have been filed related to Facebook’s use of pictures of minors without their parent or guardian’s consent.208 “Clicking the ‘like’ button on an advertiser’s Facebook page allows a user’s name and photo to be seen by the user’s Facebook friends who visit the advertiser’s page. When a user ‘likes’ a particular

202  Note, White v. Samsung Electronics America: The Ninth Circuit Turns a New Letter in California Right of Publicity Law, 15 Pace L. Rev. 161 (Fall 1994). 203   White, 971 F.2d at 1398–9, referring to Carson v. Here’s Johnny Portable Toilets, Inc., 698 F.2d 831 (6th Cir. Mich. Feb. 1, 1983) (“Here’s Johnny!—Worlds Foremost Commodian” [sic] slogan for portable toilets found to violate Johnny Carson’s common law unfair competition and invasion of privacy and publicity rights). 204   Rhett H. Laurens, Year of the Living Dead: California Breathes New Life into Celebrity Publicity Rights, 24 Hastings Comm. & Ent. L.J. 109 (2001). 205   Noam Cohen, Use My Photo? Not Without Permission, The New York Times (Oct. 1, 2007), www.nytimes.com/2007/10/01/technology/01link.html (“One moment, Alison Chang, a 15-year-old student from Dallas, is cheerfully goofing around at a local church-sponsored car wash, posing with a friend for a photo. Weeks later, that photo is posted online and catches the eye of an ad agency in Australia, and Alison appears on a billboard in Adelaide as part of a Virgin Mobile advertising campaign”). 206   Bradley Johnson, Kia Ad Gets Carmaker Sued: Tourist Surprised His Image Used sans Permission, Advertising Age (Aug. 25, 1997), http://adage.com/article/news/kia-ad-carmaker-suedtour​ist-surprised-image-sans-permission/70997/. 207   Ben Hooper, North Carolina Woman Sues after Facebook Photo Used in Beer Ad Campaign, UPI (Feb. 23, 2017), www.upi.com/North-Carolina-woman-sues-after-Facebook-photo-used-inbeer-ad-campaign/1541487865665/. (Feb. 23, 2017) (“lawsuit accuses the defendants of copyright infringement, invasion of privacy, and violation of the right of publicity”). 208   Athima Chansanchai, Facebook Sued for Using Kids’ Pics in Ads, Today News (May 5, 2011), www.today.com/money/facebook-sued-using-kids-pics-ads-123460; www.nbcnews.com/ id/38884528/ns/technology_and_science-tech_and_gadgets/#.WVwAaoqQw6c (August 27, 2010) (“lawsuit filed Thursday in Los Angeles challenges a Facebook feature that allows members to note that they like an advertised service or product. Facebook broadcasts those endorsements to the user’s friends”).

BARFIELD PRINT.indd 453

13/11/2018 11:53

454  Research handbook on the law of virtual and augmented reality

Figure 14.2 The images in dispute for the Vanna White right of publicity case, taken from the Exhibits in the public court filings

BARFIELD PRINT.indd 454

13/11/2018 11:53

Advertising legal issues in virtual and augmented reality  455 company or product, this also appears in friends’ news feeds.”209 The Supreme Court of Canada recently ruled that a class action lawsuit against Facebook for use of names and images of up to 1.8 million Canadian users without consent in “Sponsored Stories” could proceed in Canada despite the fact that Facebook’s terms of use require that all lawsuits must be filed in California.210 Existing precedent regarding use of the identities of real people (whether famous celebrities or less famous college athletes) in video games is likely to be relevant for use in virtual reality advertisements or augmented reality advertisements. Most lawsuits involving right of publicity/privacy type claims and video games turn on whether the relevant portion of the general public can easily identify the person involved based on the image or other data provided, such as a slogan or catchphrase such as “Here’s Johnny” to reference Johnny Carson.211 Thus, they are highly fact specific. The Grand Theft Auto V video game and Electronic Arts college sports games provide useful examples. There have been two lawsuits involving celebrities who claimed their images were used in the Grand Theft Auto V game. First, the actress Lindsay Lohan filed a lawsuit in New York (her state of residence) alleging that the Grand Theft Auto V game used her likeness without her permission (Figure 14.3). Specifically, Lohan noted the similarity between the photo of her shown on the left in Figure 14.3 and the image of the video game character “Lacey Jonas” who was used in the marketing and advertising material for the game, even though Lacey Jonas had a relatively minor role in the game—she appears during a game mission when players have to help her escape paparazzi in the fictional city of Los Santos, San Andreas.212 As another example of a right of publicity action, reality television actress Karen Gravano, known for the show Mob Wives, alleged the character “Antonia Bottino” in Grand Theft Auto V was based on her image and life story. Gravano alleged that, in addition to using her image, portrait, voice, and likeness, the character uses the same phrases she uses; the character’s father mirrors Gravano’s own father; that the character’s story about moving out west to safe houses mirrors Gravano’s fear of being ripped out of her former life and being sent to Nebraska; that the character’s story about dealing with the character’s father cooperating with the state government is the same as Gravano dealing with the repercussions of her father’s cooperation; and that the character’s father not

209   Complaint, Nastro lawsuit against Facebook, cited in www.today.com/money/facebooksued-using-kids-pics-ads-123460 (May 5, 2011). 210  www.theglobeandmail.com/news/national/supreme-court-gives-thumbs-up-to-privacylawsuit-against-facebook/article35444477/ (June 23, 2017). 211   For a discussion of the right of publicity to virtual avatars, see Elina Slavin, Modern Application of the Right of Publicity to Virtual Avatars, Seton Hall Law Student Scholarship, Paper 78 (2010). 212   Mike Epstein, Lindsey Lohan Allowed to Sue Rockstar Games over GTA V Character, Court Says, Digital Trends (Mar. 17, 2016, 11:11 AM), www.digitaltrends.com/gaming/grandtheft-auto-v-lawsuit-moves-forward/; Ben Gilbert, Lindsey Lohan Doesn’t Look Like Anyone in “Grand Theft Auto V,” Rules New York State Court, Business Insider (Sep. 1, 2016), www. businessinsider.com/lindsay-lohan-loses-lawsuit-against-grand-theft-auto-makers-2016-9.; Lohan v. Take Two Interactive, Index No 156443/2014 (March 11, 2016) (N.Y. Supreme Court, Judge Kenney) (denied Motion to Dismiss on statute of limitations grounds—ruling available at https://www. documentcloud.org/documents/2757547-Lohan.html)

BARFIELD PRINT.indd 455

13/11/2018 11:53

456  Research handbook on the law of virtual and augmented reality

Source:  Available as Exhibits in the public court filing in Lohan v. Take-Two Interactive Software, Case No 156443/2014, Amended Complaint filed Oct. 8, 2014, Ex. B-3 and Ex. D-5

Figure 14.3  Lindsay Lohan photo (left) and video game marketing material image of “Lacey Jonas” (right), triggering a right of publicity action letting the character do a reality show is the same as Gravano’s father publicly decrying her doing a reality show.213

A five-judge panel of the New York Appellate Division dismissed both lawsuits on the ground that the New York Civil Rights Law Section 51 required that the “name, portrait or picture” be used. The court explained: [D]efendants never referred to Gravano by name or used her actual name in the video game, never used Gravano herself as an actor for the video game, and never used a photograph of her.214 As to Lohan’s claim that an avatar in the video game is she and that her image is used in various images, defendants also never referred to Lohan by name or used her actual name in the video game, never used Lohan herself as an actor for the video game, and never used a photograph of Lohan.215

The court further reasoned that even if the depictions were of the actual plaintiffs, as alleged, the game was a “work of fiction and satire” versus “advertising” or “trade” within 213   Gravano v. Take-Two Interactive Software, Inc., 2016 NY Slip Op 05942 [142 AD3d 776], September 1, 2016, Appellate Division, First Department, available at www.courts.state.ny.us/repor ter/3dseries/2016/2016_05942.htm. 214   See Costanza v. Seinfeld, 279 A.D.2d 255 (N.Y. App. Div. 1st Dep’t Jan. 4, 2001). 215   Id at 255.

BARFIELD PRINT.indd 456

13/11/2018 11:53

Advertising legal issues in virtual and augmented reality  457 the meaning of Section 51, and therefore were entitled to First Amendment protection. Finally, with respect to Lohan’s claim regarding the use of her image in advertisements, which would certainly be “advertising” within the meaning of Section 51, the court held: “The images are not of Lohan herself, but ‘merely the avatar’ in the game that Lohan claims is a depiction of her (see Costanza at 255 [the ‘use of the character in advertising was incidental or ancillary to the permitted use,’ and therefore was not commercial])”.216 On this ground the court seems to be reaching, given that the “Lacey Jonas” character had such a small role in the game itself. How could the “Jonas” character have been so important for the advertising and marketing materials if the character had no connection to Lohan? Unlike the situation in Costanza v. Seinfeld,217 the case upon which the court relied, where “George Costanza” was one of the four main characters in the entire Seinfeld television series, the “Jonas” character was not a major character in the Grand Theft Auto V game and therefore her use in advertising seems unlikely to have been “incidental.” Perhaps the case should have been allowed to go to the jury for a decision on the merits of whether the avatar image in the advertisements was sufficiently similar to Lohan’s image in her photograph?218 In any event, based on other cases, it does not seem safe for virtual or augmented reality advertisers to rely too heavily on the Lohan case as a safe harbor. Different outcomes were reached in lawsuits involving the Electronic Arts EA Sports games. In Keller v. EA, Arizona State quarterback Sam Keller filed a class action suit against Electronic Arts and the NCAA alleging that his image and likeness were used improperly.219 Reportedly, [a]lthough Keller’s name did not appear in the game, there was little doubt that he was the inspiration for the Arizona State quarterback in its 2005 edition. The virtual player shared Keller’s jersey number, 9, as well as his height, weight, skin tone, hair color and home state. The virtual quarterback even had the same playing style, as a pocket passer. “That was what made it so cool,” said Keller, who transferred to Nebraska in 2006. “It was so blatant.”220

The Keller litigation was settled for $20 million.221 Similarly, in O’Bannon v. NCAA, former UCLA college basketball player Ed 216   Gravano v. Take-Two Interactive Software, Inc., 2016 NY Slip Op 05942, p.2 [142 AD3d 776], September 1, 2016, Appellate Division, First Department, www.courts.state.ny.us/reporter/3dser ies/2016/2016_05942.htm/. 217   Costanza v. Seinfeld. 218   For an argument that the right of publicity should be strengthened for digital personas, see Anthony Pessino, Mistaken Identity: A Call to Strengthen Publicity Rights for Digital Personas, 4 Va. Sports & Ent. L.J. 86 (2004). 219   See Keller v. Electronic Arts, Inc., C 09-1967 CW, 2010 WL 530108, at 1 (N.D. Cal. Feb. 8, 2010) aff’d sub nom. In re NCAA Student-Athlete Name & Likeness Licensing Litig., 724 F.3d 1268 (9th Cir. 2013) (discussing similarities between videogame characters and “real-life college football athletes”). 220  http://forums.operationsports.com/forums/ncaa-football/332987-more-sam-keller-lawsuit. html. 221   See also Ben Bolas, EA Sports: If It’s In the Game, It’s In the Courtroom, Moorad Sports L. J. Blog (Aug. 27, 2013), http://lawweb2009.law.villanova.edu/sportslaw/?p=2312 (discussing implications of lawsuit against EA and NCAA’s decision not to renew contract with EA). Press Release, NCAA, NCAA Reaches Settlement in EA Video Game Lawsuit (June 9, 2014) ($20

BARFIELD PRINT.indd 457

13/11/2018 11:53

458  Research handbook on the law of virtual and augmented reality O’Bannon brought an antitrust class action against the NCAA on behalf of up to 100,000 current and former student athletes who had appeared on Electronic Arts’ EA Sports basketball and football games since 2003.222 The court ordered the NCAA to pay the class $46 million,223 and the case later settled.224 Although the EA Sports  cases  also involved antitrust claims against the NCAA, if the players’ names, images, or likenesses had not been used in the games there would have been no cause of action to begin with. Thus, the EA Sports cases offer a cautionary indication that just because an avatar is used in a virtual reality or augmented reality ad without the name of the individual, there may still be exposure to right of publicity/privacy claims.

6.  FALSE OR DECEPTIVE ADVERTISING The primary legal authority creating the right to sue for misleading and false actions is the Lanham Act225—the same statute that regulates the use of trademarks. In particular, Section 43(a)(1)(b) of the Act provides a cause of action against anyone who uses a “false or misleading description of fact, or false or misleading representation of fact . . . in commercial advertising or promotion [that] misrepresents the nature, characteristics, qualities, or geographic origin of . . . goods, services, or commercial activities.”226 False advertising claims may be filed by a regulatory agency or by a private party whose interests may be negatively affected by the challenged activity. The US Federal Trade Commission, for example, has wide latitude to take action against marketing techniques that it deems misleading or deceptive, even if they are not explicitly prohibited by law. Most states also have consumer protection statutes or common law rights that offer similar or even greater protections. For example, California’s Consumers Legal Remedies Act of 1970 restricts most type of misrepresentation and also requires guilty parties to pay all costs and legal fees.227 Furthermore, some states have statutes patterned after the Restatement of Unfair ­ illion ­ m settlement), cited in Meghan Rose Price, Casenote: O’Bannon v. National Collegiate Athletic Association: A Cinderella Story, 22 Jeffrey S. Moorad Sports Law Journal 179 (2015). 222   Alexander Lodge, Note: Who’s Afraid of the Big Bad NCAA? . . . The Ed O’Bannon v. NCAA Decision’s Impact on the NCAA’s Amateurism Model, 41 Iowa J. Corp. L. 775 (Spring, 2016). 223   Christian Dennie, Article: He Shoots, He Scores: An Analysis of O’Bannon v. NCAA On Appeal and the Future of Intercollegiate Athletics, 93 N.C.L. Rev. 90 (2015) (analyzing the district court’s opinion). 224   Original codefendants Electronic Arts and the Collegiate Licensing Company reportedly settled for $40 million. 225   15 U.S.C. §§ 1101 et seq. 226   15 U.S.C. § 1125. 227   Vachon Law Firm, Consumers Legal Remedies Act at http://thelemonlawcalifornia.com/ consumer-law-legal-resources/consumers-legal-remedies-act/ (last accessed May 2, 2017): “The California Consumers Legal Remedies Act (‘CLRA’) is the name for California Civil Code §§ 1750 et seq. The CLRA declare[s] unlawful several “methods of competition and unfair or deceptive acts or practices undertaken by any person in a transaction intended to result or which results in the sale or lease of goods or services to any consumer. Forbidden practices include misrepresenting the source of the good and services, representing reconditioned goods as new, advertising goods

BARFIELD PRINT.indd 458

13/11/2018 11:53

Advertising legal issues in virtual and augmented reality  459 Competition,228 which provides injunctive relief to any person “likely to be damaged” by the defendant’s false advertising. It is easy to imagine how a plaintiff could characterize virtual reality or augmented reality content as false advertising. The simplest way for potentially misleading advertising to occur in these media is by mistake, for example if the images are inaccurate and unrealistic. Virtual and augmented reality tools are still an emerging technology: Even the most cutting-edge augmented reality and virtual reality devices remain in their early stages and are still relatively expensive for most consumers. More accessible devices, such as the wearable mounts that enable virtual reality experiences on mobile phones, deliver impressive, but less precise, experiences. As sophisticated as all of these devices are, they have inherent limitations on how realistic they make virtual images appear. Consequently, ambitious augmented reality or virtual reality content creators may try to convey more data than they are able to effectively render, resulting in imprecise output. That, in turn, might end up conveying information that is false and has a material impact on a consumer. Recently, a company was found to have misrepresented the size of its product by packaging it in an overly large container.229 Under certain circumstances, an augmented or virtual image representation of a product may be unrealistically (and misleadingly) small or large in a way that affected a consumer’s purchasing decision. Advertisers might also take shortcuts during the complex task of creating virtual imagery for advertisements. For example, in 2012, British regulators banned L’Oréal from running ads containing two photos of Julia Roberts and Christy Turlington.230 L’Oréal’s marketers had digitally enhanced both photos to the point that it could not prove to the regulators’ satisfaction that the advertised makeup products were able to produce without having the expected demand in stock, representing a repair is needed when it is not, representing rebates that have hidden conditions, and misrepresenting the authority of a salesman to close a deal. The CLRA claim is attractive to potential plaintiffs because Cal. Civ. Code § 1780 allows consumers who suffer damage as a result of a practice declared unlawful by § 1770 to obtain actual damages (the total award of damages in a class action shall be more than $1,000); an order enjoining the methods, acts, or practices; restitution of property; punitive damages; court costs and attorney’s fees; and any other relief that the court deems proper. A prevailing plaintiff gets to recover his attorney’s fees, but a prevailing defendant usually may not recover his attorney’s fees”) (footnotes omitted). 228   RESTATEMENT (THIRD) OF UNFAIR COMPETITION § 7 (199); CAL. BUS. & PROF. CODE § 17200 (discussing the California state unfair competition law). 229   In Re: McCormick & Company, Inc., Pepper Products Marketing & Sales Practices Litigation, 2016 WL 6078250, No 15-cv-2188 (D.D.C. Oct. 17, 2016). 230   Stephanie Crosson, Regulating California Businesses: Are Current Laws Best Serving the State’s Economic Interests or Has the Sun Set on California? Photoshop Flawless: Is Excessive Digital Alteration of Commercial Photography Fraud and Deceptive Advertising? 20 Nexus J. Op. 67, 70–71 (2014/2015) (“In another controversy surrounding photo alteration in an advertisement, two advertising campaigns by cosmetics company L’Oreal were banned in the United Kingdom. The first was an ad for Lancome, a company owned by L’Oreal, which featured actress Julia Roberts and claimed to ‘recreate the aura of perfect skin.’ The second was an ad for Maybelline, a company also owned by L’Oreal, which featured model Christy Turlington and promoted a product called ‘The Eraser.’ The British Advertising Standards Authority was responsible for the banning of the advertisements. The Authority announced that it ‘could not conclude that the ad image accurately illustrated what effect the product could achieve, and that the image had not been exaggerated by digital post production techniques’”) (footnotes omitted).

BARFIELD PRINT.indd 459

13/11/2018 11:53

460  Research handbook on the law of virtual and augmented reality results like the ones shown. Fashion companies are likewise criticized on a regular basis for altering photos of clothing models to give them physical features so extreme as to be anatomically impossible. The difficulty of precise 3D rendering—not to mention the commercial and societal pressures that result from these photo alterations—could likewise lead to augmented or virtual reality ads that are similarly unrealistic. In the automotive realm, for example, a digital artist may leave out or oversimplify important features when rendering a product, in ways that mislead consumers into believing things about the product that are not accurate. By definition, digitally enhancing physical reality is a fundamental element of what is done by virtual and augmented reality. These types of situations, therefore, are ones that marketers using virtual images could experience, and that could result in legal disputes. For a Lanham Act Section 43(a) false advertising claim involving a different augmented reality ad placed in the same location as the original ad (similar to the virtual reality ads in the Spider-Man movie discussed previously), a plaintiff must essentially prove that “but for” the false advertising resulting from the augmented reality displays, it would have received the business that instead went to the false advertiser.231 Marketers, or the sponsors who pay substantial amounts of money to legitimately place their logos on billboards and other forms of advertisement, could pursue a cause of action against augmented reality advertisers under Section 43(a) of the Lanham Act. Further, a Lanham Act Section 43(a) action may be particularly relevant for disputes where defendant’s apparent projection of a virtual image onto a plaintiff’s billboard presents a false or misleading advertisement attributable to the owner of the physical billboard. At the very least, advertising in augmented reality is likely to cause confusion as to the particular sponsor’s affiliation with the goods virtually advertised on the plaintiff’s property. A sponsor forced to compete with virtual or augmented reality advertising may have standing to assert Lanham Act Section 43(a) claims because the public may perceive that the virtual ad is associated with the sponsor’s products. This would cause commercial injury to the sponsor by reducing the value of the sponsor’s ad on their property. The potential deceptive nature of augmented reality advertising in real space may also lend itself to trademark and trade dress infringement claims because the virtual image projected by the augmented reality display may change the appearance or meaning of the plaintiff’s ad. Further, a Lanham Act Section 43(a) claim may be particularly relevant if virtual images are used in “ambush advertising.”232 As noted earlier, ambush advertising occurs when a company, though not an official sponsor of an event, uses the event in order to draw attention to their ad.233 If the augmented reality advertiser creates the misleading impression that it is in fact affiliated with the event in some way (or is affiliated in a way other than as permitted by the event sponsor or promoter), a Lanham Act Section 43(a) action may exist. 231   See Mosler Safe Co. v. Ely-Norris Safe Co., 273 US 132, 139 (1927) (fact that defendant falsely represented that its safes had explosion chamber held not to establish that plaintiff maker of explosion chamber safes was injured, since plaintiff failed to prove that “customers had they known the facts would have gone to the plaintiff rather than to other competitors in the market”). 232   Stratford Homes, Inc., v. Lorusso, No. 94-CV-517E(M), 1995 WL 780977 (W.D.N.Y. 1995). 233   Ambush Marketing Definition, www.businessdictionary.com/definition/ambush-marketing. html (last visited Feb. 27, 2017).

BARFIELD PRINT.indd 460

13/11/2018 11:53

Advertising legal issues in virtual and augmented reality  461 In a review of federal cases dealing with the constitutionality of regulations for advertisements in video games,234 Seth Grossman concluded: “these cases hold that video games are considered highly protected speech for the purposes of First Amendment analysis only if they have certain characteristics such as narratives, themes, and sophisticated visual and auditory elements.”235 Lacking these elements, Grossman concluded,236 advertisements in video games will not be likely to receive the “heightened First Amendment protection given movies, books, and some video games,” but instead typically receive “the less restrictive standards for evaluating limitations on commercial speech.”237 The decision in Hoffman v. Capital Cities/ABC, Inc.238 suggests that this analysis is also likely to be applicable to other types of advertisements, in addition to video game ads. In the Hoffman lawsuit, the actor Dustin Hoffman claimed violation of his common law right of publicity, violation of the Lanham Act, and unfair business practices because of a Los Angeles Magazine (“LAM”) article that featured computer modified photographs of various actors, including Hoffman, wearing designer clothing.239 Hoffman argued that the photograph was a “false portrayal.” The court concluded that the speech consisting in part of altered images of Hoffman contained expressive elements, that it was not commercial speech despite the fact that there was a “Shopper’s Guide” at the back of the magazine and the manufacturer of the shoes had placed an ad in the magazine,240 and therefore that it was entitled to full protection under the First Amendment. The Court of Appeals for the Ninth Circuit explained: These facts are not enough to make the “Tootsie” photograph pure commercial speech. If the altered photograph had appeared in a Ralph Lauren advertisement, then we would be facing a case much like those cited above. But LAM did not use Hoffman’s image in a traditional advertisement printed merely for the purpose of selling a particular product. Insofar as the record shows, LAM did not receive any consideration from the designers for featuring their clothing in the fashion article containing the altered movie stills. Nor did the article simply advance a commercial message. “Grand Illusions” appears as a feature article on the cover of the magazine and in the table of contents. It is a complement to and a part of the issue’s focus on Hollywood past and present. Viewed in context, the article as a whole is a combination of fashion photography, humor, and visual and verbal editorial comment on classic films and famous actors. Any commercial aspects are “inextricably entwined” with expressive elements, and so they cannot be separated out “from the fully protected whole.”

The court’s analysis is consistent with Grossman’s conclusions on the regulation of advertisements which integrate commercial with noncommercial speech.241 That is, when artistic expressive elements are combined with noncommercial speech, the combined expression may be eligible for full protection under the First Amendment. 234   Seem e.g., Video Software Dealers Ass’n v. Maleng, 325 F. Supp. 2d 1180 (W.D. Wash. 2004); Am. Amusement Mach. Ass’n, 244 F.3d at 577–8; Wilson v. Midway Games, Inc., 198 F. Supp. 2d 167, 181 (D. Conn. 2002). 235  Grossman, supra note 16, at 228. 236   Id. 237   Id. 238   Hoffman v. Capital Cities/ABC Inc. et al., 255 F.3d 1180 (9th Cir. 2001). 239   Id at 1183. 240   Hoffman, 255 F.3d at ___. 241  Grossman, supra note 16.

BARFIELD PRINT.indd 461

13/11/2018 11:53

462  Research handbook on the law of virtual and augmented reality Where the virtual advertisement is combined with noncommercial speech and includes expressive elements, the more likely it is that the combined commercial and noncommercial speech will receive heightened protection from government regulations.242 However, it is still important that the ad not be deceptive or false. In the Hoffman case, the court found that the magazine article made it reasonably clear to readers that Hoffman’s image had been altered through their stated use of the “digital magic” of computers and because they showed the original still after the altered photograph. Under the second prong of the Central Hudson test,243 regulations restricting advertising that is deceptive generally are constitutional. The liability for false and deceptive advertising using virtual images may be substantial. For example, the financial liability for defendants found to have falsely advertised can include the plaintiff’s advertising costs to correct the misleading impression left by the false advertisement. Attorney fees and expenses are also recoverable.244 The use of virtual images for advertising has the potential to be especially misleading given how easy it is to manipulate digital information. For example, a virtual ad may be designed to cover the trademark or brand name of an ad,245 or designed to distort the message or appearance of a physical ad. A virtual ad projected in the real world that contained false or misleading information, under Central Hudson,246 would likely be subject to government regulation and, absent strong protected expressive content,247 would not receive full First Amendment protection.248

7.  OTHER ADVERTISING REGULATIONS 7.1  Government Regulation Any regulation of the use of virtual images for advertising should consider not only the current use of virtual ads inserted into film or television, but also the increasing capability of technology to insert virtual ads into the space surrounding the consumer. While both uses of virtual ads have the potential to distort messages and confuse consumers, they also 242   Stephen Totilo, Real-World “Pac-Man” Replaces Joystick with Virtual Goggles, www.vhl. com/news/articles/1504622/20050623/index.jhtml (last visited Dec. 11, 2005) (discussing the use of a head mounted display that allows Pacman to be projected within the real world). 243   In 1996, the Supreme Court refined the rules laid out in Central Hudson in a case known as 44 Liquormart, Inc. v. Rhode Island, 517 U.S. 484 (1996). In that case, the state of Rhode Island prohibited all advertising of prices by liquor stores. The case was heard on the basis of a challenge to the ban on price advertising. The court ruled that when a state entirely prohibits publication of truthful, nonmisleading commercial messages for reasons unrelated to preservation of the fair bargaining process, a judge may look to a wider application of the First Amendment than simple regulation of commercial speech. 244   See, e.g., U-Haul Intern. v. Jartran, Inc., 601 F. Supp. 1140, 1148-1149 (D. Ariz. 1984), affd, 793 F.2d 1034 (9th Cir. 1986). 245  Mann, supra note 23. 246   Cent. Hudson, 447 U.S. at 566. 247   An example of “commercial speech” with strong expressive content would be the trademark “The Slants” at issue in the Supreme Court decision in Matal v. Tam, Slip Op. No 15-1293 (June 19, 2017). 248   Such an ad would also be subject to a Lanham Act § 43(a) claim, see infra section 6.

BARFIELD PRINT.indd 462

13/11/2018 11:53

Advertising legal issues in virtual and augmented reality  463 have the potential to provide consumers with valuable comparative information to assist in buying decisions. Further, the use of virtual images projected into the environment or inserted into film or TV may be used for purposes other than advertising. Therefore, when promulgating rules to regulate virtual and augmented reality advertising, government regulators should consider whether current laws sufficiently protect the consumer, advertiser, and other parties who may use virtual images for nonadvertising purposes. For example, the digital alteration of the physical world may constitute a valid artistic statement that should be encouraged (First Amendment protection).249 In this case, overly restrictive regulations on the use of virtual images might constrain the ability of artists to use new media technologies to make a particular point or to better illustrate an issue of public importance. Further, any regulation of advertising using virtual images should not unnecessarily inhibit advertisers from using digital technology to convey creative and powerful commercial messages or entertaining content. The economic incentive provided by the ability to participate in advertising could be rendered meaningless if others could take existing images without permission and use and manipulate them at will, even if their alterations added some value to the original. It is also the case that the advertising industry is in many ways self-regulating, and to the extent that self-regulation works, government regulations will not be necessary.250 Indeed, there are several organizations involved in the advertising industry’s self-regulation effort; these include the National Advertising Division of the Better Business Bureau, Inc., its appellate body, the National Advertising Review Board, the Children’s Advertising Review Unit, the national television networks, and trade associations in many industries.251 However, even with self-regulation by the advertising industry, disputes involving advertising still make their way to court. Now that marketers have begun using virtual images for advertising, there will inevitably be disputes related to the use of such virtual images for advertising. Thus far, the United States has not adopted any specific legislation aimed at virtual or augmented reality advertising. As a possible indication of how the US government might treat advertising using virtual images, in response to a complaint filed by the consumer advocate group Commercial Alert on deceptive advertising on the internet, the FTC provided guidance to the search industry advocating disclosure of paid advertising embedded within search results.252 The FTC recommended that the search industry include clear and conspicuous disclosures on paid advertising in order to alert the consumer to what they are viewing when an unsolicited ad appears on their screen. This recommendation was compatible with the FTC’s already established rule on unfair and deceptive acts and

  See generally Mann, supra note 23.   Jeffrey S. Edelstein, Self-Regulation of Advertising: An Alternative to Litigation and Government Action, 43 IDEA 509 (2003). 251   Id at 510. 252   Letter to Gary Ruskin, Executive Director, Commercial Alert, from Donald Clark, Secretary, FTC (June 27, 2002) (“Although the staff is not recommending Commission action at this time, we are sending letters to search engine companies outlining the need for clear and conspicuous disclosures of paid placement, and in some instances paid inclusion, so that businesses may avoid possible future Commission action”), www.ftc.gov/sites/default/files/documents/clos​ ing_letters/commercial-alert-response-letter/commercialalertletter.pdf. 249 250

BARFIELD PRINT.indd 463

13/11/2018 11:53

464  Research handbook on the law of virtual and augmented reality practices affecting commerce.253 The idea of disclosure also seems to fit well with the paradigm of virtual advertising given the ease with which digital technology allows reality to be altered and how easy it is to include appropriate disclosures in digitally altered images. Although the FTC initially rejected the need for additional regulation for product placement advertising,254 more than a decade later the FTC issued a reminder to both social media “influencers” and brand marketers to “clearly and conspicuously disclose their relationships to brands when promoting or endorsing products through social media.”255 The FTC noted: [I]f there is a “material connection” between an endorser and an advertiser—in other words, a connection that might affect the weight or credibility that consumers give the endorsement—that connection should be clearly and conspicuously disclosed, unless it is already clear from the context of the communication. A material connection could be a business or family relationship, monetary payment, or the gift of a free product.256

Given the potential for marketers to use virtual advertising to deceive consumers, some countries have already decided to regulate virtual advertising. An examination of that effort may provide insight into how the United States may eventually regulate advertising using virtual images. The European Broadcasters Union (EBU) has produced a memorandum on virtual advertising, which provides a useful framework for the regulation of advertising using virtual images.257 The memorandum defines virtual advertising “as the use of electronic (imaging) systems which alter the broadcasting signal by substituting, or adding, venue advertising in the television picture.”258 While the rules of television ads may not apply directly to virtual advertising, virtual advertising is unlikely to be left   15 U.S.C. § 45(a)(1) (2000).   Gary Ruskin, FTC Caves in to Advertisers & Broadcasters on Product Placement, http:// archive.commercialalert.org/news/archive/2005/02/ftc-caves-in-to-advertisers-broadcasters-on-pro​ duct-placement. 255   Press Release, FTC, FTC Staff Reminds Influencers and Brands to Clearly Disclose Relationship (April 19, 2017), www.ftc.gov/news-events/press-releases/2017/04/ftc-staff-reminds-inf​luencers-brandscl​early-disclose. 256   Id. 257   European Broadcasting Union Legal Department, “EBU Preliminary Views on Regulatory Aspects of New Advertising Techniques” (Nov. 6, 2002), p.4, http://195.46.210.103/faq/documents/ leg_pp_new_advertising_techniques_061102.pdf (“To ensure that virtual advertising is used in a balanced way which is satisfactory for the viewer, in May 2000 the EBU adopted a Memorandum comprising a series of rules and principles . . . A number of basic principles may be recalled here:   – the presence of virtual advertising must be indicated to viewers by appropriate means, normally at the beginning and/or end of the programme;   – no virtual advertising may be inserted into the signal by the organizer of the event, his agents or any third party without the prior agreement of the broadcaster(s) holding the transmission rights;   – virtual advertising may be inserted only on surfaces at the venue which are customarily used for advertising; it may also be inserted on the field of play/surface, but only outside competition times;   – virtual advertising must not be used for products or services for which television advertising is prohibited in the country where the broadcaster is established.”) 258   EBU Memorandum on Virtual Advertising (May 25, 2000) at p.1 (available as an Appendix to the EBU Nov. 6, 2002 letter, supra). 253 254

BARFIELD PRINT.indd 464

13/11/2018 11:53

Advertising legal issues in virtual and augmented reality  465 unregulated. The proposed rules advocated by the EBU that are of particular relevance to advertising using virtual images are: (1) the use of virtual advertising must not change the quality of the program, or transform or impair the appearance of the venue where an event is taking place; (2) virtual advertising may be inserted only on surfaces at the venue which are customarily used for advertising; (3) virtual advertising may not appear on persons; (4) virtual advertising may not be inserted in a way that obscures, even in part, the television audience’s view of the action/performance; and (5) virtual advertising should be in keeping with the overall look of the venue and should not have greater prominence that the advertising seen at the venue.259 Applying these rules to advertising using virtual images, most augmented reality advertising will not be in line with the EBU suggested regulations. The EBU has also suggested that broadcasters inform their viewers in an appropriate manner, in accordance with national law and practice, of the use of virtual advertising.260 The disclosure of virtual advertising advocated by the EBU is compatible with the FTC’s rule on unfair and deceptive acts and practices. Certain other foreign regulators follow the EBU’s recommendations and either ban virtual signage altogether or issue strict regulations for the technology’s use.261 They were concerned that empowering broadcasters to digitally alter television signals could lead broadcasters to distort newscasts, change people’s faces, or engage in other deceptive practices.262 As more and more Americans become exposed to virtual and augmented reality advertising, and as legal issues develop concerning its use, we expect further guidance from the FTC and, likely, industry initiatives to try to forestall formal legislative action or regulations. 7.2  Municipal Regulations (Billboards) The location, content, and size of physical ads have been heavily regulated,263 and virtual or augmented reality projected ads in the physical world are likely to raise many of the same concerns. One particular type of advertisement that has been regulated is the outdoor billboard which consists of stationary structures, either standing independently or attached to a building.264 In response to the proliferation of billboards in real space, a   EBU Memorandum, Proposed Rules on pp.2–4.   Id. 261   Theresa McEvilly, Comment, Virtual Advertising in Sports Venues and the Federal Lanham Act 43(a): Revolutionary Technology Creates Controversial Advertising Medium, 8 Seton Hall J. Sports L. 603, 613–14 (1998) (“Portuguese television station and EBU member RTP used SciDel’s virtual advertising technology during the broadcast of a soccer game between Belgian and Portuguese teams in July 1997. The Portuguese regulators were displeased with RTP’s decision and subsequently warned the station not to use the technology again”) (citations omitted). 262   Id. 263   See, e.g., Chad Bogan, Adaptive, Understanding Sign Code Regulations for LED Sign Permits, Adaptive Display’s Digital Signage Blog (Mar. 19, 2015), www.adaptivedisplays.com/ blog/bid/361881/Understanding-Sign-Code-Regulations-for-LED-Sign-Permits; David Zahniser, “Appellate Court Hands L.A. a Legal Victory on Digital Billboards,” Los Angeles Times (March 10, 2016). For an example of one community’s response to regulating billboards see City Club of Portland Report: Billboard Regulation in Portland (Sept. 6, 1996), http://pdxscholar.library.pdx. edu/cgi/viewcontent.cgi?article=1478&context=oscdl_cityclub. 264   See generally, Florida Outdoor Advertising Association, http://foaa.org/rules-and-regulations; 259 260

BARFIELD PRINT.indd 465

13/11/2018 11:53

466  Research handbook on the law of virtual and augmented reality number of communities have enacted regulations imposing restrictions on the design and use of billboards.265 An interesting policy issue for communities to address is whether they should enact similar laws to regulate virtual ads projected in real space especially when used for augmented reality advertising. Augmented reality advertising will allow consumers to access information projected within the real world, or in advertising parlance, “outdoors.” Many local governments have sought to restrict the use of outdoor advertising by limiting sign spacing and location,266 prohibiting signs along roads designated as scenic corridors, limiting nighttime illumination of signs, and imposing moratoria on new sign construction.267 Lavey v. Two River is representative of a municipality’s power to regulate outdoor signs.268 In Lavey, an outdoor advertising company brought forward a Section 1983 action (that is, a civil action for deprivation of rights secured by the Constitution, such as due process) against the city, alleging that an ordinance regulating outdoor signs violated the First Amendment and federal and state due process clauses.269 The court held that the ordinance did not violate the First Amendment and that the ordinance was not so vague as to violate due process.270 Many brick and mortar regulations might not be directly applicable to current augmented reality advertising. For example, the size of the projected virtual image in real space might not be limited by the size restrictions imposed on real-world billboards.271 Current augmented reality systems coupled with a head-worn display lack the ability to project an image of the size approaching that of many local zoning ordinances. For example, in

Regulations for the Control of Outdoor Advertising in North Carolina, https://connect.ncdot.gov/ resources/Asset-Management/StateMaintOpsDocs/Outdoor%20Advertising%20Manual%20Part%​ 201.pdf. 265   Metromedia, Inc. v. City of San Diego, 453 U.S. 490 (9th Cir. 1981) (a divided court approved a ban on billboards contained in the city’s comprehensive sign ordinance, but held the sign ordinance unconstitutional because it also contained provisions found to violate the free speech clause); Prime Media, Inc. v. Brentwood, 398 F.3d 814 (6th Cir. 2005); Lombardo v. Warner, 353 F.3d 774 (9th Cir. 2003); Whiteco Outdoor Adver. v. City of Tucson, 972 P.2d 647 (Ariz. Ct. App. 1998) (discussing whether a charter city, in the exercise of its general regulatory police powers, may ban light fixtures mounted on the bottom of existing billboards). 266   Lavey v. Two Rivers, 171 F.3d 1110 (7th Cir. 1999) (outdoor advertising company and its president brought § 1983 action against city, alleging that ordinance regulating outdoor signs violated First Amendment and federal and state due process clauses; United States District Court held that: (1) ordinance did not violate First Amendment; (2) ordinance was not so vague as to violate due process; and (3) absence of intent element did not render ordinance unconstitutional). 267   Cafe Erotica of Florida, Inc. v. St. Johns County, 360 F.3d 1274 (11th Cir. 2004). 268   Lavey, 171 F.3d at 1110. 269   Section 1983 of Title 42 of the U.S. Code is part of the Civil Rights Act of 1871 and provides: “Every person who, under color of any statute, ordinance, regulation, custom, or usage, of any State or Territory or the District of Columbia, subjects, or causes to be subjected, any citizen of the United States or other person within the jurisdiction thereof to the deprivation of any rights, privileges, or immunities secured by the Constitution and laws, shall be liable to the party injured in an action at law, suit in equity, or other proper proceeding for redress.” 270   Id. 271   Some zoning ordinances restrict the size of signs to 1,200 square feet in manufacturing areas, and cap the size of nonadvertising signs within 200 feet of highways and parks at 500 square feet.

BARFIELD PRINT.indd 466

13/11/2018 11:53

Advertising legal issues in virtual and augmented reality  467 Lamar Tennessee, LLC v. City of Hendersonville,272 a case involving a dispute about the size of an outdoor billboard, the applicable city ordinance allowed a maximum permissible size for billboards of 80 square feet; this may be beyond the capability of current see-through displays (given the FOV provided by such displays) unless one is so far from the display that it subtends a small visual angle in the viewer’s field of view. If a future technology allowed such displays to be as large as desired, then it is likely they would be subject to restrictions under local zoning ordinances. Another form of outdoor advertising regulation involves limiting the placement of billboards along scenic corridors.273 Advertising using augmented reality displays could also be subject to similar restrictions. Although individuals not wearing an augmented reality display would not be able to see the virtual advertisement, the court would have to determine if public policy was satisfied by regulating the projection of virtual images within designated scenic or historic spaces. Legislative provisions regulating the size, height, location, manner of construction, and maintenance of advertising structures are considered to be valid where such provisions have a reasonable tendency to protect the public safety, health, morals, or general welfare.274 Would these public policy interests also allow municipalities to exercise their police power to regulate virtual ads projected within real space? Do aesthetic considerations alone constitute a sufficient basis for prohibiting off-premises advertising structures? While a number of courts have asserted that the promotion of scenic beauty promotes the public welfare and constitutes a sufficient basis for the exercise of the police power,275 other courts have held that although aesthetic considerations may be taken into account by the legislative body, such considerations alone do not constitute a sufficient justification for the exercise of the police power.276 In the US, the Federal Highway Beautification Act of 1965 is perhaps the best known prohibition of off-premises advertising structures, and is the basis for numerous state statutes prohibiting such structures near interstate highways. State statutes have usually been enacted in part to obtain the financial incentives offered by the Act to states that comply with its provisions.277 Thus, the outcome of traditional billboard advertising laws and regulations as applied to virtual reality and augmented reality is difficult to predict.

8. SUMMARY Sometimes the old ways are the best.278 Sometimes they are not. The increasing use of virtual and augmented reality advertising raises a host of issues under products liability 272   Lamar Tennessee, LLC. v. Hendersonville, Slip Copy, No. M2003-00415-COA-R3-CV (Tenn. Ct.App. Jan. 11, 2005). 273   Heideman v. South Salt Lake City, 348 F.3d 1182, 1192 (10th Cir. 2003); Wheeler v. Comm’r of Highways, 822 F.2d 586 (6th Cir. 1987) (ordinance allowed signs relating to any “activity” on premises), cert denied, 484 U.S. 1007 (1978). 274   See generally Murphy, Inc. v. Westport, 131 Conn. 292 (1944). 275   Whiteco Outdoor Adver., 972 P.2d at 674. 276   Murphy, 131 Conn. at 292. 277   Federal Highway Beautification Act of 1965, 23 U.S.C. § 131. 278   Kincade, gamekeeper in Skyfall movie (released 2012): “If all else fails, sometimes the old ways are the best.” Also, Eve character: “Sometimes the old ways are best.”

BARFIELD PRINT.indd 467

13/11/2018 11:53

468  Research handbook on the law of virtual and augmented reality laws, intellectual property laws, the common law of trespass and nuisance, and federal regulations regarding false advertising. To what extent should advertising using virtual images be considered a new technological advancement necessitating the need for new law, and to what extent is current law capable of handling disputes involving this new form of advertising? Advertising in virtual and augmented reality is still in its youth as a mass market technological capability. Advertising in each has the potential to be a “revolutionary” technology that can offer advertising agencies and sponsors new ways to generate revenue and target product ads to the most appropriate markets. Augmented reality technology may also provide consumers with the ability to occlude or edit ads displayed in the real world, thus giving consumers more freedom as to which information they see and access in the world.279 Future technological developments may also allow consumers to interact with ads in real time as they walk through a mall or down a city street, and to view virtual ads that are specifically targeted to particular individuals.280 However, even though there are many potential benefits of augmented reality advertising, the technology does pose ethical and legal considerations. There is concern regarding the potential for deception and that advertisements using virtual images may encroach into the space surrounding the user, creating a form of digital blight. The ultimate policy question is, of course: What kind of world do we want to live in? Will government regulations be necessary to protect society from the potentially harmful effects of advertising using virtual images? Conversely, if there are no augmented reality regulations, should the space in which a virtual image is projected be considered a “commons,” free for all to use?281 Existing laws and regulations involving torts, intellectual property, and false and deceptive advertising will have much to say about advertising using virtual images, but they may not offer the final word.

SOURCES Cases 44 Liquormart, Inc. v. Rhode Island, 517 U.S. 484 (1996) Alice Corp. v. CLS Bank International, 573 U.S. __, 134 S. Ct. 2347 (2014) Am. Amusement Mach. Ass’n, 244 F.3d at 577–8 America Online, Inc. v. IMS, 24 F.Supp.2d 548, 550 (E.D.Va.1998) AMF, Inc. v. Sleekcraft Boats, 599 F.2d 341 (9th Cir. 1979) Angerman v. Burick, 791 N.E.2d 984 (Ohio 2003) Bradley v. American Smelting and Refining Co., 709 P.2d 782 (Wash. Sup. Ct. 1985) Cafe Erotica of Florida, Inc. v. St. Johns County, 360 F.3d 1274 (11th Cir. 2004) Carson v. Here’s Johnny Portable Toilets, Inc., 698 F.2d 831 (6th Cir. Mich. Feb. 1, 1983) City of Modesto Redevelopment Agency v. Superior Court, 119 Cal. App. 4th 28 (Cal. Ct. App. 2004) CompuServe, Inc., v. Cyber Promotions, Inc., 962 F. Supp. 1015 (S.D. Ohio 1997) Costanza v. Seinfeld, 279 A.D.2d 255 (N.Y. App. Div. 1st Dep’t Jan. 4, 2001).

  See generally Mann, supra note 23.   Mike Butcher, The World of Minority Report Is About to Be Unlocked, TechCrunch (Apr. 19, 2017), https://techcrunch.com/2017/04/19/the-world-of-minority-report-is-about-be-unlockedand-th​e-key-is-inside-your-phone/. 281   See David Lange, Recognizing the Public Domain, 44 Law & Contemp. Probs. 147 (Autumn 1981). 279 280

BARFIELD PRINT.indd 468

13/11/2018 11:53

Advertising legal issues in virtual and augmented reality  469 Dr. Seuss Enters., LP v. Penguin Books USA, Inc., 109 F.3d 1394 (9th Cir. 1997) E.S.S. Entertainment v. Rockstar Games, 25 Berkeley Tech. L.J. 429 FTC v. Winstead Hosiery Co., 285 U.S. 483, 493 (1922) Google, Inc. v. Zuili, case number CBM2016-00008 Gravano v. Take-Two Interactive Software, Inc., 2016 NY Slip Op 05942 [142 AD3d 776], September 1, 2016, Appellate Division, First Department, available at www.courts.state.ny.us/reporter/3dseries/2016/2016_05942. htm Hildebrand v. Watts, 1997 Del. Ch. Lexis 32 (Del. Ch. 1997) Heideman v. South Salt Lake City, 348 F.3d 1182, 1192 (10th Cir. 2003) Hoffman v. Capital Cities/ABC Inc. et al., 255 F.3d 1180 (9th Cir. 2001) Horn v. City of Birmingham, 648 So.2d 607 (Ala. Civ. App. 1994) Hotmail Corp. v. Van$ Money Pie, Inc., No. C-98 JW PVT ENE, 1998 WL 388389 (N.D. Cal. 1998) In re McCormick & Company, Inc., Pepper Products Marketing & Sales Practices Litigation, 2016 WL 6078250, No. 15-cv-2188 (D.D.C. Oct. 17, 2016) In re NCAA Student-Athlete Name & Likeness Licensing Litig., 724 F.3d 1268 (9th Cir. 2013) Intel Corp. v. Hamidi, 71 P.3d 296 (Cal. 2003) Keller v. Electronics Arts, Inc., C 09-1967 CW, 2010 WL 530108, at 1 (N.D. Cal. Feb. 8, 2010) aff’d sub nom Lamar Tennessee, LLC. v. Hendersonville, Slip Copy, No M2003-00415-COA-R3-CV (Tenn.Ct.App. Jan. 11, 2005) Lavey v. Two Rivers, 171 F.3d 1110 (7th Cir. 1999) Lewis Galoob Toys, Inc. v. Nintendo of America, Inc., 964 F.2d 965, 967 (9th Cir. 1992) Location Based Services, LLC. v. Niantic, Inc., Civil Action No. 2:17-cv-411, E.D. Texas—Marshall Division, filed May 9, 2017 Lombardo v. Warner, 353 F.3d 774 (9th Cir. 2003) Matal v. Tam Slip Op. Metromedia, Inc. v. City of San Diego, 453 U.S. 490 (9th Cir. 1981) Midler v. Ford Motor Co. 1988, U.S. Court of Appeals, 9th Circuit, 849.F.2d 460 Mosler Safe Co. v Ely-Norris Safe Co., 273 US 132, 139 (1927) Murphy, Inc. v. Westport, 131 Conn. 292 (1944) O’Bannon v. NCAA (O’BANNON v. NATIONAL COLLEGIATE ATHLETIC ASSOCIATION: A CINDERELLA STORY, 22 Jeffrey S. Moorad Sports Law Journal 179 (2015)) Pestey v. Cushman, 788 A.2d 496, 502 (Conn. 2002) Prime Media, Inc. v. Brentwood, 398 F.3d 814 (6th Cir. 2005) RecogniCorp, LLC. v. Nintendo Co., Ltd. and Nintendo of America Inc., Civil Action 2016-1499, U.S. Court of Appeals for the Federal Circuit, opinion issued April 28, 2017 Register.com, Inc. v. Verio, Inc., 126 F. Supp. 2d 238, 250 (S.D.N.Y. 2000) Rosetta Stone v. Google, Inc. 676 F.3d 144 (4th Cir. 2012) Sherwood 48 Assocs. 76 Fed. Appx. at 390 Stratford Homes, Inc., v. Lorusso, No 94-CV-517E(M), 1995 WL 780977 (W.D.N.Y. 1995) The Shelburne v. Crossan Corp., 95 N.J. Eq. 188 (N.J. Ch. 1923) Thrifty-Tel, Inc. v. Bezenek, 54 Cal. Rptr. 2d 468, 473 (Cal. Ct. App. 1996) U-Haul Intern. v. Jartran, Inc., 601 F. Supp. 1140, 1148-1149 (D. Ariz. 1984), affd, 793 F.2d 1034 (9th Cir. 1986) Video Software Dealers Ass’n v. Maleng, 325 F. Supp. 2d 1180 (W.D. Wash. 2004) Wheeler v. Comm’r of Highways, 822 F.2d 586 (6th Cir. 1987) White v. Samsung Electronics America: The Ninth Circuit Turns a New Letter in California Right of Publicity Law, 15 Pace L. Rev. 161 (Fall 1994). 971 F.2d at 1398–9 Whiteco Outdoor Adver. v. City of Tucson, 972 P.2d 647 (Ariz. Ct. App. 1998) Wilson v. Interlake Steel Co., 32 Cal. 3d 229, 233–4 (1982) Wilson v. Midway Games, Inc., 198 F. Supp. 2d 167, 181 (D. Conn. 2002)

Statutes Computer Fraud and Abuse Act, 18 U.S.C. §1030 (2001) Economic Espionage Act, 18 U.S.C. §1831 (1996) Federal Highway Beautification Act of 1965, 23 U.S.C. §131 Lanham Act, 15 U.S.C. §1114(1) 15 U.S.C. §45(a)(1) (2000) 15 U.S.C. §1052(a) 15 U.S.C. §§1101 15 U.S.C. §1125

BARFIELD PRINT.indd 469

13/11/2018 11:53

470  Research handbook on the law of virtual and augmented reality 15 U.S.C. §1127 17 U.S.C. §106(1) and (2) 17 U.S.C. §120(a)

Regulations The California Consumers Legal Remedies Act (California Civil Code) §§ 1750 et seq California Code, Civil Code—CIV § 3344 Kristen M. Ploetz, Light Pollution in the United States: An Overview of the Inadequacies of the Common Law and State and Local Regulation, 36 NEW ENGLAND LAW REVIEW 987 (2002) 37 C.F.R. Sec. 202.11 § 1780 Section 1983 of Title 42 of the U.S. Code, part of the Civil Rights Act of 1871

BARFIELD PRINT.indd 470

13/11/2018 11:53

15.  Data privacy legal issues in virtual and augmented reality advertising S.J. Blodgett-Ford and Mirjam Supponen*

. . . it was, admittedly, a little difficult to make my way through the streets, and when I saw a miraculously clear stretch of sidewalk I took it. I wondered at the time why the elderly man I pushed aside getting to the sidewalk gave me such a strange look. “Watch it, buster!” he called. He was waving a signpost, but of course it was graffiti-covered. I wasn’t in a mood to worry about some minor civil ordinance. I walked past— And WOWP a blast of sound shook my skull and FLOOP a great supernova flare of light burned my eyes, and I went staggering and reeling as tiny, tiny elf voice shouted like needles in my ears Mokie-Koke, Mokie-Koke, MokieMokieMokie-Koke! And went on doing it, with variations, for what seemed like a hundred years or more. Stenches smote my nose. Subsonic shivers shook my body . . . while my ears were still ringing and my eyes still stinging with that awful blast of sound and light, I picked myself up from where I lay sprawled on the ground. “I warned ya,” yelled the little old man from a safe distance.1

In the 1984 science fiction/satire The Merchant’s War by Frederik Pohl, advertising agencies rule the world and there are virtually no legal or ethical limits on advertisers’ ability to influence consumers. In this dystopian scenario, advertisers use powerful virtual and augmented reality technology that blasts the visual, auditory, olfactory, and even temporal senses of unsuspecting pedestrians—literally knocking them out and brainwashing them. The book’s antihero, Tennison Tarb, absentmindedly walks into such an advertising zone and, as a result, becomes addicted to “Mokie-Koke.” While we may not (yet) be at that point, virtual reality and augmented reality technologies clearly offer new ways to target experiences, messages, and promotional offers to individual consumers, and to collect and use data regarding consumers’ interactions with such content. Advertisers can leverage existing massive data sets2—at petabyte (1015 bytes) scales—that include an almost unimaginable level of fine-grained detail about consumer behavior and preferences over an extended period of time. Online data collection of consumer behavior *  The authors are extremely grateful to Woodrow Barfield and Marc Blitz for their valuable comments and feedback, and to Mariah Figlietti Vasquez for her research assistance. 1   Frederik Pohl, The Merchant’s War 47 (St. Martin’s Press 1984). 2   Kalev Leetaru (Contributor Opinion), Why Are We So Afraid of Petabytes? Forbes (Jan. 17, 2017), www.forbes.com/sites/kalevleetaru/2017/01/17/why-are-we-so-afraid-of-petabytes/#5f14bc365875) (“How have we reached a world in which five years ago Google engineers were casually running sorting benchmarks on 50PB, three years ago Facebook was generating 4 petabytes of new data per day and today real world companies are storing hundred petabyte archives in Google’s BigQuery platform, yet outside Silicon Valley I hear data scientists talk about analyzing terabytes as pushing the boundaries of the possible?”); see also Tino Tereshko (Google employee), 15 awesome things you probably didn’t know about Google BigQuery (Oct. 21, 2016), https://medium.com/google-cloud/15-awesome-things-youprobably-didnt-know-about-google-bigquery-6654841fa2dc (“Some of BigQuery’s customers include Spotify, New York Times, Coca Cola, Viant, Motorola, Kabam Games, Vine, Disney Interactive, Lloyds Bank, Citigroup, Sharethis, Wepay, Zulily, and many many others”).

471

BARFIELD PRINT.indd 471

13/11/2018 11:53

472  Research handbook on the law of virtual and augmented reality has been occurring for more than two decades.3 Social media expanded the volume and variety of personal data content collected (including data about relationships and connections to friends and family).4 Facebook has already entered the virtual reality market via Facebook 360, which “allows publishers and content creators to share immersive stories, places and experiences with their fans” using the Oculus Rift or Samsung Gear VR devices.5 Such “content creators” will certainly include advertisers, given the massive ad revenue tech giants receive. For example, Facebook and Google were predicted to receive more than $100 billion from advertising revenue in 2017—almost half of the global spending on digital advertising.6 Perhaps most significantly, smartphones and other mobile devices have dramatically expanded the granularity of data collected,7 including geolocation data,8 and virtual and augmented reality advertising is highly likely to continue, and expand, such practices. Many smartphones and applications track consumers and share data with third parties, typically for advertising purposes, in addition to sharing it with the device manufacturer and the app developer.9 Moreover, mobile applications may track the consumer even when the consumer is not actively using the application, as long as the application is still running in the background.10 Some applications (and devices) may track users even when location sharing is supposedly disabled.11 Virtual and augmented reality technologies similarly can enable such tracking, including interactions between consumers and advertisements, down to the level of where consumer’s eyes move—how long their eyes stay at a particular area—what their facial expression indicates about their reaction, and

  Amazon was founded in 1994.   Facebook launched in 2004.  5  https://facebook360.fb.com/.  6   Lucy Handley, Facebook and Google Predicted to Make $106 Billion from Advertising in 2017, Almost Half of World’s Digital Ad Spend, CNBC (Mar. 21, 2017), available at www.cnbc. com/2017/03/21/facebook-and-google-ad-youtube-make-advertising-in-2017.html.  7   The Apple iPhone launched in 2007.  8   See, for example, Section, Getting the User’s Location, Apple’s Developer Guide, https://devel​ oper.apple.com/library/content/documentation/UserExperience/Conceptual/LocationAwarenessPG/ CoreLocation/CoreLocation.html.  9   Narseo Vallina-Rodriguez and Srikanth Sundaresan, 7 in 10 Smartphone Apps Share Your Data with Third-Party Services, Scientific American (May 30, 2017), www.scientificamerican. com/article/7-in-10-smartphone-apps-share-your-data-with-third-party-services/. 10   Andrew J. Hawkins, Uber Wants to Track Your Location Even When You Are Not Using the App: Some Riders Feel “Icky” About the New Permission, The Verge (Nov. 30, 2016), www.­sci​ entificamerican.com/article/7-in-10-smartphone-apps-share-your-data-with-third-party-services/ and see Uber’s Privacy Policy at www.uber.com/legal/privacy/users/en/ (“Location Information: When you use the Services for transportation or delivery, we collect precise location data about the trip from the Uber app used by the Driver. If you permit the Uber app to access location services through the permission system used by your mobile operating system (‘platform’), we may also collect the precise location of your device when the app is running in the foreground or background. We may also derive your approximate location from your IP address”). 11   Zack Whittaker, AccuWeather Caught Sending User Location Data, Even When Location Sharing Is Off, ZDNet (Aug. 22, 2017), www.zdnet.com/article/accuweather-caught-sendinggeo-location-data-even-when-denied-access/ (“A security researcher has found that the popular weather app sends private location data without the user’s explicit permission to a firm designed to monetize user locations”).  3  4

BARFIELD PRINT.indd 472

13/11/2018 11:53

Data privacy legal issues in virtual and augmented reality advertising  473 even brainwave patterns. For example, the data aggregator Google has the capability to continually track its users through its “free” services, especially since the introduction of its Google Glass. “With this additional data, Google Glass could take targeted advertising to an entirely new level. Such advertising would likely be unnervingly accurate and would interact with almost everything the user interacts with.”12 As virtual and augmented reality technology and data collection and use practices continue to become more powerful, a key question will be whether there are any legal limits to prevent the world Pohl described from becoming a reality. This chapter considers certain legal constraints on virtual and augmented reality advertising. After providing an overview of US and European Union data privacy regulations, selected issues that are likely to be especially relevant to virtual reality and augmented reality advertising will be highlighted, such as: (1) the foundational principles of notice and choice; (2) children’s advertising; (3) the collection and use of geolocation data; and (4) the collection and use of biometric data.13

1.  LEGAL FRAMEWORK—US AND EU Privacy and data protection laws, regulations, and industry standards already play a significant role in online, offline, and mobile device advertising, and they will certainly be important to virtual reality and augmented reality advertisements, particularly any such advertisements that collect and use geolocation data or biometric data. Data protection laws and regulations provide individuals with rights regarding their personal information and limit the ways in which organizations, including advertisers, may collect and use this information.14 A global and connected world requires compliance with more than one set of national rules, and the vast international differences can make compliance burdensome. As data has increasingly become one of the most valuable assets in the marketing sector, it is imperative that organizations understand how to navigate the data privacy field and avoid expensive and reputation-damaging pitfalls. That basic principle will continue to be true for virtual reality and augmented reality advertising, and certain aspects of that technology will bring acute compliance challenges for advertisers. In this chapter, we will provide a general overview of the existing legal framework in the United States and the European Union (“EU”), including the EU General Data Protection Regulation (“GDPR”) that took effect in May of 2018, then focus on selected issues that may be especially relevant to virtual reality and augmented reality advertising, such as geolocation, geofencing and beacons, and biometrics and facial recognition.

12   Michael S. Wagner, Note, Google Glass: A Preemptive Look at Privacy Concerns, 11 J. on Telecomm. & High Tech. L. 475, 489 (2013). 13   In this chapter, the authors have elected to take a practical approach to the laws, regulations, and industry standards as they exist versus a normative or prescriptive approach, such as a conceptual analysis of “virtual persons” as outlined by Jonathan W. Penney in Privacy and the New Virtualism, 10 Yale J. L. & Tech. 194 (2007). 14   Peter Carey, Data Protection: A Practical Guide to UK and EU Law 1 (Oxford University Press 3rd ed. 2009).

BARFIELD PRINT.indd 473

13/11/2018 11:53

474  Research handbook on the law of virtual and augmented reality To say that data privacy and security protections across and within countries and states form a “patchwork” is an understatement. For virtual reality and augmented ­reality advertisers in a global online market, such laws, regulations, and industry standards may more accurately be characterized as an n-set Venn diagram with a relatively large “n” value.15 It is almost impossible to grasp how such laws, regulations, and industry standards overlap and diverge as a whole at an abstract level. There are general laws in many countries,16 and in some regions,17 that relate to the collection, use, storage, and sharing of information or data that can be used to identify a specific person, which is commonly referred to as “Personally Identifiable Information” (or “PII”) in the United States and “personal data” in Europe. There are category-based laws regarding certain types of data that is deemed especially sensitive, such as health-related data,18 financial data,19 credit data,20 educational records,21 children’s data,22 and even video rental data.23 There are also laws related to the collection and use of data by the government, which may require that private businesses share data they collect with law enforcement agencies,24 and government guidelines regarding cybersecurity.25 There are state level consumer protection laws,26 which may require written information security programs,27 notices of data breaches, and certain security measures. There are laws specific to advertising, such as anti-spam laws,28 and laws that prohibit an advertiser 15   For an example of a six-set Venn diagram, see https://eagereyes.org/techniques/­venn-diagrams. Unless otherwise indicated, all websites cited in this chapter were last visited between June and August 2017. 16   For example, Canada Personal Information Protection and Electronic Documents Act (PIPEDA), S.C. 2000, c. 5 (2015); United Kingdom Data Protection Act and the Australian Privacy Principals. 17   Data Protection Directive in the European Union. 18   Health Insurance Portability and Accountability Act of 1996 (HIPAA), 45 C.F.R. Sections 160–64. 19   Gramm–Leach–Bliley Act (GLBA), 15 U.S.C. Sections 6801–9. 20   Fair and Accurate Credit Transactions Act (FACTA), 15 U.S.C. § 1601 et. seq., 1681 et. seq. 21   Family Educational Rights and Privacy Act (FERPA), 20 U.S.C. § 1232(g). 22   Children’s Online Privacy Protection Act (COPPA), 15 U.S.C. §§ 6501–6. 23   Video Privacy Protection Act (VPPA). The VPPA was amended to allow rental services such as Netflix to share customers’ viewing data on social media sites. For a general discussion, see Joe Mullin, Congress Tweaks US Video-Privacy Law So Netflix Can Get on Facebook, Ars Technica (Dec. 21, 2012), https://arstechnica.com/tech-policy/2012/12/congress-tweaks-us-video-privacy-lawso-netflix-can-get-on-facebook/. Another question of potential interest is whether movie-watching in augmented or virtual reality could be considered comparable to watching movies in the physical world. The Washington Post (Jan 19, 2016), www.washingtonpost.com/news/answer-sheet/ wp/2016/07/19/pokemon-go-sparks-concern-about-childrens-privacy/?utm_term=.b07577f2ba8f. 24   Uniting and Strengthening America by Providing Appropriate Tools Required to Intercept and Obstruct Terrorism (USA PATRIOT) Act of 2001, Pub. L. No. 107-56. 25   National Institute of Standards and Technology (NIST) Risk Management Framework. 26   California Consumer Legal Remedies Act (CLRA), California Civil Code Sec. 1750 et seq., Unfair Competition Law, California Business and Professions Code Sec. 17200 et seq. 27  Massachusetts M.G.L. chapter 93H, 201 CMR 17.00: STANDARDS FOR THE PROTECTION OF PERSONAL INFORMATION OF RESIDENTS OF THE COMMONWEALTH. 28   Controlling the Assault of Non-Solicited Pornography and Marketing (CAN-SPAM) Act of 2003, Pub. L. No. 180–87.

BARFIELD PRINT.indd 474

13/11/2018 11:53

Data privacy legal issues in virtual and augmented reality advertising  475 from electronically texting or calling a consumer’s mobile phone without appropriate consent.29 There are international guidelines,30 as well as frameworks,31 to consider. Furthermore, there are industry standards,32 certifications,33 and requirements that,34 as a practical matter, are enforced by the major players in the advertising and marketing industry. Perhaps most importantly when it comes to mobile devices and online platforms, the device and operating system manufacturers (Apple, Google, Samsung, Nokia, and so on), wireless network providers (Verizon, T-Mobile, AT&T, China Mobile, and so on), and the dominant online platforms (Facebook, Google, Amazon, Alibaba, and so on) have tremendous control over the options available for consumers and for advertisers, and even if the names of some of the players may change due to acquisitions, divestitures, and consolidation, the impact these behemoths exert is likely to continue to be massive. Device manufacturers, network providers, and online platforms already play a key role in practice in shaping consumer privacy rights and protections—and they are likely to be the leaders in virtual and augmented reality advertising as they extend their technological reach in that direction.35 With that important caveat in mind, we turn to the legal framework in the US, and then that in the EU, before addressing some specific areas relevant to virtual and augmented reality advertising. 1.1  US Basic Legal Framework In the United States, although there is a constitutional right to privacy under the U.S. Constitution and many state constitutions,36 and data privacy protections exist under state tort law and criminal laws37—as well as a plethora of federal and state statutes and regulations—when it comes to day to day commercial advertising and marketing, the most relevant federal legal constraints often lie in the following sources:   Telephone Consumer Protection Act (TCPA) of 1991, 47 U.S.C. Sec. 227.   OECD Guidelines on the Protection of Privacy and Transborder Flows of Personal Data. 31   European Union—United States Privacy Shield (Privacy Shield). 32   Digital Advertising Alliance (DAA) code, European Digital Advertising Alliance (eDAA) code, Network Advertising Initiative (NAI) code, and the Direct Marketing Association (DMA) code. 33   TRUSTe Privacy Certification Standards. 34   Payment Card Industry (PCI) Data Security Standards. 35   Lee A. Bygrave, Data Privacy Law: An International Perspective 107 (Oxford University Press 2014). 36   Privacy Protections in State Constitutions, National Conference of State Legislatures (May 5, 2015), www.ncsl.org/research/telecommunications-and-information-technology/privacyprotections-in-state-constitutions.aspx. Note that the U.S. Constitution does not provide protection from private invasions of privacy. Rather, it protects people from surveillance by the government. See Griswold v. Connecticut, 381 U.S. 479 (1965). One question of potential interest is whether the acts of an avatar in his or her house within virtual reality would be considered protected by the U.S. Constitution. For a brief discussion regarding the need for a warrant to search a house in Second Life, see Marc J. Blitz, Stanley in Cyberspace: Why the Privacy Protection  of  the  First Amendment Should Be More Like that of the Fourth, 62 Hastings L. J. 357–400 (2010). 37   So-called revenge porn statutes are an example. See, e.g., Danielle Citron, Hate Crimes in Cyberspace (Harvard University Press 2014) and the Cyber Civil Rights Initiative, www. cybercivilrights.org/revenge-porn-laws/. 29 30

BARFIELD PRINT.indd 475

13/11/2018 11:53

476  Research handbook on the law of virtual and augmented reality ●

Section 5 of the Federal Trade Commission Act (“FTC Act”), which prohibits “unfair and deceptive acts or practices in or affecting commerce.”38 ● Children’s Online Privacy Protection Act (“COPPA”),39 which restricts the collection, use, or disclosure of personal information (including IP addresses) of children under the age of 13. ● Controlling the Assault of Non-Solicited Pornography and Marketing Act (“CANSPAM”),40 which sets rules for commercial email messages, including “opt-out” rules. ● Telephone Consumer Protection Act (“TCPA”),41 which restricts telephone solicitations (including texts to smartphones) that use automatic dialing systems, which have been broadly defined to include any technology with the capacity to dial random or sequential numbers.42 The FTC has enforcement authority for each of the Acts set forth above except for the TCPA, which is enforced by the Federal Communications Commission (“FCC”), and CAN-SPAM, where the FCC has responsibility for wireless devices and the FTC handles email messages. In enforcing Section 5 of the FTC Act in an advertising and marketing context, the FTC often compares the actual practices of an advertiser (or its authorized agents) with what the advertiser (or the agents) promise to consumers in the applicable privacy policy or “privacy notice.”43 The FCC also regulates common carriers such as Internet Service Providers (“ISPs”), and has in the past taken steps to increase its scope in the data privacy area. Many of the privacy-related federal laws and regulations do not preempt states from having similar or stricter laws.44 When the FCC’s proposed strong privacy regulations for ISPs were withdrawn after the 2016 US federal elections,45 many states tried to fill the

  15 U.S.C. Sec. 45.   15 U.S.C. Sec. 6501 et seq. 40   15 U.S.C. Sec. 7701 et seq. 41   47 U.S.C. Sec. 227. 42   There is litigation regarding the exact scope of this definition, notwithstanding the FCC’s position: Dial “C” for Confusion: Courts Split on TCPA Definition of Autodialer, Lexology (Mar. 15, 2016), www.lexology.com/library/detail.aspx?g=a2b689ea-6ce1-4d1c-b77b-036cd38b9ab9. 43   See “Enforcing Privacy Promises” for a list of relevant press releases from the FTC regarding its actions: www.ftc.gov/news-events/media-resources/protecting-consumer-privacy/enforcing-pri​ vacy-promises. 44   For a useful and concise overview of federal and state privacy laws, see Daniel J. Solove and Paul M. Schwartz, Privacy Law Fundamentals (IAPP 2017). Preemption is covered on pp.187–93. For an excellent summary of state internet privacy laws, see www.ncsl.org/research/ telecommunications-and-information-technology/state-laws-related-to-internet-privacy.aspx (last updated June 20, 2017). 45   Cecilia Kang, Congress moves to overturn Obama-era online privacy rules, New York Times, Mar. 28, 2017 (“Congress completed its overturning of the nation’s strongest internet privacy protections for individuals on Tuesday in a victory for telecommunications companies, which can track and sell a customer’s online information with greater ease. In a 215-to-205 vote largely along party lines, House Republicans moved to dismantle rules created by the Federal Communications Commission in October. Those rules, which had been slated to go into effect later this year, had required broadband providers to receive permission before collecting data on a user’s online 38 39

BARFIELD PRINT.indd 476

13/11/2018 11:53

Data privacy legal issues in virtual and augmented reality advertising  477 gap. Days after US President Donald Trump signed legislation allowing ISPs to sell the personal data of subscribers, numerous states introduced specific legislation to protect their residents from such collection, including Connecticut, Illinois, Kansas, Maryland, Massachusetts, Minnesota, Montana, New York, Washington, and Wisconsin.46 The New York legislation, for example, would prohibit ISPs from selling customer browsing history and other personal information to third parties.47 California has strong consumer protection laws related to privacy in an online world, which are likely already directly applicable to, or will be extended to cover, virtual and augmented reality advertising. Under the California Online Privacy Protection Act (“CalOPPA”), California also requires “commercial Web site” operators who collect personally identifiable information about individual consumers in California to “conspicuously post” a privacy policy on their website.48 The privacy policy must disclose how the website responds to a web browser “Do Not Track” signal or similar mechanisms, and whether third parties are or may be conducting such tracking on the operator’s site or service.49 Virtual and augmented reality advertising will often not be delivered via a “commercial Web site.” Instead, mobile applications, which typically are not browserbased and therefore do not recognize “Do Not Track” signals, and wearable devices such as smartphones and Google Glass-type products, will undoubtedly be key delivery mechanisms for virtual and augmented reality ads. This opens up a potential loophole for virtual and augmented reality advertisers, because the term “operator” does not include “any third party that operates, hosts, or manages, but does not own, a Web site or online service on the owner’s behalf or by processing information on behalf of the owner.”50 This means that, in delivering targeted virtual reality and augmented reality advertisements for third parties (not themselves) on internet-connected mobile or wearable devices, key industry players such as Verizon might not be deemed to be “operators” covered by this particular California law. Similarly, the California “Shine the Light”51 law requires that companies and data brokers52—who have also become increasingly powerful for targeted advertising53— ­activities”), www.nytimes.com/2017/03/28/technology/congress-votes-to-overturn-obama-era-onli​ ne-pr​ivacy-rules.html. 46   Eyragon Eidam and Jessica Mulholland, 10 States Take Internet Privacy Matters into Their Own Hands, Government Technology Magazine, Apr. 10, 2017, www.govtech.com/policy/10States-Take-Internet-Privacy-Matters-Into-Their-Own-Hands.html. 47   Press Release from Senator Timothy M. Kennedy, Not in New York: Kennedy introduces bill to ban internet providers from selling your personal information, Apr. 4, 2017 (noting “health insurance companies could potentially purchase the search history of customers and raise their rates, or even cancel their coverage due to searches about a preexisting condition”), www.nysenate.gov/ newsroom/press-releases/timothy-m-kennedy/not-new-york-kennedy-introduces-bill-ban-internet. 48   Calif. Bus. & Prof. Code § 22575(a). 49   Calif. Bus. & Prof. Code § 22575. 50   Calif. Bus. & Prof. Code § 22577(c). 51   Calif. Civil Code § 1798.83. 52   “The sale, rental, exchange, or lease of personal information for consideration to businesses is a direct marketing purpose of the business that sells, rents, exchanges, or obtains consideration for the personal information.” Calif. Civil Code § 1798.83(e)(2) (excerpted). 53   Brian Naylor, Firms are buying, sharing your online info. What can you do about it? NPR All Things Considered radio broadcast, July 11, 2016 (“One of the largest brokers, Acxiom, reported over

BARFIELD PRINT.indd 477

13/11/2018 11:53

478  Research handbook on the law of virtual and augmented reality that do business with California residents must allow customers to opt out of information sharing or disclose in detail how personal information is shared for direct marketing purposes. In the data broker role, ISPs such as Verizon would be covered by the law to the extent that they sell personal information for consideration to other businesses for “direct marketing” purposes. On its face, however, the law is specific to only physical mail, electronic mail, or telephone marketing because “[d]irect marketing purposes” is defined as “the use of personal information to solicit or induce a purchase, rental, lease, or exchange of products, goods, property, or services directly to individuals by means of the mail, telephone, or electronic mail for their personal, family, or household purposes” (emphasis added).54 Thus, the “Shine the Light” law could allow targeted virtual reality and augmented reality advertising experiences, such as on Facebook 360, to stay in the darkness unless the law is extended by legislation, or regulatory agency or enforcement authority interpretation. Facebook’s use of terms such as “publishers and content creators” and “immersive stories, places and experiences”55 in describing the Facebook 360 technology is no accident. It is strongly in the interests of Facebook to avoid using the more loaded phrase “advertising” when communicating with the general public and potential regulators. Facebook is less circumspect when communicating with potential advertisers directly and it is clear that advertisers are interested in using Facebook 360 video in ads.56 In response to a question in the Facebook Business “Advertiser Help Center,” Facebook responded: “The option to use 360 videos in ads is still being tested, however, you are able to boost posts with 360 video. [An] article from our Help Center goes over the steps on how to do this.”57 However, the California Consumer Privacy Act of 2018, Cal. Civ. Code Sec. 1798.100 to Sec. 1798.198 (“CCPA”), which takes effect from January 1, 2020, is a more comprehensive law and may close some of the potential gaps when it comes to virtual and augmented reality advertising. In addition to giving California consumers broad data access and data deletion rights, the CCPA requires that covered organizations provide a clear and conspicuous “Do Not Sell My Personal Information” link on their home pages so California consumers can easily opt out of the sale of their personal data. US-based virtual and augmented reality advertisers and their agents must carefully navigate US federal and state data privacy laws, including potential loopholes as technology moves faster than the law. However, a narrow geographic view does not make sense when most of the brands that would be likely to be early adopters of virtual and $800 million in revenue last year”), www.npr.org/sections/alltechconsidered/2016/07/11/485571291/ firms-are-buying-sharing-your-online-info-what-can-you-do-about-it. See also FTC Report, Big data: a tool for inclusion or exclusion? (Jan. 2016), www.ftc.gov/system/files/documents/reports/ big-data-tool-inclusion-or-exclusion-understanding-issues/160106big-data-rpt.pdf and the earlier FTC report Data brokers: a call for transparency and accountability (2014), www.ftc.gov/system/ files/documents/reports/data-brokers-call-transparency-accountability-report-federal-trade-­comm​ ission-may-2014/140527databrokerreport.pdf. 54   Calif. Civil Code § 1798.83(e)(2) (excerpted). 55  https://facebook360.fb.com/ 56   Grace Caffyn, Advertisers Take a Liking to 360-Video Ads, DigiDay Dec. 8, 2016, available at https://digiday.com/uk/360-video-ads-turning-advertiser-heads/. 57   Available at www.facebook.com/business/help/community/question/?id=10153902626538338, linking to www.facebook.com/business/help/1730249197231921/?ref=u2u.

BARFIELD PRINT.indd 478

13/11/2018 11:53

Data privacy legal issues in virtual and augmented reality advertising  479 augmented reality technology for advertising are global, and such advertising will often be targeted using global ISPs and technology platforms such as Facebook and Google. Thus, we also consider the situation in Europe,58 because European laws are generally viewed as being much more strict than United States laws, and they will become even more severe now that GDPR has taken effect. 1.2  European Union Basic Legal Framework The European Union generally is viewed as taking a more stringent approach to data privacy than the US. In addition to personal data protection being viewed as a fundamental right in the EU, the legislation is more comprehensive and, from a corporate perspective, compliance can be more burdensome. For example, EU data protection legislation covers both public and private sectors, applies to both automated and manual processing of personal data, and regulates international data transfers. It is also interpreted to have a more comprehensive reach due to the broad definitions of “processing” and “personal data.”59 In fact, what qualifies as “personal data” under EU data privacy law often is surprising to US-based advertisers and businesses. The question of what is considered “personal data” is important, because it helps determine the extent to which European data protection law affects the business operations of advertisers. “Personal data” refers to “any information relating to an identified or identifiable natural person.”60 The term “personal” can be misleading because “personal data” refers to all data about a person and not just data about their personal life.61 Personal data can be anything from a name, email address, IP address, mobile device ID, or social media post to an avatar identifier. Images of an individual can also be personal data if the individual in the images is recognizable.62 Furthermore, an identifiable person is someone who can be identified directly or indirectly.63 For example, a person’s name could directly identify a person whereas a description of a person’s profession, age, and location could identify someone indirectly. Advertisers no longer need an individual’s name or email address to categorize that individual based on their socioeconomic, psychological, or behavioral 58   Many other countries also have relevant laws either in place or in process. For a helpful matrix of data privacy laws around the world, see Global Privacy Handbook, Baker & McKenzie (2017), http://globalprivacymatrix.bakermckenzie.com/. Also, in Justice KS Puttaswamy (Retd.) & Anr. v. Union of India & Ors (an extremely lengthy written opinion) the India Supreme Court reversed decade-old precedent and held that the right to privacy is a fundamental right because it is “an intrinsic part of the right to life.” The case involved a challenge to governmental regulations requiring that Indian citizens submit biometric data in order to access certain governmental services. Murali Krishnan, Breaking: Right to Privacy is a Fundamental Right Under Article 21, Supreme Court, Bar & Bench, Indian Legal News, Aug. 24, 2017, https://barandbench.com/ right-privacy-fundamental-right-supreme-court/. 59   Lee A. Bygrave, supra note 35, at 100. 60   Article 2(a), DPD. 61   Christopher Kuner, European Data Protection Law: Corporate Compliance and Regulation 91–2 (Oxford University Press 2007). 62   Article 29 Data Protection Working Party, Opinion 4/2007 on the Concept of Personal Data, WP 136 at 8, http://ec.europa.eu/justice/data-protection/article-29/documentation/opinion-recommen​dat​ ion/files/2007/wp136_en.pdf. 63   Article 2(a), DPD.

BARFIELD PRINT.indd 479

13/11/2018 11:53

480  Research handbook on the law of virtual and augmented reality patterns and make effective advertising decisions about them.64 Furthermore, advanced tracking tools make it easy to track the patterns and activities of mobile devices and link them to their users. Today, the ability to identify an individual is increasingly dependent on who is matching the data, what their technical capabilities are, and what data sets are being aggregated.65 With this background in mind, the European Court of Justice (“CJEU”) has held that even dynamic IP addresses fall under this broad definition of personal data in certain circumstances.66 In practice, this means that advertisers in the augmented and virtual reality contexts will find it nearly impossible to claim that they are not processing personal data. Under the umbrella of “personal data,” EU data privacy law also identifies certain special categories that require extra protection due to their relation to fundamental rights and freedoms.67 These types of data are generally referred to as “sensitive data” and include information about racial or ethnic origin, political opinions, religious or philosophical beliefs, trade union membership, and data concerning health or one’s sex life.68 However, the definition of sensitive data now also includes biometric and genetic data.69 The general rule is that the processing of such data is prohibited, except under certain limited circumstances.70 From an advertiser’s perspective, sensitive data generally requires an individual’s express consent prior to any data collection or processing activities.71 From 1995 until 2018, the central instrument of EU data privacy law was the Data Protection Directive (95/46/EC) (“DPD”). Drafted decades before the era of advanced advertising technology, it was challenging to apply to modern activities. Enormous technological advances in the marketing sector have motivated local data protection authorities to actively monitor any potential privacy threats.72 In response, the EU data privacy field underwent major reform in an attempt to catch up with these new technological developments, and the challenges in finalizing the e-Privacy Regulation show that the process is ongoing. In May 2018, the General Data Protection Regulation (“GDPR”) replaced the DPD. The GDPR was designed to simplify compliance with EU data privacy laws because organizations generally (subject to variations or “derogations”) need only comply with one regulation instead of 28 variations of the old directive.73 However, due to its 64   Opinion 4/2007 on the Concept of Personal Data, WP 136 at 14, http://ec.europa.eu/justice/ data-protection/article-29/documentation/opinion-recommendation/files/2007/wp136_en.pdf. 65   Recital 26, DPD; Kuner, supra note 61, at 91. 66   Patrick Breyer v. Bundesrepublik Deutschland, No C-582/14, at ¶49 (Second Chamber Oct. 19, 2016), http://curia.europa.eu/juris/document/document.jsf ?docid=184668&doclang=EN. 67   Article 29 Data Protection Working Party, Advice Paper on Special Categories of Data (“Sensitive Data”), at 4. http://ec.europa.eu/justice/data-protection/article-29/documentation/ other-document/files/2011/2011_04_20_letter_artwp_mme_le_bail_directive_9546ec_annex1_en.​ pdf. 68   Article 8(1), DPD. An interesting question is whether these types of data gathered about an avatar would also be considered “sensitive data” under EU data protection law. 69   Article 9(1), GDPR. 70   Article 8(1), DPD; Article 9(1), GDPR. 71   Article 8(2)(a), DPD; Article 9(2)(a), GDPR. 72  Carey, supra note 14, at 307. 73   Under EU law, “regulations” are directly applicable across the EU. “Directives” are not directly applicable, but set out an objective or policy that Member States must implement

BARFIELD PRINT.indd 480

13/11/2018 11:53

Data privacy legal issues in virtual and augmented reality advertising  481 extraterritorial nature, organizations are not necessarily able to avoid the applicability of the GDPR by locating outside of Europe. Organizations, including advertisers, based outside the EU are subject to the GDPR if they offer goods or services to data subjects in the EU or if they monitor the behavior of EU data subjects.74 According to the GDPR, “monitoring” includes the tracking of individuals on the internet, including to create profiles, “particularly in order to take decisions concerning her or him or for analyzing or predicting her or his personal preferences, behaviors and attitudes.”75 The incredibly wide scope catches a number of advertisers that previously could simply ignore EU data privacy laws. Now, organizations that are not established within the EU, but that are caught by the GDPR’s long-arm jurisdiction, may need to appoint a representative in the EU.76 Moreover, the regulation codifies a number of rights for all EU data subjects and requires organizations to transparently and accurately disclose how they collect, use, store, and share information about individuals.77 There are also more severe consequences for noncompliance under the new regulation. Breaching certain GDPR obligations may cost an organization up to 4 percent of global revenue or 20 million euros in sanctions, whichever is higher.78 When data collected by virtual reality and augmented reality advertising through internet-connected devices is considered to be “personal data,” the requirements of the GDPR also apply, except where the e-Privacy Directive has specifically covered a certain issue.79 When this occurs, the data protection principles regarding data quality, data subject rights, security, international data transfers, and other requirements set forth in the GDPR will continue to be fully applicable. This interpretation is consistent with the general rule that lex specialis overrides lex generalis in EU law.80 As of the time of this writing, the e-Privacy Directive is also undergoing major reform. It is intended to be repealed and replaced by the new Regulation on Privacy and Electronic Communications (“e-Privacy Regulation”), which had not reached its final form as of the time of writing. The draft regulation suggests that significant changes will take place in the privacy and electronic communications field. For instance, a version of the draft regulation introduces a wide extraterritorial scope, applies to both traditional telecoms operators and over-the-top communications services (such as WhatsApp, Facebook

into national law within a certain time frame. For general information, see European Union: Regulations, Directives and other acts at https://europa.eu/european-union/eu-law/legal-acts_en. 74   Article 3(2), GDPR. 75   Recital 24, GDPR. 76   Article 27(1), GDPR. 77   Individual rights provided by the GDPR include the right to be informed, the right of access, the right to rectification, the right to erasure, the right to restrict processing, the right to data portability, the right to object, and rights regarding automated decision making and profiling. See, e.g., Articles 13–18 and 20–21, GDPR. 78   See Articles 83(4)–(5), GDPR. 79   Recital 10, e-Privacy Directive; Article 29 Working Party Opinion 2/2010 on online behavioral advertising, WP171, at 8. http://ec.europa.eu/justice/data-protection/article-29/documentation/ opinion-recommendation/files/2010/wp171_en.pdf. 80   Article 29 Working Party, Opinion 2/2010 on online behavioral advertising, WP171, at 10. http://ec.europa.eu/justice/data-protection/article-29/documentation/opinion-recommendation/fil​ es/2010/wp171_en.pdf.

BARFIELD PRINT.indd 481

13/11/2018 11:53

482  Research handbook on the law of virtual and augmented reality Messenger, Skype, and so on), introduces new rules for communications data, simplifies the cookie rules, and introduces a more effective enforcement system mirroring the GDPR.81 Despite the regulation’s draft status, the EU for some time maintained its ambitious goal to have the regulation implemented concurrently with the GDPR—a deadline that even the Council of the EU admitted was “unrealistic” and that was not, in fact, met.82 Although the final content of the e-Privacy Regulation remains unclear, advertisers and other businesses using cookies and direct marketing should closely follow the developments of the e-Privacy Regulation. Its significance for advertisers and marketers may match that of the GDPR. Virtual reality and augmented reality advertisers have to deal not only with evolving technologies, but also with evolving laws and regulations. After addressing the foundational principles of notice and choice, we will consider the collection and use of geolocation data, and the collection and use of biometric data, due to their likely importance for virtual and augmented reality advertising. Special challenges involving advertising to children will also be noted, as applicable.

2. NOTICE AND CHOICE IN VIRTUAL REALITY AND AUGMENTED REALITY ADVERTISING The foundation of advertising-related data privacy regulations in both the US and the EU is the basic concept that consumers should be given notice of key points regarding collection, use, and sharing of their data before their data is collected, and choice about whether or not to allow such collection, use, and sharing. In defining what these principles mean, the two regions differ in many details, including the level of notice required, how transparent it must be to consumers, and when advertisers must require “opt-in” consent, which can be much more challenging to obtain, versus only offering an “opt-out” option. The US generally favors an “opt-out” model while the EU has a strong preference for “opt-in” consent when sensitive personal data or “profiling” is involved, which will often be the case for virtual and augmented reality advertisements that collect, use, and share information about a consumer’s activities, interests, and interactions with the ads themselves. The way such issues have been playing out already in online and mobile advertising is likely to be highly relevant to virtual and augmented reality advertising as well, but the fine level of balance needed between technology and consumer interest will become even more precarious. As a US example of the importance of notice and consent as technology evolves, in 2017 the FTC announced a $2.2 million payment by Vizio, Inc. in a settlement regarding Vizio’s use of smart televisions to track consumer data without the viewer’s informed consent. According to the FTC’s complaint, Vizio smart televisions: 81   See, e.g., Articles 2–3, 8 and 23, Proposal for a Regulation of the European Parliament and of the Council concerning the respect for private life and the protection of personal data in electronic communications and repealing Directive 2002/58/EC (Regulation on Privacy and Electronic Communications). 82   Council of the European Union, Progress Report on the Regulation on Privacy and Electronic Communications, at 6. http://data.consilium.europa.eu/doc/document/ST-9324-2017-INIT/en/pdf.

BARFIELD PRINT.indd 482

13/11/2018 11:53

Data privacy legal issues in virtual and augmented reality advertising  483 capture second-by-second information about video displayed on the smart TV, including video from consumer cable, broadband, set-top box, DVD, over-the-air broadcasts, and streaming devices . . . In addition, VIZIO facilitated appending specific demographic information to the viewing data, such as sex, age, income, marital status, household size, education level, home ownership, and household value, the agencies allege. VIZIO sold this information to third parties, who used it for various purposes, including targeting advertising to consumers across devices.83

The FTC alleged that “VIZIO touted its ‘Smart Interactivity’ feature that ‘enables program offers and suggestions’ but failed to inform consumers that the settings also enabled the collection of consumers’ viewing data.”84 Clearly, use of similar smart technologies can be leveraged for targeting for virtual and augmented reality advertisements. Future televisions may well enable virtual and augmented reality experiences directly for consumers.85 If they do so, we can be confident that they will try to collect as much data as technologically possible and legally permissible, and to use that data to customize advertising experiences for consumers. Amazon’s Alexa, Google Home, Apple’s Siri, and Microsoft’s Cortana currently record ambient conversations and use that data for machine learning/artificial intelligence.86 Without some applicable legal restriction, it is likely that virtual and augmented reality advertising technologies will do the same thing and become more advanced in their ability to process and use such data. In addition to the FTC’s general enforcement authority, the CAN-SPAM Act is perhaps the most closely relevant to virtual and augmented reality advertising and marketing. Among other things, the CAN-SPAM Act requires an opt-out option that would likely be expanded to apply to targeted virtual and augmented reality advertising experiences, even if it is a stretch to characterize them as electronic mail messages. Rather than sending emails or push notifications to consumer devices, augmented and virtual reality allow the advertiser to project virtual products and banners into physical spaces. For example, individuals wearing Google Glass equipment could receive location-specific advertisements, notifications, and offers right in front of their eyes, depending on where they are walking 83   FTC v. Vizio, Inc., No. 2:17-cv-00758, slip op. (D. N.J. filed Feb. 6, 2017), www.ftc.gov/ system/files/documents/cases/170206_vizio_2017.02.06_complaint.pdf. 84   VIZIO to Pay $2.2 Million to FTC, State of New Jersey to Settle Charges it Collected Viewing Histories on 11 Million Smart Televisions Without Users’ Consent, Federal Trade Commission (Feb. 6, 2017), www.ftc.gov/news-events/press-releases/2017/02/vizio-pay-22-million-ftc-state-newjersey-set​tle-charges-it. 85   Chris Jancelewicz, Watching TV in virtual reality: 5 things you can expect, Global News (Mar. 27, 2017), http://globalnews.ca/news/3329677/watching-tv-in-virtual-reality-5-things-youcan-expect/ (“It’s coming, folks. In December 2016, TV channel Showcase announced a new innovative show, Halcyon, which would combine VR elements with the typical TV show fare. Made up of 10 short-form webisodes and five interactive VR episodes, Halcyon is an interactive police procedural with a twist. You, as a viewer, can truly be part of the show through virtual reality. (This experience is optimized and only available for Oculus Rift and Samsung Gear VR, powered by Oculus. In other words, you need to have a Samsung phone and an Oculus VR set to participate.)”) 86   Tim Moynihan, Alexa and Google Home Record What You Say. But What Happens to That Data? Gear (Dec. 5, 2016) (several times a day, Amazon uses the entire stack of Alexa queries to educate its AI about dialects and casual speech. Microsoft has mysterious fake apartments(!) set up to record and understand natural speech patterns. Google slices and dices the audio it’s already captured, then remixes it to help train its system. All these methods are meant to make your voice assistant smarter in the coming years), www.wired.com/2016/12/alexa-and-google-record-your-voice/.

BARFIELD PRINT.indd 483

13/11/2018 11:53

484  Research handbook on the law of virtual and augmented reality or which stores they are visiting.87 The CAN-SPAM Act could easily be extended to cover these notifications and banners, even if they are not technically email messages. The CANSPAM Act carries hefty penalties, with each separate electronic message subject to fines of up to approximately $40,000 (more if certain aggravating factors are present), and both the advertiser who sent the message and the company whose product or service was promoted in the message can be liable.88 Moreover, criminal penalties—including ­imprisonment— may be imposed for accessing someone else’s computer to send spam without permission or taking advantage of open relays or open proxies without permission.89 However, as indicated previously, application of the CAN-SPAM Act to immersive virtual and augmented reality advertisements may be a bit of a stretch because the CANSPAM Act currently only applies to “any electronic mail message the primary purpose of which is the commercial advertisement or promotion of a commercial product or service (including content on an Internet website operated for a commercial purpose)” (emphasis added).90 Advanced virtual reality and augmented reality advertisements using technologies such as Google Glass and beyond, like the Mokie-Koke ad in The Merchant’s War, will likely be far more immersive than an email message. The FTC requires that email message advertisers use accurate header information and subject lines,91 identify the email messages as ads, tell recipients where the sender is located physically, include clear and conspicuous opt-out options, and honor opt-out requests promptly. But such guidelines are not all directly applicable to virtual and augmented reality advertising. What does it mean to “use accurate header information and subject lines” in a virtual reality or augmented reality advertisement? Does some type of visual and/or aural message need to be sent before the consumer interacts with the ad, identifying the message as an ad and stating the subject matter, like a modern equivalent of the old man waving a sign at Tennison Tarb? Would it be necessary to tell consumers experiencing a virtual or augmented reality ad where the sender is located physically? Assuming a virtual or augmented reality ad was for Coke Zero, does it really help the consumer to know that Coca-Cola is headquartered in Atlanta, Georgia, with offices in many cities worldwide? The FTC clearly prioritizes opt-out requirements. For example, in 2016, the FTC’s administrative complaint against Turn, Inc. (“Turn”)92 resulted in a consent order ­requiring Turn to modify its privacy policy and give meaningful opt-out options to consumers.93 Turn’s privacy policy indicated that consumers could block targeted advertising by using their web browser’s settings to block or limit cookies but instead, allegedly acting 87   Pop Culture Detective, ADmented Reality—Google Glasses Remixed with Google ads, YouTube (Apr. 5, 2012), www.youtube.com/watch?v=_mRF0rBXIeg. 88   CAN-SPAM Act: A Compliance Guide for Business, Federal Trade Commission (published in September 2009 and “edited march 2017 to reflect inflation-adjusted civil penalty maximums”), www.ftc.gov/tips-advice/business-center/guidance/can-spam-act-compliance-guide-business. 89   Id. 90   15 U.S.C. Sec. 7702(2)(A). 91   CAN-SPAM Act, supra note 88. 92   In 2017, Turn Inc. was acquired by a Singapore telecommunications company—Singtel’s digital ad division, Amobee. Singtel-Owned Amobee Strengthens Data, Analytics and Media Buying Capabilities, Amobee (Apr. 10, 2017), www.amobee.com/company/newsroom/amobee-completesacquisi​tion-of-turn-to-expand-digital-marketing-technology-solutions. 93   Digital Advertising Company Settles FTC Charges It Deceptively Tracked Consumers Both

BARFIELD PRINT.indd 484

13/11/2018 11:53

Data privacy legal issues in virtual and augmented reality advertising  485 under instructions from Verizon, Turn used unique identifiers to track millions of Verizon Wireless customers after they blocked or deleted cookies from websites.94 According to the FTC complaint,95 Turn participated in a Verizon Wireless program that allowed Turn and its clients to access demographic information provided by Verizon Wireless about Verizon Wireless users. To create a shared identifier allowing Verizon Wireless and companies participating in the program to uniquely identify each Verizon Wireless user, Verizon Wireless appended unique identifiers known as tracking headers (“X-UIDH headers”) to its users’ mobile Internet traffic. Verizon Wireless injected these X-UIDH headers into all unencrypted web requests for more than 100 million consumers on the Verizon Wireless data network. During the relevant time period, Verizon Wireless users had no means to prevent the transmission of the X-UIDH header.96

Turn then synced the X-UIDH header with other identifiers, including cookies and device advertising identifiers. That synching allowed Turn to maintain the link between the consumer’s browser or device and an identifier associated with behavioral, demographic, or tracking data—even after a consumer had deleted cookies, reset the device advertising identifier, or both. As a result, even if a consumer deleted cookies or reset the device advertising identifier, Turn was still able to recognize the user. In addition, the FTC charged that Turn could recreate unique cookies even after a user had deleted them. Turn’s practices were similar to the use of Flash cookies, which can be respawned to evade blocking by browser settings.97 As a basic principle, if consumers are told they can opt out of virtual reality and augmented reality advertising using a certain method, that statement must be true and advertisers must respect the consumer’s decision to opt out, or the advertiser (and their agents) may be subject to liability. Clear and conspicuous opt-out requirements are perhaps the most useful requirement of CAN-SPAM that should be extended to non-email virtual and augmented reality advertisements. Under current practices, such opt-out options are typically at the bottom of an email message in relatively small text, with a hypertext link. It seems likely that some type of opt-out option notice would also need to be displayed in a virtual or augmented reality advertisement, but it is not clear that such a text display is either the only opt-out method or the best opt-out method. Should a consumer be forced to first see an advertisement when walking down the street and only then be able to opt out of seeing advertisements for such product or service (or from such advertiser) in the future? Or would it make more sense to have a consolidated opt-out option, like the FCC’s “Do Not Call” list? If a virtual reality or augmented reality advertisement is basically equivalent to an email message, then the FTC CAN-SPAM rules described above will apply. However, if Online and Through Their Mobile Devices, Federal Trade Commission (Dec. 20, 2016), www.ftc.gov/ news-events/press-releases/2016/12/digital-advertising-company-settles-ftc-charges-it-deceptively. 94   FTC Complaint, In the Matter of Turn Inc. (No 152 3099), www.ftc.gov/system/files/­ documents/cases/turn_inc_final_complaint.pdf. 95   Id at ¶¶8–10. (FTC Complaint). 96   Id at ¶8. (FTC Complaint). 97   Chris Jay Hoofnagle, Ashkan Soltani, Nathaniel Good, Dietrich J. Wambach, and Mika D. Ayenson, Symposium: Privacy and Accountability in the 21st Century: Behavioral Advertising: The Offer You Cannot Refuse, 6 Harv. L. & Pol’y Rev. 273 (2012).

BARFIELD PRINT.indd 485

13/11/2018 11:53

486  Research handbook on the law of virtual and augmented reality the ad is targeted via a mobile phone or other wireless device, then the FCC may have primary jurisdiction, although even then the FTC can still take action against unfair and deceptive practices under its general Section 5 authority under the FTC Act. The FCC has implemented rules under both CAN-SPAM and the TCPA to help consumers stop unwanted calls, texts, and faxes.98 There is a national “Do Not Call” registry,99 administered by the FCC, so consumers can register their numbers. Consumers also can file an informal complaint with the FCC online if an advertiser has violated any of the FCC rules regarding unwanted calls or text messages.100 In 2015, the FCC strengthened the rules, giving consumers the right to revoke their consent to receive “robocalls” and “robotexts” “in any reasonable way at any time.”101 Private enforcement is also available. For example, class action against the Caribbean Cruise Line and its codefendants resulted in a settlement of more than $56 million—approximately $500 per call to people who received calls offering a “free cruise” in return for taking a short survey.102 In a virtual reality advertisement, a “sponsored avatar” could easily communicate commercial messages to a consumer, either verbally or by other means. Thus, it seems likely that targeted virtual reality or augmented reality advertisements will be subject to enforcement actions under CAN-SPAM (whether by the FTC or the FCC) and the TCPA (by the FCC). The FTC and the FCC both view the “opt-out” rights of consumers as of fundamental importance for online and mobile advertising and marketing messages, which suggests “opt-out” options will also be key for virtual and augmented reality advertising. Similarly, in the EU, the GDPR emphasizes the importance of transparency of data processing operations in the context of profiling and automated decisionmaking as well as more generally throughout the regulation.103 In fact, the principle of “fair and transparent” processing is one of the fundamental principles of EU data protection law.104 Although the general principle of transparency is not new, the GDPR provides stricter requirements and a more comprehensive list of topics on which data subjects must be informed. The GDPR’s transparency requirements are somewhat contradictory. On one hand, the regulation requires an extensive amount of information to be provided; on the other, this information should be provided in a concise, intelligible, and easily accessible way.105 These requirements and the long list of topics that need to be disclosed will present

 98   Stop Unwanted Calls, Texts and Faxes, Federal Communications Commission (Last Updated/ Reviewed Sept. 8, 2017), www.fcc.gov/consumers/guides/stop-unwanted-calls-texts-and-faxes.  99  www.donotcall.gov/. 100   Consumer Complaint Center, Federal Communications Commission, https://­consumercom​ plaints.fcc.gov/hc/en-us/signin?return_to=https%3A%2F%2Fconsumercomplaints.fcc.gov%2Fhc​ %2Fen-us%2Farticles%2F203332190-Spam-Unwanted-Email-and-Text-Messages. 101   FCC Strengthens Consumer Protections Against Unwanted Calls and Texts, Federal Communications Commission (Jun. 18, 2015), www.fcc.gov/document/fcc-strengthens-consumerprotec​tions-against-unwanted-calls-and-texts. 102   Herb Weisbaum, How You Could Get $500 Per Call for Those Unwanted Messages Six Years Ago, NBC News (Jan. 17, 2017), https://www.nbcnews.com/business/consumer/how-you-couldget-50​0-call-those-unwanted-messages-six-n707651. 103   Recital 71, GDPR. 104   See Article 5(1)(a), GDPR. 105   Article 12(1) and Recital 58, GDPR.

BARFIELD PRINT.indd 486

13/11/2018 11:53

Data privacy legal issues in virtual and augmented reality advertising  487 practical challenges to advertisers in the augmented and virtual reality spaces.106 One way to solve this paradox might be to provide layered privacy notices, use icons in addition to traditional privacy notices, and/or provide “just-in-time” privacy notices.107 The requirement of transparency means advertisers in augmented and virtual reality should carefully review their procedures, policies, and public-facing documents to ensure that individuals are provided with the legally required information in a user-centric fashion.108 According to the GDPR, organizations must inform individuals of: (1) the identity and contact details of the data controller; (2) the contact details of the DPO (if the organization has one); (3) the purposes of the data processing and the legal basis for the processing; (4) the legitimate interests of processing; (5) the recipients or categories of recipients of the personal data; and (6) details about any data transfers outside the EU. In addition, the individuals must be informed of: (7) the data retention period or how it is calculated; (8) the existence of individual rights (such as the right to access, rectification, or erasure); (9) the right to withdraw consent; (10) the right to complain to a supervisory authority; and (11) the logic and consequences of the processing where automated decisionmaking, including profiling, is involved.109 The requirement of providing “meaningful information” about automated decisionmaking, including profiling, is relevant for advertisers in the context of augmented and virtual reality.110 From a transparency perspective, advertisers must inform individuals about the existence of automated decisionmaking, including profiling, and “meaningful information about the logic involved, as well as the significance and the envisaged consequences of such processing for the data subject.”111 Therefore, advertisers may be required to inform consumers about the basic algorithmic logic used.112 In addition to the GDPR, and at least until the e-Privacy Regulation is finalized and takes effect, virtual reality and augmented reality advertisers will need to comply with the Electronic Communications Privacy Directive or “e-Privacy Directive” (Directive 2002/58/ EC as amended by Directive 2009/136/EC), which primarily regulates the processing of personal data in the electronic communications sector (such as for telecommunications providers). This directive is particularly important to advertisers because of the so-called cookie rule. Article 5(3) of the e-Privacy Directive requires informed consent before the storing or accessing of information stored in the terminal equipment of an individual. Any storage of cookies or other similar technical equipment or subsequent use of such previously stored equipment to access information must comply with the informed consent  CIPL, Recommendations on Transparency, Consent and Legitimate Interest Under GDPR, at 7.   See International Association of Privacy Professionals (“IAPP”) Webinar “Enhanced Notice Requirements under the GDPR: How to Put Them into Practice,” recorded on August 18, 2017 (Speakers: Elaine Call, CIPP/US, CIPM, CIPT, Senior Privacy Counsel, Sunovion; Sayoko Blodgett-Ford, CIPP/US, Member & Chief Privacy Officer, GTC Law Group; Ilana Westerman, CEO, Create with Context) (available at iapp.org website). 108   See CIPL, Recommendations on Transparency, Consent and Legitimate Interest Under GDPR, at 6. 109   Articles 13–14, GDPR. 110   Article 13(2)(f), GDPR. 111   Article 12(2)(f), GDPR. 112   Article 13(2)(f), GDPR; Recital, 60, GDPR. See also CIPL, Recommendations on Transparency, Consent and Legitimate Interest Under GDPR, at 8–9. 106

107

BARFIELD PRINT.indd 487

13/11/2018 11:53

488  Research handbook on the law of virtual and augmented reality requirement.113 This “information” accessed and/or stored does not have to be personal information as defined by the DPD or the GDPR. Rather, the terminal equipment of users and information contained therein is regarded as part of the users’ private sphere, which triggers the obligations of Article 5(3).114 Advertising network providers must therefore provide “clear and comprehensive information” to users and obtain the users’ prior informed opt-in consent before information in the users’ devices are accessed.115 Another issue that is of high relevance to virtual and augmented reality advertisers involves the use of “profiling” to target ads, or offers, to particular individual consumers and exclude others. The GDPR’s provisions surrounding profiling and automated decisiontaking are almost certain to affect advertising in virtual and augmented reality. Profiling was one of the most controversial topics during the GDPR’s drafting process, and the result represents a compromise.116 According to Article 4(4) of the GDPR, “profiling” refers to any: (1) automated processing of personal data that consists of (2) the use of that data to evaluate certain personal aspects relating to an individual. Examples include analyzing or predicting aspects regarding the individual’s performance at work, financial situation, health, personal preferences, interests, reliability, behavior, location, or movements.117 For example, there is increasing interest in using fitness trackers as part of a larger health ecosystem,118 which could certainly be used to target health-based virtual and augmented reality advertising, perhaps in combination with data from virtual assistant or Internet of Things (IoT) devices—such as Alexa or Siri—which track ambient noises in a room. These could detect a high level of coughing, sneezing, throat clearing, and nose blowing and use machine learning and geolocation to display an ad for a cold and flu remedy (perhaps using geolocation to direct the user to a local drug store or to free fast delivery using Amazon Fresh) on a user’s virtual or augmented reality device. Such 113   Article 29 Working Party Opinion 2/2010 on online behavioral advertising, WP 171, at 8. Available at: http://ec.europa.eu/justice/data-protection/article-29/documentation/opinion-recommendation/ files/2010/wp171_en.pdf. 114   Recital 24, e-Privacy Directive; Article 29 Working Party Opinion 2/2010 on online behavioral advertising, WP 171, at 9. Available at: http://ec.europa.eu/justice/data-protection/article-29/ documentation/opinion-recommendation/files/2010/wp171_en.pdf. 115   Article 5(3), e-Privacy Directive; Article 29 Working Party Opinion 04/2012 on Cookie Consent Exemption, WP 194, at 2. Available at: http://ec.europa.eu/justice/data-protection/­ article-29/documentation/opinion-recommendation/files/2012/wp194_en.pdf; See also Article 29 Working Party Opinion 2/2010 on online behavioral advertising, WP 171, at 3. Available at: http:// ec.europa.eu/justice/data-protection/article-29/documentation/opinion-recommendation/files/2010/ wp171_en.pdf. 116   Rita Heimes, Top 10 Operational Impacts of the GDPR: Part 5—Profiling, IAPP (Jan. 20, 2016), https://iapp.org/news/a/top-10-operational-impacts-of-the-gdpr-part-5-profiling/: “A hotly contested provision of the GDPR, the “profiling” restrictions ultimately adopted were narrower than initially proposed.” 117   Article 4(4) and Recital 71, GDPR. 118   See, e.g., Sonia Mehrotra and Arun Pereira, GOQii: envisioning a new fitness future, Indian School of Business, December 31, 2016, Harvard Business School case study. The “Karma points” in the GOQii system translate to donations for “a social cause.” See https://goqii.com/us-en (“Inspired by Mahatma Gandhi’s journey of 390 KM, for every 390 steps tracked through your activity tracker band, you earn 1 Karma Point. Karma Points convert to monetary donations that’ll go to international relief efforts sponsored by our donor partner, Oxfam India. To contribute, you can choose any cause that’s listed in the Karma section of your GOQii App”).

BARFIELD PRINT.indd 488

13/11/2018 11:53

Data privacy legal issues in virtual and augmented reality advertising  489 tracking of a consumer’s behavior to determine preferences for the purposes of delivering targeted advertisements in a virtual or augmented reality context would be considered profiling under the GDPR. The GDPR restricts decisions that are solely based on automated processing, including profiling. Examples of such decisions include the automatic refusal of an online credit application or e-recruiting practices without human intervention.119 It is also possible that the offering of products or services at different prices to different individuals based on profiling could qualify as such a decision, but this interpretation has not yet been confirmed. Individuals have the right to not be subject to decisions that are based solely on automated processing—including profiling—that produce legal effects concerning them or that otherwise significantly affect them.120 However, such processing is permitted if it is: (1) necessary for entering into (or for the performance of) a contract between the individual and the data controller; (2) authorized by EU or EU Member State law; or (3) based on the explicit consent of the individual.121 When the advertiser relies on consent or the performance of a contract, it will have to ensure that suitable measures have been implemented to safeguard the data subject’s rights, freedoms, and legitimate interests. At a minimum, this includes the right for the consumer to obtain human intervention, to express his or her point of view, and to contest the decision.122 The GDPR further restricts automated decisionmaking based on sensitive data. Such processing is only allowed (1) if the individual has given his or her explicit consent for one or more specified purposes or (2) where the processing is necessary for reasons of substantial public interest, on the basis of EU or EU Member State law.123 Note that the European Data Protection Board may issue additional guidance with regard to profiling.124 Finally, if Frederick Pohl’s vision is prescient, one might expect that virtual and augmented reality advertisers may wish to use so-called “subliminal advertising.” Thus far, such advertising has not proven easy to implement effectively,125 but if that changes one can also expect calls for increased regulation, as occurred in the 1950s, when such calls led to a ban on subliminal advertising in Canada, the United Kingdom, Australia, and the European Union. Although there is no statute banning subliminal advertising in the US (and there are First Amendment issues in efforts to ban the use of subliminal messages in other media content), the FCC has commented on the issue for radio and TV ads at least: “We sometimes receive complaints regarding the alleged use of subliminal techniques in radio and TV programming. Subliminal programming is designed to be perceived on a subconscious level only. Regardless of whether it is effective, the use of subliminal perception is inconsistent with a station’s obligation to serve the public interest because the broadcast is intended to be deceptive.”126

  Recital 71, GDPR.   Article 22, GDPR. 121   Article 22, GDPR. 122   Article 22(3), GDPR. 123   Article 22(4), GDPR. 124   Recital 72, GDPR. 125   Does Subliminal Advertising Actually Work? BBC News Magazine, Jan. 20, 2015, www. bbc.com/news/magazine-30878843. 126   Federal Communications Commission Record, 2001; see www.umich.edu/~onebook/pages/ frames/legalF.html for a discussion of subliminal message laws and controversies. 119 120

BARFIELD PRINT.indd 489

13/11/2018 11:53

490  Research handbook on the law of virtual and augmented reality When it comes to the bedrock data privacy law principles of notice and choice, virtual and augmented reality advertisers will have to face rapidly evolving, and often ambiguous, global laws and regulations, while trying to increase the accuracy of their “profiling” techniques for targeted ads. The challenge will be even greater for advertising to children (or collection of data from children as part of a more general ad campaign), or advertising that using geolocation data or biometric data.

3.  CHILDREN’S ADVERTISING AND CHILDREN’S DATA Virtual reality and augmented reality advertisements for products and services could easily be targeted to children, as shown in the patent application below, which illustrates an augmented reality advertising display that optionally appears to a mother and daughter shopping in a grocery store using an augmented reality device attached to the shopping cart or a wearable gesture-based recognition system.127 The pending published utility patent application from Magic Leap, Inc.128 describes potential advertising and marketing-related uses of the augmented reality technology: The AR system may render virtual content which replicates a celebrity appearance. For example, the AR system may render a virtual appearance of a celebrity chef at a supermarket. The AR system may render virtual content which assists in cross-selling of products. For example, one or more virtual affects may cause a bottle of wine to recommend a cheese that goes well with the wine. The AR system may render visual and/or aural affects which appear to be proximate the cheese, in order to attract a shopper’s attention. The AR system may render one or more virtual affects in the field of the user that cause the user to perceive the cheese recommending certain crackers. The AR system may render virtual friends who may provide opinions or comments regarding the various produces (e.g., wine, cheese, crackers). The AR system may render virtual affects within the user’s field of view which are related to a diet the user is following. For example, the affects may include an image of a skinny version of the user, which is rendered in response to the user looking at a high calorie product. This may include an aural oral reminder regarding the diet.129

Augmented reality ads targeted at children could be a “gamelike” experience like Pokémon Go.130 Niantic reportedly charges sponsored PokéStops up to 50 cents per “visit” by a Pokémon Go user.131 Companies such as Starbucks, Sprint, and Boost Mobile and others 127   One question of potential interest would be product placement in the context of children and augmented or virtual reality. 128   “Methods and Systems for Creating Virtual and Augmented Reality,” U.S. Application No 14/738,877 (filed June 13, 2015). 129   “Methods and Systems for Creating Virtual and Augmented Reality,” U.S. Patent Application No 14/738,877, PCT/US2015/035718, Publication No. 20160026253, application filed June 13, 2015. 130   See Ron Cheng, Is “Pokémon Go” Keeping Tabs On Your Children? Forbes (Aug. 22, 2016), www.forbes.com/sites/roncheng/2016/08/22/is-pokemon-go-keeping-tabs-on-your-­children/​ #4bebeb96247c; Valerie Strauss, Pokémon Go Sparks Concern About Children’s Privacy, Washington Post (July 19, 2016), www.washingtonpost.com/news/answer-sheet/wp/2016/07/19/ po​kemon-go-sparks-concern-about-childrens-privacy/?utm_term=.97c67443b762. 131   Paul Tassi, “Pokémon GO” is Charging Sponsored PokéStops up to 50 Cents per “Visit,”

BARFIELD PRINT.indd 490

13/11/2018 11:53

Data privacy legal issues in virtual and augmented reality advertising  491

Source:  A drawing from US Patent Publication No 20160026253 (available from the US Patent Office on USPTO.gov)

Figure 15.1  An augmented reality friendly monster character bursting out of the cereal boxes in the grocery store aisle for a young child who touches a particular box

have paid to sponsor gyms.132 The Magic Leap patent application also contemplated the “gamification” of advertising, showing a young girl finding an augmented reality fictional monster character called “Gerald” who was “hiding behind cucumbers” and inviting the child to click “to collect points.” At the US federal level, the Children’s Online Privacy Protection Act (“COPPA”) is the most relevant law for virtual and augmented reality advertisements targeted at children or that otherwise collect children’s data, even if they are so-called general audience ads and not specifically directed at children. In 2013, the FTC updated regulations under COPPA to expand the definition of “personal information” to include persistent Which Seems Like a Bad Deal, Forbes Opinion (June 2, 2017), www.forbes.com/sites/insert​ coin/2017/06/02/pokemon-go-is-ripping-off-its-sponsored-pokestops-charging-up-to-50-cents-pervisit/#77c3d4402159. 132   Pokémon Go: Where are Sponsored Gyms Located? http://heavy.com/games/2017/06/poke​ mon-go-where-are-sponsored-gyms-locations/ (Noting “in Japan, over 3,000 McDonald’s ­locations are sponsored gyms”); Ben Brown, 17 Businesses That are Cashing in on Pokemon Go (&  How You Can Too) Marketing blog, Sept. 7, 2017, www.bitcatcha.com/blog/2016/19-businesses-cas​ hing-pokemon-go/.

BARFIELD PRINT.indd 491

13/11/2018 11:53

492  Research handbook on the law of virtual and augmented reality

Source:  A drawing from US Patent Publication No. US20160026253 (available from the U.S. Patent Office on USPTO.gov).

Figure 15.2  A child finds an augmented reality character133 i­dentifiers that can be used to recognize a user over time and across websites, such as cookie IDs and internet protocol (“IP”) 13addresses.134 This was a significant development 133   The figure is further described in the second column of page 113, paragraph 1445 of the published patent application: “FIG. 89D shows another scenario 8908 in which the child is gesturing toward a bonus virtual icon 8942, in the form of a friendly monster or an avatar. The AR system may render unexpected or bonus virtual content to the field of view of the child’s AR system 8903 to provide a more entertaining and engaging user experience for the child.” 134   See 16 C.F.R. 312-2 (“Definitions”) (“Personal information”—item (7) “A persistent

BARFIELD PRINT.indd 492

13/11/2018 11:53

Data privacy legal issues in virtual and augmented reality advertising  493 for online a­ dvertisers—and for virtual and augmented reality advertisers—because the prior definition of personal information generally included only information about a person’s specific identity or activities in the physical world. With this expansion, COPPA broadened the definition to include online identifiers and other online activities in order to better reflect the reality of data collection practices online.135 The revision “effectively prohibits covered operators from using cookies, IP addresses, and other persistent identifiers to track users across websites for most advertising purposes, amassing profiles on individual users, and serving online behavioral advertisements on COPPA-covered websites. Covered operators can, however, use persistent identifiers to support the internal operations of a website.”136 Undoubtedly because of COPPA, Niantic states it does not allow sponsored locations to appear in the Pokémon Go game for players known to be under 13 years old.137 In addition to the FTC, state attorney generals can initiate enforcement actions to protect the personal information of children. For example, the New York Attorney General took action in 2016 against Viacom, Mattel, Hasbro, and Jumpstart,138 because the companies allowed third party marketers and advertising companies to track children’s online i­ dentifier that can be used to recognize a user over time and across different Web sites or online services. Such persistent identifier includes, but is not limited to, a customer number held in a cookie, an Internet Protocol (IP) address, a processor or device serial number, or unique device identifier”); FTC Press Release, FTC Strengthens Kids’ Privacy, Gives Parents Greater Control Over Their Information by Amending Children’s Online Privacy Protection Rule, Federal Trade Commission (Dec. 19, 2012), www.ftc.gov/news-events/press-releases/2012/12/ftc-strengthens-kids-privacy-giv​ es-parents-greater-control-over; see final rule amendments at www.dwt.com/FTC-AnnouncesCOPPA-Rule-Changes-01-17-2013/; 16 C.F.R. 312-2 (Definitions) (“Collects or collection means the gathering of any personal information from a child by any means, including but not limited to:   (1) Requesting, prompting, or encouraging a child to submit personal information online;   (2) Enabling a child to make personal information publicly available in identifiable form. An operator shall not be considered to have collected personal information under this paragraph if it takes reasonable measures to delete all or virtually all personal information from a child’s postings before they are made public and also to delete such information from its records; or   (3) Passive tracking of a child online”). 135   Complying with COPPA: Frequently Asked Questions at 5, Federal Trade Commission (Mar. 2015), www.ftc.gov/tips-advice/business-center/guidance/complying-coppa-frequently-ask​ ed-questions. 136   A.G. Schneiderman announces results of “Operation Child Tracker,” ending illegal online tracking of children at some of nation’s most popular kids’ websites, Attorney General Eric T. Schneiderman (Sept. 13, 2016), https://ag.ny.gov/press-release/ag-schneiderman-announcesresults-operation-child-tracker-ending-illegal-online [hereinafter A.G. Schneiderman]. 137   Niantic Labs, A Note About Sponsored Locations (Dec. 14, 2016) (“So far we have launched 3,700+ of the 10,500 Sprint, Boost Mobile and Sprint at RadioShack stores as PokéStops, Gyms, and real-world charging stations with more coming online daily. Last Thursday we also announced a partnership with Starbucks that turned 7,800 locations into PokéStops and gathering spots for Trainers. With the two combined partnerships, we will have added a total of 18,300 destinations to Pokémon GO . . . Since some of you have asked, we wanted to clarify that sponsored locations do not appear in the game for players under the age of 13”), www.nianticlabs.com/blog/sponsored_loc/. 138   Wendy Davis, Viacom, Mattel, Hasbro and Jumpstart Sanctioned for Tracking Kids, Media Post (Sept. 13, 2016), www.mediapost.com/publications/article/284649/viacom-mattel-hasbro-an​ d-jumpstart-sanctioned-fo.html.

BARFIELD PRINT.indd 493

13/11/2018 11:53

494  Research handbook on the law of virtual and augmented reality activity on children’s websites such as those for Barbie and Hot Wheels.139 According to the settlement, the participating companies agreed to, among other things, monitor third party tracking technologies, adopt procedures to vet third party data collection for COPPA compliance, and notify third parties of the fact they are operating through a website covered by COPPA. The Attorney General swore to continue investigating businesses violating COPPA and holding them accountable.140 One of the alleged violations involved JumpStart: JumpStart integrated a Facebook plug-in into the Neopets website. Facebook plug-ins are integrated into many websites and allow Facebook to track users across the Internet. Facebook uses the tracking information for serving behavioral advertising, among other things, unless the website operator notifies Facebook with a COPPA flag that the website falls is subject to COPPA. JumpStart did not notify Facebook that the Neopets website was directed to children.141

The action of the New York Attorney General offers potentially relevant guidance for virtual reality and augmented reality advertisers if children may be interacting with the ads. COPPA does not govern only websites, but also any “online service” that is “operated for commercial purposes, including any person offering products or services,” that “collects or maintains personal information from or about the users or visitors” to such online service.142 “Internet” is broadly defined to contemplate networks that use TCP/IP and “any successor protocols” to TCP/IP.143 And the FTC has made it clear that it interprets the term to cover new technologies, including, for example: ●

mobile apps that send or receive information online (such as network-connected games, social networking apps, or apps that deliver behaviorally targeted ads), ● internet-enabled gaming platforms, ● plug-ins, ● advertising networks, ● internet-enabled location-based services, ● voice-over internet protocol services, ● connected toys or other Internet of Things (“IoT”) devices.144

  A.G. Schneiderman, supra note 136.   Id. 141   Id. 142   15 U.S.C. Sec. 6501(2)(A) (“operator” means “any person who operates a website located on the Internet or an online service and who collects or maintains personal information from or about the users of or visitors to such website or online service, or on whose behalf such information is collected or maintained, where such website or online service is operated for commercial purposes, including any person offering products or services for sale through that website or online service, involving commerce”). 143   15 U.S.C. Sec. 6501(6) (“The term “Internet” means collectively the myriad of computer and telecommunications facilities, including equipment and operating software, which comprise the interconnected world-wide network of networks that employ the Transmission Control Protocol/Internet Protocol, or any predecessor or successor protocols to such protocol, to communicate information of all kinds by wire or radio”). 144   Children’s Online Privacy Protection Rule: A Six-Step Compliance Plan for Your Business, Federal Trade Commission (June 2017), www.ftc.gov/tips-advice/business-center/guidance/­child​ 139 140

BARFIELD PRINT.indd 494

13/11/2018 11:53

Data privacy legal issues in virtual and augmented reality advertising  495 Thus, virtual reality and augmented reality advertisers who target children will need to comply with COPPA and obtain “verifiable parental consent”145 before collecting personal information from a child under 13 years old (unless they qualify under one of the narrow exceptions, such as collection for internal use only). Traditional methods approved by the FTC have included having a parent or guardian do one of the following: ● ●

● ●

● ●

sign a consent form and return it by fax, mail, or scan; use a credit card, debit card, or other online payment system that provides notification of each separate transaction to the account holder; call a toll-free number staffed by trained personnel; connect to trained personnel via a video conference; provide a copy of a form of government issued ID that can be checked against a database (must delete the identification when the verification process is finished); answer a series of knowledge-based challenge questions that would be difficult for someone other than the parent or guardian to answer.146

In 2015, the FTC also approved the use of “face match to verified photo identification” (“FMVPI”) as an additional method to verify the parent’s identity.147 Thus, an advertiser who wishes to collect personal information (even IP address) from a child may “verify a picture of a driver’s license of other photo ID submitted by the parent and then comparing that photo to a second photo submitted by the parent, using facial recognition technology.”148 All of the approved methods of obtaining “verifiable parental consent” are quite challenging for advertisers to implement on a large scale, which also makes them susceptible to class actions for enforcement. For example, could an advertiser easily determine whether a parent was allowing a child to use their virtual reality headset or their augmented reality mobile device? In the summer of 2017, a class action was filed in California accusing the Walt Disney Company and others of illegally tracking children’s behavior playing online games.149 Counsel for the plaintiffs alleged:

rens-online-privacy-protection-rule-six-step-compliance [hereinafter FTC Six-Step Compliance Plan]. 145   15 U.S.C. Sec. 6501(9) (“verifiable parental consent” means “any reasonable effort (taking into consideration available technology), including a request for authorization for future collection, use, and disclosure described in the notice, to ensure that a parent of a child receives notice of the operator’s personal information collection, use, and disclosure practices, and authorizes the collection, use, and disclosure, as applicable, of personal information and the subsequent use of that information before that information is collected from that child”). 146   FTC Six-Step Compliance Plan, supra note 144. 147   FTC Press Release, FTC Grants Approval for New COPPA Verifiable Parental Consent Method, Federal Trade Commission (Nov. 19, 2015), www.ftc.gov/news-events/press-relea​ ses/2015/11/ftc-grants-approval-new-coppa-verifiable-parental-consent-method. (“Based on an application submitted by Riyo Verified Ltd., the Commission has approved the use of ‘face match to verified photo identification’ (FMVPI) as a method to verify that the person providing consent for a child to use an online service is in fact the child’s parent.”) 148   FTC Six-Step Compliance Plan, supra note 144 at Step 4—final bullet in FTC list. 149   Ashley Cullins, Disney Accused of Illegally Tracking Children Via Apps in New Lawsuit,

BARFIELD PRINT.indd 495

13/11/2018 11:53

496  Research handbook on the law of virtual and augmented reality App developers and their SDK-providing partners can track children’s behavior while they play online games with their mobile devices by obtaining critical pieces of data from  the mobile devices, including “persistent identifiers,” typically a unique number linked to a specific mobile device . . . These persistent  identifiers allow SDK providers to detect a child’s activity across multiple apps and platforms on the internet, and across different devices, effectively providing a full chronology of the child’s actions across devices and apps. This information is then sold to various third-parties who sell targeted online advertising.150

Interestingly, the Disney Children’s Privacy Policy section regarding persistent identifiers indicates that they generally rely on a COPPA exception for collection for “internal purposes only,”151 stating that such data is principally used only in order to provide access to features and activities, to customize content and improve the sites and mobile applications, to provide research and analysis for performance, and for anonymous reporting.152 Disney states: “In the event we collect (or allow others to collect) such information from children on our sites and applications for other purposes, we will notify parents and obtain consent prior to such collection.”153 It seems unlikely that most virtual or augmented reality advertisers could credibly say that the data they collect is for “internal purposes only” unless they are truly never sharing such data or acting on behalf of other parties. At the state level there is also the California’s Privacy Rights for California Minors in the Digital World Act,154 which allows minors (defined as anyone under 18 years old)155 to remove, or to request and obtain removal of, certain content or information posted on an internet web site, online service, online application, or mobile application. It also prohibits an operator of a web site or online service directed to minors from marketing or advertising products or services that minors are legally prohibited from buying.156 The law also prohibits marketing or advertising certain products based on personal information specific to a minor or knowingly using, disclosing, compiling, or allowing a third party to do so. Although mobile applications are more clearly covered, the California privacy rights for minors law also has the potential gap in coverage discussed above, in that the term “operator” does not include “any third party that operates, hosts, or manages, but does not own, a Web site or online service on the owner’s behalf or by processing informaHollywood Reporter (Aug. 4, 2017), www.hollywoodreporter.com/thr-esq/disney-accused-illeg​a​l​l​y​ -tracking-children-apps-new-lawsuit-1026881. 150   Id (quoting the Complaint). 151   Children’s Privacy Policy, The Walt Disney Company (updated July 14, 2016), https:// privacy.thewaltdisneycompany.com/en/for-parents/childrens-online-privacy-policy/. 152   Id. 153   Id. 154   Calif. Bus. & Prof. Code §§ 22580–22582. 155   Calif. Bus. & Prof. Code Sec. 22580(d). 156   Defined as “an Internet Web site, online service, online application, or mobile application, or a portion thereof, that is created for the purpose of reaching an audience that is predominately comprised of minors, and is not intended for a more general audience comprised of adults. Provided, however, that an Internet Web site, online service, online application, or mobile application, or a portion thereof, shall not be deemed to be directed at minors solely because it refers or links to an Internet Web site, online service, online application, or mobile application directed to minors by using information location tools, including a directory, index, reference, pointer, or hypertext link.” Calif. Bus. & Prof. Code Sec. 22580(e).

BARFIELD PRINT.indd 496

13/11/2018 11:53

Data privacy legal issues in virtual and augmented reality advertising  497 tion on behalf of the owner.”157 The CCPA may help close that gap, at least for California residents, by mandating opt in consent to sale of personal information if an organization has “actual knowledge” that a consumer is under 16 years old.

4.  GEOLOCATION, GEOFENCING, AND BEACONS Suppose you are a woman aged 18–24 and you visit a clinic that offers abortion care. Maybe you use Google maps to find it. And while you are waiting for your appointment, you might scroll through Facebook, Instagram, Twitter, or the news. An ad pops up on your phone. It could be for anything—clothing, college courses, a competition. Even if you scroll past it, marketers can capture your ID. If you have location services enabled for that app, marketers can capture your location as well. Anti-choice activists have paid for information about people who fit this profile . . . Once they have your advertising ID, they can send you anti-choice messages like ads for these crisis pregnancy centers.158

Smartphones and GPS-tracking technologies allow marketers to gather an abundance of information about individuals. It has never been easier to track consumer movements and daily routines, because they are almost always equipped with what are effectively personal tracking devices. Information about where consumers live, work, spend free time, and what businesses they frequent is extremely valuable to advertisers, especially when movements can be tracked in real time. Location data combined with augmented reality presents advertisers with novel opportunities. As an example, Pokémon Go allowed businesses to establish “sponsored PokéStops” to lure more players into those b ­ usinesses.159 This real time collection of geolocation data allows marketers to predict consumer habits and preferences and provide location-based advertising using geofencing and beacon technology. However, tracking the exact location of individuals also encompasses significant privacy risks, because geolocation data may provide intimate details about an individual’s private life.160 For example, the continuous tracking of movements may reveal that an individual frequently visits the hospital or attends political demonstrations, or allow otherwise invasive tracking.161 The tracking may therefore indirectly expose data about an individual’s health or political affiliations or other data categorized as sensitive under US law,162 or under EU law.163 The lack of transparency, awareness, and meaningful

  Calif. Bus. & Prof. Code § 22580(f).   How Geo-Fencing Works . . . and How It Can Be Abused, Rewire Multimedia, June 8, 2017, video available at www.youtube.com/watch?v=G3WgvzV5evw. 159  Tassi, supra note 131. 160   Article 29 Working Party, Opinion 13/2011 on Geolocation Services on smart mobile devices, WP185, at 19. http://ec.europa.eu/justice/policies/privacy/docs/wpdocs/2011/wp185_en.pdf. 161   Alex Hern, Uber Employees “Spied on Ex-Partners, Politicians and Beyoncé”, The Guardian (Dec. 13, 2016), www.theguardian.com/technology/2016/dec/13/uber-employees-spying-ex-partnerspol​iticians-beyonce. 162   Curt Woodward & Hiawatha Bray, A Company Sent Anti-Abortion Ads by Phone. Massachusetts Wasn’t Having It, Boston Globe (Apr. 4, 2017), www.bostonglobe.com/business/2017/04/04/healeyhalts-digital-ads-targeted-women-reproductive-clinics/AoyPUG8u9hq9bJUAKC5gZN/story.html. 163   See Article 29 Working Party, Opinion 13/2011 on Geolocation Services on smart mobile devices, WP185, at 7. http://ec.europa.eu/justice/policies/privacy/docs/wpdocs/2011/wp185_en.pdf 157 158

BARFIELD PRINT.indd 497

13/11/2018 11:53

498  Research handbook on the law of virtual and augmented reality consent may be one of the biggest challenges in the collection and processing of geolocation data in virtual reality and augmented reality advertising. Virtual ads can be projected into the real space surrounding the consumer via augmented reality, and such ads can be targeted based on location and even the identity of the person nearby.164 For example, NewAer promises that “[w]ith our SDK you can customize digital signs based upon who is nearby them. This can be triggered from a mobile app, a customer reward card or an iBeacon. With or without connectivity to the Internet.”165 A growing number of commercial enterprises specialize in creating virtual and augmented reality advertising.166 Businesses may seek to display extra information or content when consumers are in or near their stores, or when product packaging is scanned, as has already been happening with beacons and QR codes,167 or in other locations where they want to target potential sales, for example, a running shoe company at a marathon. For example, a nonprofit or political organization may seek to stop negative augmented reality advertising near their offices or ads, as was seen with the prohibition of “geofencing” around Massachusetts healthcare facilities.168 As explained by the Massachusetts Attorney General: Geofencing creates a virtual “fence” around a specified location that is tripped when a person crosses the “fence” with a phone or other mobile device. Once the geofence is tripped, an advertiser will attempt to display an ad in an open app or web browser on the person’s mobile device. The ad is typically tailored to that location and other information about the user. The mobile device also may be tagged so that advertisements can be directly pushed to it whenever the same app or browser page is opened in the future. Consumers may not realize when they installed these apps that the app would disclose their location information for purposes unrelated to the app, including advertising.

When combined with virtual reality or augmented reality advertising, geofencing is science fiction as science fact—straight out of The Merchant’s War—except that it can be even more specifically targeted to a particular individual. There have been various federal legislative initiatives in the US to specifically regulate the collection of geolocation data, thus far without success.169 For example,

164   Mike Butcher, The World of Minority Report is About to be Unlocked, TechCrunch (Apr. 19, 2017), https://techcrunch.com/2017/04/19/the-world-of-minority-report-is-about-be-unlocked-andthe-key-is-inside-your-phone/. 165   The NewAer Advantage, NewAer, https://newaer.com/. 166   United States-based PVI, Inc. (Princeton Video Imaging) markets the L-Vis system (LiveVideo Imaging System). See Digital Broadcasting.com, Princeton Video Image, Inc. Storefront, PVI Overview In-Studio Broadcast Services, SMT, http://www.sportvision.com/pvi-overview (last visited Feb. 27, 2017). 167   Kate Kaye, Ad Dollars Help Fuel Proliferation of Beacons and Proximity Sensors, Advertising Age (Feb. 10, 2017), http://adage.com/article/datadriven-marketing/ad-dollars-fuel-proliferation-bea​ co​ns-proximity/307951/. 168   Maura Healey, AG Reaches Settlement with Advertising Company Prohibiting ‘Geofencing’ Around Massachusetts Healthcare Facilities, Mass.gov (Apr. 4, 2017), www.mass.gov/ago/newsand-updates/press-releases/2017/2017-04-04-copley-advertising-geofencing.html. 169   An official US government site that tracks geolocation privacy legislation may be found at www.gps.gov/policy/legislation/gps-act/. Geolocation Privacy Legislation, GPS.GOV.

BARFIELD PRINT.indd 498

13/11/2018 11:53

Data privacy legal issues in virtual and augmented reality advertising  499 the Commercial Privacy Bill of Rights Act of 2011 would have provided an affirmative opt-in consent for the collection of sensitive information, including geolocation of children and minors.170 President Obama released a draft in 2015,171 but it appears to have stalled. The Geolocation Privacy and Surveillance Act of 2013 (“GPS Act”), which was reintroduced on July 27, 2017,172 would establish a legal framework for use of and access to geolocation information. The GPS Act would “prohibit businesses from disclosing geographical tracking data about its customers to others without the customers’ permission.”173 The Location Privacy Protection Act of 2015 (“LPPA”) was reintroduced in November of 2015 but does not appeared to have progressed since that time.174 The LPPA would prohibit companies from collecting or disclosing geolocation information without affirmative express consent.175 There does appear to be renewed interest in a comprehensive federal privacy law by technology giants such as Facebook and Google, who would prefer a standardized, and stable, approach (provided it is not too restrictive from their perspective) versus evolving state by state variations. The FTC has played a role in enforcing consumer geolocation privacy. Although location-based advertising is not currently subject to any specific federal regulations, FTC guidelines and industry standards do require that location information not be collected or used without consumer consent. The FTC has been involved in various cases where the collection of geolocation information was an issue. For instance, FTC v. Goldenshores Technologies was the first case in which the collection and sharing of geolocation data was the primary focus of the FTC. In Goldenshores, a flashlight app collected its users’ location information without disclosing it in its privacy policy. This decision provided detailed instructions regarding the collection of geolocation information. The settlement agreement required that the application obtain an affirmative express consent from users before collecting such data.176 The FTC formally included geolocation information as “sensitive” in its privacy report from March 2012, in which it also strongly encouraged organizations to acquire consent before the collection or sharing of such data.177 170   Commercial Privacy Bill of Rights Act of 2011, S. 799, 112th Cong. § 202 (2011); also see www.congress.gov/bill/113th-congress/senate-bill/2378. 171   Analysis of the Consumer Privacy Bill of Rights Act, CDT (Mar. 2, 2015), https://cdt.org/ insight/analysis-of-the-consumer-privacy-bill-of-rights-act/. 172   Geolocation Privacy Legislation, supra note 169. 173   Geolocation Privacy Legislation, supra note 169. 174   Geolocation Privacy Legislation, supra note 169, at Location Privacy Protection Act. 175   The Location Privacy Protection Act of 2015, s. 2270. 176   Goldenshores Technologies, LLC, and Erik M. Geidl, In the Matter of Federal Trade Commission (Apr. 9, 2014), www.ftc.gov/enforcement/cases-proceedings/132-3087/goldenshorestech​nologies-llc-erik-m-geidl-matter. 177   FTC Report, Protecting Consumer Privacy in an Era of Rapid Change: Recommendations for Businesses and Policymakers (2012), www.ftc.gov/sites/default/files/­docu​ ments/reports/federal-trade-commission-report-protecting-consumer-privacy-era-rapid-change-rec ommendations/120326privacyreport.pdf. See also FTC Report, Mobile Privacy Disclosures: Building Trust through Transparency (2013), www.ftc.gov/sites/default/files/documents/repor​ ts/federal-trade-commission-report-protecting-consumer-privacy-era-rapid-change-recommendatio ns/120326privacyreport.pdf (in which the FTC further emphasized the requirement of notice and affirmative express consent).

BARFIELD PRINT.indd 499

13/11/2018 11:53

500  Research handbook on the law of virtual and augmented reality Moreover, in February 2015, the FTC published additional guidelines regarding the collection of user location data.178 There has also been some geolocation-related legislative and enforcement activity at the state level. Montana was the first state that passed a geolocation privacy bill in 2013, focusing on governmental inquiries.179 A Geolocation Privacy Protection bill was introduced in Illinois in 2017, and generally provided that a private entity cannot collect, use, store, or disclose geolocation information from a location-based application on a device without first receiving affirmative express consent after complying with specified notice requirements.180 In Massachusetts, a landmark geolocation action involved targeted messages from antiabortion groups, such as messages saying “You Have Choices” and “Pregnancy Help” sent to smartphones of women who were near reproductive health clinics.181 According to the ad agency’s chief executive: “We can reach every Planned Parenthood in the US.” The ads could follow women after they left the clinic, and appear on their cellphones for as long as a month.182 One campaign was aimed at women aged 18 to 24 years old who had been near one of 140 abortion clinics in five cities, and reportedly delivered 2.4 million ad “impressions” and 10,000 “clicks.” Massachusetts Attorney General Maura Healey brought an enforcement action, alleging that the ads illegally used consumer health data,183 because the advertiser targeted people entering reproductive health facilities and methadone clinics.184 The Attorney General argued that using geofencing technology to target ads to consumers would run afoul of the state’s consumer protection laws by tracking consumers’ physical location near or within medical facilities, disclosing that location to third-party advertisers, and targeting individuals with potentially unwanted advertising based on inferences about their private, sensitive, and intimate medical or physical condition, all without their knowing consent.185

The parties settled, and the advertising agency that created the campaign agreed to cease sending such messages. Under the settlement, the company cannot use location data about people who are near a Massachusetts healthcare facility to send them advertisements based on a medical condition.186 Because the advertisements related to sensitive

178   The FTC published guidelines on the collection of location information when the app is not in use. The FTC recommends that mobile apps should clearly disclose their data collection when the app in question is not in active use. Amanda Koulousias, Location, location, location, Federal Trade Commission (Feb. 11, 2015), www.ftc.gov/news-events/blogs/business-blog/2015/02/ location-location-location?utm_source=govdelivery. 179   Did You Know Montana was a Leader on Privacy Law? Neither did Montana, The Washington Post (Aug. 28, 2013), www.washingtonpost.com/news/the-switch/wp/2013/08/28/did-you-knowmontana-was-a-leader-on-privacy-laws-neither-did-montana/?utm_term=.63a103ed3b92. 180  www.ilga.gov/legislation/BillStatus.asp?DocNum=3449&GAID=14&DocTypeID=HB&Le gId=105236&SessionID=91&GA=100. 181   Woodward and Bray, supra note 162. 182   Id. 183  Healey, supra note 168. 184   Allison Grande, Mass. AG Pact Bars Ad “Geofences” at Abortion Clinics, Law360 (Apr. 6, 2017). 185   Id. 186   Id; Healey, supra note 168.

BARFIELD PRINT.indd 500

13/11/2018 11:53

Data privacy legal issues in virtual and augmented reality advertising  501 health information,187 it is not clear whether the Attorney General would have brought an enforcement action for less sensitive targeted virtual reality and augmented reality advertisements. In addition to geofencing, beacon technology also can be leveraged for virtual reality and augmented reality ads. Beacons and other proximity sensors are “small devices that communicate with mobile apps in physical environments.”188 They sense the presence of objects, such as mobile devices, without actually being in physical contact. They are becoming ubiquitous: “pegged at 13 million sensors as of the last quarter of 2016, [the number] has nearly tripled in a year since Q4 2015 when there were 5 million counted.” And advertisers are a key factor in that growth: According to a new report from ProximityDirectory, 42% of the 272 companies offering proximity services operate ad networks that enable advertising targeted based on location data intercepted through beacons, Wi-Fi sensors and other technologies. Around 30% of those firms allow advertisers to target mobile and online ads to people whose devices have been spotted by proximity sensors in specific locations, or measure whether they showed up in a location after being exposed to an ad.189

Beacon-based technologies may be used in targeting virtual reality and augmented reality advertisements just as easily as targeting traditional mobile device ads or text messages. Use of beacons has become increasingly controversial from a data privacy perspective, in both the US and the EU. For example, in 2014, beacons were placed in NYC phone booths for testing purposes.190 The city’s mayor ordered the removal of the beacons within days based on consumer concerns that the beacons would track the movement of users.191 A study by Pew found that location data has become even more precious and more valued by consumers in the smartphone era. Even though the majority of consumers will share their personal information in exchange for something of value, many feel unhappy about what happens to the information collected about them, especially if the data collected is used for purposes other than those for which they were originally collected.192 On the other hand, even though consumers care about privacy, consumers are in most cases ready to give up some privacy if they perceive that they are receiving some benefit in return, for example in the form of price breaks or loyalty programs.193 Virtual reality and augmented reality advertisers can be expected to view use of beacons as “fair game.” Reportedly, 187   Press release, April 4, 2017, www.mass.gov/ago/news-and-updates/press-releases/2017/201704-04-copley-advertising-geofencing.html (“‘While geofencing can have positive benefits for consumers, it is also a technology that has the potential to digitally harass people and interfere with health privacy,’ said AG Healey. ‘Consumers are entitled to privacy in their medical decisions and conditions. This settlement will help ensure that consumers in Massachusetts do not have to worry about being targeted by advertisers when they seek medical care.’”) 188   Kate Kayle, Ad Dollars Help Fuel Proliferation of Beacons and Proximity Sensors, Advertising Age (Feb. 10, 2017), http://adage.com/article/datadriven-marketing/ad-dollars-fuelproliferati​on-beacons-proximity/307951/. 189   Id. 190  www.vanityfair.com/news/politics/2014/10/new-york-city-phone-booth-beacons-tracking. 191   See https://iapp.org/news/a/public-beacons-make-appearance-in-nyc-quickly-taken-down. 192   See the Pew research at http://www.pewinternet.org/2016/01/14/privacy-and-informationsha​ring/. 193  https://iapp.org/news/a/should-you-really-be-that-scared-of-your-in-store-retail-analytics/.

BARFIELD PRINT.indd 501

13/11/2018 11:53

502  Research handbook on the law of virtual and augmented reality apps have increasingly incorporated ultrasonic tones to track consumers. They ask permission to access your smartphone microphone, then listen for inaudible “beacons” that emanate from retail stores, advertisements, and even websites. If you’re not paying attention to the permissions you grant, you could be feeding marketers information about your online browsing, what stores you go to, and what products you like and dislike without ever realizing it.194

The FTC’s first case against a retail tracking company, which eventually led to a settlement, involved Nomi Technologies (“Nomi”).195 According to the FTC, Nomi Technologies’ privacy statement was inconsistent with Nomi’s actual practices. Nomi provided technology that allowed its clients to track customers by, for example, providing data on how much time customers spent in different sections of the store. Nomi’s privacy policy stated that consumers could opt out of such retail tracking. However, not all of its clients’ retail locations offered such an option. Moreover, Nomi did not require that its clients provide notice to their customers. The FTC concluded that these factors rendered Nomi’s privacy policy deceptive. Even though the case focused on the company’s deceptive privacy statement, it can also be viewed as the FTC’s first step in implementing the online world’s notice and choice regime to retail tracking. Virtual reality and augmented reality advertisers should be prepared to meet an affirmative requirement to disclose their tracking practices and acquire consumers’ consent before using beacon technology in the United States, and, as will be discussed below, in the EU as well. The e-Privacy Directive regulates the use of geolocation data in the EU, but it has been interpreted to only apply to data processed by public electronic communication services and networks (that is, telecommunications operators).196 According to the Directive, location data may only be processed when made anonymous, or with the user’s or subscriber’s consent “to the extent and for the duration necessary for the provision of a value added service.”197 Information society services (companies providing location services and applications using base stations, GPS, and WiFi) are therefore explicitly excluded from the scope of the e-Privacy Directive.198 Although not specifically mentioned in the DPD, the prevalent view in the European Union is that geolocation data should be interpreted as “personal data,” because companies can indirectly identify individuals.199 It is not yet considered “sensitive data,” but it has received an increasing amount of attention in recent years due to its high value and the 194   Lily Hay Newman, Hundreds of Apps Can Listen for Marketing “Beacons” You Can’t Hear, Wired (May 2, 2017), www.wired.com/2017/05/hundreds-apps-can-listen-beacons-canthear/ (“Beacon technology is also showing up in more physical locations. While the researchers didn’t find any ultrasonic tones being broadcast out on a sampling of television programming from seven countries, they did find that four of the 35 retail stores they visited around Germany did have beacons installed”). 195   In re Nomi Technologies, Inc. www.ftc.gov/enforcement/cases-proceedings/132-3251/nomitechn​ologies-inc-matter. 196   Article 29, Working Party Opinion 13/2011 on Geolocation services on smart mobile devices, WP 185, at 7. http://ec.europa.eu/justice/policies/privacy/docs/wpdocs/2011/wp185_en.pdf. 197   Article 9, e-Privacy Directive. 198   Article 1(2), e-Privacy Directive; Article 29 Working Party Opinion 13/2011 on geolocation services on smart mobile devices, WP185, at 9. http://ec.europa.eu/justice/policies/privacy/docs/ wpdocs/2011/wp185_en.pdf 199   See Article 2(a) and Recital 26, DPD; Article 29 Working Party Opinion 13/2011 on

BARFIELD PRINT.indd 502

13/11/2018 11:53

Data privacy legal issues in virtual and augmented reality advertising  503 high risks associated with the data.200 The sensitivity of processing patterns of location data also means that advertisers should base their geolocation processing activities on an individual’s prior informed consent.201 The French Data Protection Authority (“CNIL”) has also highlighted the requirement of prior consent before the collection or processing of individuals’ geolocation data for commercial purposes.202 The high standards for obtaining valid consent mean that advertisers cannot hide the consent in the small print of general terms and conditions, and the possibility to opt out alone may not be sufficient.203 In order to obtain valid consent, advertisers should obtain a freely given, specific, and informed indication of the data subject’s wishes.204 The individual should also be informed about the terms of the processing of his or her geolocation data. Furthermore, a new consent should be obtained if the purposes of the processing materially change, for example, if the company suddenly wishes to share the collected data with third parties.205 Data subjects should also be able to easily withdraw their consent.206 As an example, Niantic Labs was threatened with a lawsuit in Germany following the launch of the location-based application Pokémon Go. The Federation of German Consumer Organizations was prepared to sue Niantic if Niantic did not alter its terms of service and privacy policy to comply with European data protection law. The Federation criticized the lack of transparency and the insufficient consents obtained from consumers.207 The Irish Data Protection Commissioner also reacted to the launch of Pokémon Go by publishing guidelines for applications collecting geolocation data. The Commissioner acknowledged that businesses often wish to utilize customer location data to provide location-based content and advertising. According to the Commissioner, the precise pattern of an individual’s movements has the potential to reveal highly intimate details about that individual’s personal life. Businesses may find such data valuable because they allow the specific targeting of services to specific individuals, which poses serious privacy risks. The Commissioner further highlighted that informed consent was often the most appropriate basis for processing personal location data.208 The GDPR does not radically change the treatment of geolocation data. However, it does explicitly include location data in the definition of personal data.209 This means that g­ eolocation services on smart mobile devices, WP185, at 9. http://ec.europa.eu/justice/policies/ privacy/docs/wpdocs/2011/wp185_en.pdf. 200   Dr. Andres Guadamuz, Pokemon Go: Augmented Reality Tests IP, WIPO: Magazine (Feb. 2017) www.wipo.int/wipo_magazine/en/2017/01/article_0005.html. 201   Article 29 Working Party Opinion 13/2011 on geolocation services on smart mobile devices, WP185, at 14. Available at: http://ec.europa.eu/justice/policies/privacy/docs/wpdocs/2011/ wp185_en.pdf. 202   CNIL, Commerce et données personnelles: Le suivi des consommateurs. www.cnil.fr/sites/ default/files/atoms/files/_commerce-donnees_perso_suivi_consommateurs.pdf 203   Opinion 13/2011, at 14. 204   Article 2(h), DPD. 205   Opinion 13/2011, at 15. 206   Opinion 13/2011, at 16. 207  http://fortune.com/2016/07/20/pokemon-go-germany-privacy/. 208   See Data Protection Commissioner, Guidance Note for Data Controllers on Location Data. www.dataprotection.ie/viewdoc.asp?DocID=1587&ad=1. 209   Article 4(1), GDPR.

BARFIELD PRINT.indd 503

13/11/2018 11:53

504  Research handbook on the law of virtual and augmented reality the processing of such data is subject to the expansive data protection requirements set forth in the GDPR. Furthermore, for organizations relying on consent, the GDPR makes it considerably harder for advertisers and other organizations to obtain such valid consent. The new elevated consent must be “freely given, specific, informed, and unambiguous” and given by a “statement or by a clear affirmative action.”210 Opt-outs, pre-ticked boxes, or inactivity continue to be insufficient bases for the collection of geolocation data.211 Furthermore, a virtual reality or augmented reality advertiser relying on the individual’s consent must be able to demonstrate that the data subject has given his or her consent to the data protection operation, and data subjects have the right to withdraw their consent at any time.212 In order to meet the specificity requirement, the consent must be clearly distinguishable from other matters.213 In other words, consent from a covered European resident cannot be buried or hidden in terms and conditions.214 Geolocation will almost inevitably be a significant part of virtual and augmented reality advertising, and must be implemented in accordance with US and EU laws and regulations. What other techniques can such advertisers use to make the ad targeting even more effective? Biometrics and facial recognition techniques provide one answer, and also raise unique data privacy challenges.

5.  BIOMETRICS AND FACIAL RECOGNITION Hello, Mr Yakamoto. Welcome back to the Gap. How did those assorted tank tops work out for you?215

In addition to geolocation data collection, facial, optic, biometric, haptic, audio, and Internet of Things (“IoT”) data collection and use is on the rise. For example, advertisers often seek to collect data about consumer interactions with their ads, in order to measure the campaign’s effectiveness and return on investment.216 Virtual reality and

  Article 4(11) and Recital 32, GDPR.   Recital 32, GDPR. 212   Articles 7(1), 7(3), and Recital 42, GDPR. 213   Article 7(2), GDPR. 214   The draft e-Privacy Regulation addresses the tracking of the location of terminal equipment. However, the Article 29 Working Party, the European Data Protection Supervisor, and the European Parliament’s Civil Liberties Committee have heavily criticized the proposed “device tracking” exception in Article 8(2)(b). It is likely that the final version of the e-Privacy Regulation will take a much more stringent approach to tracking. See, e.g., Jennifer Baker, LIBE Submits More Than 800 Amendments to ePrivacy Regulation, IAPP, https://iapp.org/news/a/ libe-submits-more-than-800-amendments-to-eprivacy-regulation/. 215   Minority Report movie (2002). Clips are available on YouTube, for example at www.youtube. com/watch?v=ITjsb22-EwQ. Kalev Leetaru, From Cellphone Tracking to Facial Recognition: Minority Report is Here, Forbes (Oct. 20, 2016, 1:38 PM), www.forbes.com/sites/kalevleetaru/2016/10/20/ from-cellphone-tracking-to-facial-recognition-minority-report-is-here/#1aa98dbb115a. 216   For example, see Comscore’s tools for advertisers: www.comscore.com/Products/AdvertisingAnalytics (“In today’s media ecosystem, media buyers and sellers must be able to measure, evaluate and optimize every aspect of a campaign—regardless of the platform. comScore provides the end-to-end tools and insights needed to maximize and prove ROI.”). 210 211

BARFIELD PRINT.indd 504

13/11/2018 11:53

Data privacy legal issues in virtual and augmented reality advertising  505 augmented reality advertisers have opportunities to collect and leverage additional data. The company SightCorp offers real time facial analysis and eye tracking to help businesses “[d]etect and measure facial expressions like happiness, surprise, sadness, disgust, anger and fear” and “[u]nderstand what motivates . . . customers and what grabs their attention, by analyzing head movements, eye location and attention time.” 217 The augmented reality company ModiFace scans its users’ faces and allows them to virtually try on makeup before they make the decision to buy a product.218 Mannequins have been equipped with cameras and facial recognition technology to track “shoppers’ movements in some stores in order to improve layout and targeted marketing.” The facial recognition technology was also used to record shoppers’ gender, race, and estimated age.219 Such information could potentially be of great value to advertisers, who already track user interactions with ads on websites.220 Special data privacy risks and threats are involved in the processing of biometric data and facial recognition technologies because they allow individuals to be tracked and/or profiled. In the summer of 2017, the New York Times reported on a prototype virtual reality game that users can control through a combination of thoughts (using a headset that measures their electrical brain activity with electroencephalography) and eye movements.221 The article explained: “The algorithms learn from your behavior. Before playing the game, you train them to recognize when you are focusing your attention on an object. A pulse of light bounces around the virtual room, and each time it hits a small colored ball in front of you, you think about the ball. At that moment, when you focus on the light and it stimulates your brain, the system reads the electrical spikes of your brain activity.”222 Facebook “hopes to have a system that allows people to type with their thoughts five times faster than they now type using a smartphone keyboard.”223 If Facebook, which earned more than $9 billion in ad revenue in the second quarter of 2017 alone,224 is able to read its more than one billion daily active users’ brain activity and thoughts/interests even in this rudimentary manner, consider how much additional power it could offer as an advertising platform for commercial advertisements, as well as in terms of choices Facebook makes for users’ news feeds. For example, “in 2012, it was  http://sightcorp.com/   Josh Constine, Augmented Reality for Trying on Makeup is a Booming Business, TechCrunch, https://techcrunch.com/2016/01/19/facial-precognition/. 219   Liz Kilmas, “Spooky” Mannequins Outfitted with Facial Recognition Cameras Spy on Shoppers, The Blaze, www.theblaze.com/stories/2012/11/21/spooky-mannequins-outfitted-with-facial-reco​ gnition-cameras-spy-on-shoppers/. 220   Simon Hill, How Much Do Online Advertisers Really Know About You? We Asked an Expert, Digital Trends (June 27, 2015), www.digitaltrends.com/computing/how-do-advertisers-trackyou-on​line-we-found-out. 221   One area of potential interest would be the data associated with the thoughts and the set of neuronal activity and circuits. For example, it would be interesting to study whether such data are protected by the First Amendment (thought as a precursor to speech). 222   Cade Metz, A Game You Can Control with Your Mind, New York Times (Aug. 27, 2017) (online). 223   Id. 224   Marty Swant, Facebook Raked in $9.16 Billion in Ad Revenue in the Second Quarter of 2017: It Now Has 1.32 Billion Daily Active Users, AdWeek (July 26, 2017), www.adweek.com/digital/ facebook-raked-in-9-16-billion-in-ad-revenue-in-the-second-quarter-of-2017/. 217 218

BARFIELD PRINT.indd 505

13/11/2018 11:53

506  Research handbook on the law of virtual and augmented reality discovered that travel website Orbitz was showing Mac users pricier hotel options than PC users. Later the same year, the Wall Street Journal reported that the Staples website was tracking visitor’s locations and only applying price discounts if there was a competitor store within 20 miles of them.”225 In 2013, the US Senate warned: One of the primary ways data brokers package and sell data is by putting consumers into categories or “buckets” that enable marketers – the customers of data brokers – to target potential and existing customers. Such practices in many cases may serve the beneficial purpose of providing consumers with products and services specific to their interests and needs. However, it can become a different story when buckets describing consumers using financial characteristics end up in the hands of predatory businesses seeking to identify vulnerable consumers, or when marketers use consumers’ data to engage in differential pricing.226

Facebook clearly has a strong interest in facial recognition. Facebook first introduced its automatic facial recognition feature in 2011.227 The “tag suggestion” feature “deploy[ed] a sophisticated facial recognition tool to automatically match pictures with names. When a Facebook user uploads a photo of friends, the ‘tag suggestion’ feature can automatically pull up the names of the individuals in the image. The facial recognition software was developed by . . . Face.com, which Facebook acquired.”228 The Electronic Privacy Information Center advocacy group filed a complaint with the FTC over Facebook’s use of automatic tagging.229 The FTC has a consent order with Facebook that subjects the company to audits over its privacy policies for the next 20 years.230 It seems likely that Facebook would be interested in leveraging facial recognition data for advertising and marketing purposes, and its legal disputes will provide guidance to other advertisers who share that interest. Although thus far there is no federal regulation specific to the collection of biometric data, the Illinois Biometric Information Privacy Act (“BIPA”),231 which prevents collection of a customer’s biometric identifier or biometric information without first informing the subject in writing regarding the collection, purpose of use, and storage, and obtaining a written release.232 Specifically, the Illinois Biometric Information Privacy Act provides:233 No private entity may collect, capture, purchase, receive through trade, or otherwise obtain a person’s or a customer’s biometric identifier or biometric information, unless it first:  Hill, supra note 220.   U.S. Senate Committee on Commerce, Science, and Transportation, Office of Oversight and Investigations, Majority Staff, A Review of the Data Broker Industry: Collection, Use, and Sale of Consumer Data for Marketing Purposes, Staff Report for Chairman Rockefeller Dec. 18, 2013. 227   See Sarah Jacobsson Purewal, Why Facebook’s Facial Recognition is Creepy, PCWorld (June 8, 2011). 228   Somini Sengupta and Kevin J. O’Brien, Facebook Can ID Faces, But Using Them Grows Tricky, New York Times (Sept. 21, 2012). 229  https://epic.org/privacy/facebook/. 230  www.nytimes.com/2012/09/22/technology/facebook-backs-down-on-face-recognition-in-eu​ rope.html. 231   Illinois Biometric Information Privacy Act, 740 Ill. Comp. Stat. 14/1 et seq. 232   740 ILCS 14/15; www.ilga.gov/legislation/ilcs/ilcs3.asp?ActID=3004&ChapterID=57. 233  www.ilga.gov/legislation/ilcs/ilcs3.asp?ActID=3004&ChapterID=57. 225 226

BARFIELD PRINT.indd 506

13/11/2018 11:53

Data privacy legal issues in virtual and augmented reality advertising  507 (1) informs the subject or the subject’s legally authorized representative in writing that a biometric identifier or biometric information is being collected or stored; (2) informs the subject or the subject’s legally authorized representative in writing of the specific purpose and length of term for which a biometric identifier or biometric information is being collected, stored, and used; and (3) receives a written release executed by the subject of the biometric identifier or biometric information or the subject’s legally authorized representative.234

Biometric privacy laws are in place in Texas,235 as well as in Washington,236 and other states are expected to follow.237 However, there are likely to be differences in such laws that could be relevant to virtual and augmented reality advertisers. For example, unlike the Illinois and Texas statutes, Washington does not expressly cover facial recognition techniques such as recordings or scans of face geometry in its definition of “biometric identifier.”238 Washington’s definition also excludes “physical or digital photographs, video or audio recording or data generated therefrom,” as well as certain health-related data processed pursuant to Health Insurance Portability and Accountability Act of 1996. In 2015, Facebook was sued under the Illinois BIPA law; in 2016, the District Court denied Facebook’s motion to dismiss, finding that plaintiffs had stated a claim under the BIPA and rejecting Facebook’s argument that photographs and information derived from photographs were excluded from the BIPA’s definitions of “biometric identifier” and “biometric information.”239 According to the BIPA, a “biometric identifier” includes “a retina or iris scan, fingerprint, voiceprint, or scan of hand or face geometry,” but excludes items such as photographs, demographic data, and physical descriptions.240 “Biometric information” refers to “any information, regardless of how it is captured, converted, stored, or shared, based on an individual’s biometric identifier used to identify an individual.”241 The Illinois lawsuit was pending at the time of writing and is widely expected to define the rules for other major technology companies, including Google.242   740 ILCS 14/15.   Texas Statute on the Capture or Use of Biometric Identifier, Tex. Bus. & Com. Code Ann. Sec. 503.001. 236   Washington House Bill 1493, signed into law in May of 2017. 237   A Host of Biometric Privacy/Facial Recognition Bills Currently Circulating in State Legislatures, National Law Review, Feb. 22, 2017, www.natlawreview.com/article/host-biometricprivacyfacial-recognition-bills-currently-circulating-state (Alaska HB 72, Connecticut HB 5522, Montana HB 581, New Hampshire HB 523). 238  http://lawfilesext.leg.wa.gov/biennium/2017-18/Pdf/Bills/House%20Bills/1493-S.pdf (“‘Biometric identifier’ means data generated by automatic measurements of an individual’s biological characteristics, such as a fingerprint, voiceprint, eye retinas, irises, or other unique biological patterns or characteristics that is used to identify a specific individual. ‘Biometric identifier’ does not include a physical or digital photograph, video or audio recording or data generated therefrom, or information collected, used, or stored for health care treatment, payment, or operations under the federal health insurance portability and accountability act of 1996.”) 239   In re Facebook Biometric Information Privacy Litigation, Civ. Action No. 15-cv-03747-JD, N.D. Calif. 240   740 ILCS 14/10. 241   Id. 242   Meg Graham, Illinois Biometrics Lawsuits May Help Define Rules for Facebook, Google, Chicago Tribune (Jan. 17, 2017) (available at www.chicagotribune.com/bluesky/originals/ctbiometric-illinois-privacy-whats-next-bsi-20170113-story.html) (“The suit may hinge on whether 234 235

BARFIELD PRINT.indd 507

13/11/2018 11:53

508  Research handbook on the law of virtual and augmented reality Facebook continues to lobby against legislation that would bar the use of its automated facial recognition technology.243 In the EU, biometric data is almost always personal data, but it does not always qualify as “sensitive” under the GDPR. It falls within the definition of sensitive data when it is processed to “uniquely identify” an individual.244 According to the regulation, biometric data is personal data that results from “specific technical processing” relating to the physical, physiological, or behavioral characteristics of a person, which “allow or confirm the unique identification of that natural person.” Examples of this kind of data include facial images or dactyloscopic (fingerprint) data.245 Facial recognition is also considered to be within the scope of biometrics because it often allows for the unique identification of an individual.246 However, recital 51 of the GDPR clarifies that processing photographs does not automatically mean that biometric data is processed. Instead, photographs are biometric data within the meaning of the GDPR only when they are processed using special technical means that allow for the unique identification or authentication of a person.247 Thus, the regulation makes a distinction between biometric data that is used for purposes of uniquely identifying an individual and other biometric data. The general rule under the GDPR is that the processing of biometric data is prohibited.248 However, the GDPR provides ten exceptions to this prohibition. The prohibition of processing biometric data does not apply when: (1) the individual has given his/ her explicit consent; (2) the processing is necessary to carry out the obligations and exercise specific rights of the controller or of the data subject in the field of employment or social protection law so far as it is authorized by law providing for appropriate safeguards for the fundamental rights and interests of the individual; (3) the processing is necessary for vital interests of an individual; (4) processing is carried out in the course of legitimate activities by a foundation, association or other nonprofit body with a political, philosophical, religious, or trade union aim and the processing relates solely to the (former) members; (5) processing relates to personal data manifestly made public by the individual; (6) processing is necessary for the establishment, exercise, or defense of legal claims; (7) processing is necessary for reasons of substantial public interest; (8) processing is necessary for the purposes of preventive or occupational medicine or other treatment on the basis of EU or Member State law or contract; (9) processing is necessary for reasons of public interest in the area of public health; or (10) processing is necessary for archiving purposes in the public interest, scientific or historical research, or statistical purposes.249 users were sufficiently informed about how their Facebook data would be used, said Matthew Kugler, an assistant professor at the Northwestern University Pritzker School of Law.”). 243   April Glaser, Facebook is Using an “NRA Approach” to Defend its Creepy Facial Recognition Programs, Slate.com (Aug. 4, 2017). 244   Article 9, GDPR. 245   Article 4(14), GDPR. 246   Article 29 Working Party Opinion 02/2012 on facial recognition in online and mobile services, at 1. http://ec.europa.eu/justice/data-protection/article-29/documentation/opinion-­recomm​ endation/file​s/2012/wp192_en.pdf. 247   Recital 51, GDPR. 248   Article 9(1), GDPR. 249   Article 9(2), GDPR.

BARFIELD PRINT.indd 508

13/11/2018 11:53

Data privacy legal issues in virtual and augmented reality advertising  509 Explicit consent of the individual will likely be the most important exception for the advertising and marketing sectors. European data protection authorities criticized Facebook for breaching EU data protection law by introducing the facial recognition feature without first obtaining users’ consent. The feature was launched without informing users and obtaining user consent, which sparked criticism from European privacy advocates and data protection authorities. Facebook was pressured to disable the feature for new Facebook users within the EU and to delete the data it already collected.250 In the EU, advertisers and other entities processing biometric data should also prepare for more restrictive national legislation. The GDPR allows Member States to maintain or introduce further conditions and limitations to the processing of genetic, biometric, and other data concerning health.251 Moreover, processing biometric data of European residents often includes additional obligations. For example, a Data Protection Impact Assessment (“DPIA”) must be conducted prior to processing if sensitive biometric data is processed on a large scale.252 Such an assessment is also necessary where the processing is “likely to result in a high risk to the rights and freedoms of natural persons.”253 Conducting a DPIA is particularly relevant when a new data processing technology is used.254 This is important for advertisers, because the use of biometric data in the context of augmented and virtual reality is likely to contain this kind of “new technology” thought to threaten the rights and freedoms of individuals. Furthermore, although a DPIA is not strictly required for all processing activities, conducting one in unclear situations is highly recommended.255 The consequences of noncompliance are significantly more expensive than the costs of conducting a DPIA. The supervisory authority may impose an administrative fine of up to 10 million euros or 2 percent of total worldwide annual turnover, whichever is higher, for failure to conduct a DPIA.256 The purpose of a DPIA is to build and demonstrate regulatory compliance. This is done by describing the processing activities, assessing the necessity and proportionality of the processing, managing the risks to the rights and freedoms of individuals, and demonstrating appropriate compliance measures.257 No standard form exists for DPIAs, 250   Irish Data Protection Commissioner, Report of Review of Facebook Ireland’s Implementation of Audit Recommendations Published—Facebook Turns Off Tag Suggest in the EU, www.­dataprotec​ tion.ie/docs/21-09-12-Press-Release-Facebook-Ireland-Audit-Review-Report/i/1233.htm; Irish Data Protection Commissioner, Facebook Ireland Ltd, Report of Re-Audit (September 21, 2012), at 8–9; www.dataprotection.ie/documents/press/Facebook_Ireland_Audit_Review_Report_21_Sept_2012. pdf. See also www.datenschutz-hamburg.de/uploads/media/PressRelease-2011-11-10-Facebook_ BiometricDatebase.pdf; Sengupta and O’Brien, supra note 228. 251   Article 9(4), GDPR. 252   Article 35(3)(b), GDPR. Note that the term “Privacy impact assessment” is often used to refer to the DPIA. 253   Article 35(1), GDPR. 254   Article 35(1), Recitals 89 and 91, GDPR. 255   Article 29 Working Party Guidelines on Data Protection Impact Assessment (DPIA) and determining whether processing is “likely to result in a high risk” for the purposes of Regulation 2016/679, WP 248, at 7. 256   Article 29—Guidelines April 4, 2017, see ec.europa.eu/newsroom/document.cfm?doc_id=44​ 137. 257   Article 29 Working Party Guidelines on Data Protection Impact Assessment (DPIA) and

BARFIELD PRINT.indd 509

13/11/2018 11:53

510  Research handbook on the law of virtual and augmented reality but the regulation does provide the minimum requirements of such an assessment.258 According to the GDPR, the DPIA must contain at least: (1) a systemic description of the processing operations; (2) the legitimate interest pursued by the controller; (3) the necessity and proportionality of the processing operations in relation to the purposes; and (4) an assessment of the risks to the rights and freedoms of data subjects.259 The data controller must also consult the supervisory authority prior to starting any processing activities in certain situations. The supervisory authority should be consulted when the DPIA indicates that “the processing would result in a high risk in the absence of measures taken by the controller to mitigate the risk.”260 The supervisory authority will provide written advice to the controller within eight weeks if he/she believes that the processing would infringe the GDPR, especially if the controller has not adequately identified or mitigated the risk.261 In both the US and the EU, the jury is still out as to the extent of the limits on use of bio­metric data and facial recognition for virtual reality and augmented reality advertising. What we can safely predict is that there will be business pressure on advertisers to test those limits.

6. CONCLUSION Virtual reality and augmented reality advertisers must comply with existing (and expanding) advertising-related privacy laws, regulations, and industry standards in the United States and abroad. Key principles of notice and choice must be translated into the context of these new technologies, almost in real time, as the technologies continue to become more powerful. In addition, special challenges related to children’s data, geolocation, geofencing and beacons, and facial recognition and other biometric data collection are likely to be made even more acute as virtual reality and augmented reality advertising evolves. There is no doubt that, absent legal constraints, advertisers will seek to make ad targeting and analytics as effective as possible, using all available technologies and leveraging even larger volumes of data, at an even more granular level of detail. They already have more than 200 billion reasons to do so.262 From a legislative and policy perspective, how close will we get to making the advertising described in The Merchant’s War a reality? Will cutting edge technologies such as direct brain interactions,263 combined with virtual and augmented reality advertisements, be used to make the consumer’s situation even

determining whether processing is “likely to result in a high risk” for the purposes of Regulation 2016/679, WP 248, at 4. 258   Id at 19. See Annex 2 for a useful table of criteria for an acceptable DPIA. 259   Article 35(7), GDPR. 260   Article 36, GDPR. 261   Article 36, GDPR. 262  Handley, supra note 6. 263   Rolfe Winkler, Elon Musk launches Neuralink to connect brains with computers: startup from CEO of Tesla and SpaceX aims to implant tiny electrodes in human brains, The Wall Street Journal (Mar. 27, 2017), www.wsj.com/articles/elon-musk-launches-neuralink-to-connect-brainswith-co​mputers-1490642652.

BARFIELD PRINT.indd 510

13/11/2018 11:53

Data privacy legal issues in virtual and augmented reality advertising  511 worse than that described by Frederik Pohl? We do not know the answers, but for some strange reason we do feel like we need to drink a Mokie-Koke now.

SOURCES Cases Griswold v. Connecticut, 381 U.S. 479 (1965) In re Facebook Biometric Information Privacy Litigation, Civ. Action No. 15-cv-03747-JD, N.D. Calif. Justice KS Puttaswamy (Retd.) & Anr. v. Union of India & Ors Patrick Breyer v. Bundesrepublik Deutschland., No. C-582/14, at ¶49 (Second Chamber Oct. 19, 2016) TC v. Vizio, Inc., No. 2:17-cv-00758, slip op. (D. N.J. filed Feb. 6, 2017)

Statutes 15 U.S.C. Sec. 45 15 U.S.C. Sec. 7701 et seq. 740 ILCS 14/10 740 ILCS 14/15 Calif. Bus. & Prof. Code § 22575 Calif. Bus. & Prof. Code § 22575(a) Calif. Bus. & Prof. Code § 22577(c) Calif. Bus. & Prof. Code §§ 22580–2 Calif. Bus. & Prof. Code Sec. 22580(d) Calif. Bus. & Prof. Code Sec. 22580(e) Calif. Bus. & Prof. Code § 22580(f) Calif. Civil Code § 1798.83 Calif. Civil Code § 1798.83(e)(2) California Consumer Legal Remedies Act (CLRA), California Civil Code Sec. 1750 et seq. Canada Personal Information Protection and Electronic Documents Act (PIPEDA), S.C. 2000, c. 5 (2015) Children’s Online Privacy Protection Act (COPPA), 15 U.S.C. §§ 6501–6 15 U.S.C. Sec. 6501(2)(A) 15 U.S.C. Sec. 6501(6) 15 U.S.C. Sec. 6501(9) Commercial Privacy Bill of Rights Act of 2011, S. 799, 112nd Cong. § 202 (2011) Controlling the Assault of Non-Solicited Pornography and Marketing (CAN-SPAM) Act of 2003, Pub. L. No 180–7 Directive 95/46/EC of the European Parliament and of the Council of 24 October 1995 on the protection of individuals with regard to the processing of personal data and on the free movement of such data (Data Protection Directive) Article 2(a) Article 2(h) Article 8 Article 8(1) Article 8(2)(a) Recital 26 Directive 2002/58/EC of the European Parliament and of the Council of 12 July 2002 concerning the processing of personal data and the protection of privacy in the electronic communications sector (“e-Privacy Directive”) Article 1(2) Article 5(3) Article 9 Recital 10 Recital 24 Fair and Accurate Credit Transactions Act (FACTA), 15 U.S.C. §1601 et. seq., 1681 et. seq. Family Educational Rights and Privacy Act (FERPA), 20 U.S.C. § 1232(g)

BARFIELD PRINT.indd 511

13/11/2018 11:53

512  Research handbook on the law of virtual and augmented reality Gramm-Leach-Bliley Act (GLBA), 15 U.S.C. Sections 6801–9 Health Insurance Portability and Accountability Act of 1996 (HIPAA), 45 C.F.R. 160–4 Illinois Biometric Information Privacy Act, 740 Ill. Comp. Stat. 14/1 et seq. Location Privacy Protection Act of 2015, s. 2270 Massachusetts M.G.L. chapter 93H Proposal for a Regulation of the European Parliament and of the Council concerning the respect for private life and the protection of personal data in electronic communications and repealing Directive 2002/58/EC (Regulation on Privacy and Electronic Communications) Regulation (EU) 2016/679 of the European Parliament and of the Council of 27 April 2016 on the protection of natural persons with regard to the processing of personal data and on the free movement of such data, and repealing Directive 95/46/EC (General Data Protection Regulation, “GDPR”) Article 3(2) Article 4(1), 4(4), 4(11), 4(14) Article 5(1)(a) Article 7(1), 7(2), 7(3) Article 9(1), (2), (4) Article 9(2)(a) Article 12(1) Article 12(2)(f) Article 13(2)(f) Articles 13–18 and 20–1 Article 22, 22(3), 22(4) Article 27(1) Article 35(1), (3)(b), (7) Article 36 Article 83(4)–(5) Recital 24 Recital 32 Recital 51 Recital 58 Recital 60 Recital 71 Recital 72 Telephone Consumer Protection Act (TCPA) of 1991, 47 U.S.C. Sec. 227 Texas Statute on the Capture or Use of Biometric Identifier, Tex. Bus. & Com. Code Ann. Sec. 503.001 Unfair Competition Law, California Business and Professions Code Sec. 17200 et seq. United Kingdom Data Protection Act and the Australian Privacy Principals Uniting and Strengthening America by Providing Appropriate Tools Required to Intercept and Obstruct Terrorism (USA PATRIOT) Act of 2001, Pub. L. No 107-56. Video Privacy Protection Act (VPPA) Washington House Bill 1493

Regulations 16 C.F.R. 312-2 Standards for the Protection of Personal Information of Residents of the Commonwealth, 201 C.M.R. 17

BARFIELD PRINT.indd 512

13/11/2018 11:53

16.  Reordering the chaos of the virtual arena: harmonizing law and framing collective bargaining for avatar actors and digital athletes Jon M. Garon*

Serious sport has nothing to do with fair play. It is bound up with hatred, jealousy, boastfulness, disregard of all rules and sadistic pleasure in witnessing violence: in other words it is war minus the shooting. (George Orwell, Shooting an Elephant) The future ain’t what it used to be. (Yogi Berra)

LET THE GAMES BEGIN Free agency started it all. Before it was over, entertainment had changed forever. Like the professional athletes a generation earlier, the elite computer gamers playing Madden Football wanted to move players from team to team.1 As with fantasy sports leagues, the gamers wanted to create rosters that matched up from week to week. A few wanted cross-over players to join in. Soon an online game developed. Clever programmers, illustrators, and sports aficionados created “PRO” (www.pro.us), an interactive online gaming environment featuring virtual replicas of all starting professional athletes in football, baseball, basketball, soccer, hockey, tennis, boxing (full contact), and golf. Professional women’s leagues were represented along with the men’s leagues. Players could be used in their professional sport or imported to any of the other sports represented in the game. Lacrosse and gladiator fighting competitions were quickly added and proved some of the most popular. In some games, gamers would control entire teams; in others a gamer would control only his or her own character, allowing dozens of online gamers to participate at once. (This was difficult in football, but worked very well for gladiatorial fights in the Hippodrome and Roman Coliseum.) As a default design, players generally

*  An earlier version of this chapter was published as part of the 10th Annual Chapman Law Review Symposium as Jon M. Garon, Playing in the Virtual Arena: Avatars, Publicity, and Identity Reconceptualized through Virtual Worlds and Computer Games, 11 Chap. L. Rev. 465 (Spring 2008). I would like to thank my fellow panelists of that event for the insights they shared throughout the symposium: Kathy Heller, John Tehranian, Kevin Greene, Eric Farber, and the constitutional law panelists Erwin Chemerinsky, Jay Dougherty, John Eastman, and Raymond Ku. 1   See Madden 18: Introducing Mut Squads, Electronic Arts Inc: Sports, www.easports.com/ madden-nfl (“Bring your best players to the field when you team up with friends to build your Ultimate Team of NFL stars and compete online together for the win”).

513

BARFIELD PRINT.indd 513

13/11/2018 11:53

514  Research handbook on the law of virtual and augmented reality wore their own team numbers and uniforms, but gamers could reoutfit the characters with any of the available professional, minor league, or university team indicia available in the “locker room.” Celebrity Death Match became a popular minigame. Even in the open sourced, fan fiction version of PRO, some rules had to be developed. Nazi uniforms and swastikas were banned, as was all material considered blasphemous by the community. Some leagues permitted Pokémon and Yu-Gi-Oh! creatures while others insisted all avatars be based on humans. Gamers were charged a monthly subscription fee, with additional costs for purchasing uniforms in the locker room. More entrepreneurial gamers set up design studios, selling original team logo designs, uniforms, equipment, and weapons to enhance the play of the characters. Characters became avatars.2 A gamer soon became able to edit the selected athlete, changing not only the team uniform but also the physical attributions of the player. Gamers increased the skill level for the characters and avatars on their team rosters through play or purchase of digital steroids and virtual growth hormones. Using augmented reality glasses, meet-ups were held in parks across the globe. Like Ultimate Frisbee, the games had no referees. But they did have instant replay, actively running on each player’s mobile device. Nobody paid the athletes, their unions, or the professional sports leagues. Pictures, descriptions, and accounts of games—both copies and originals—were completely unauthorized. In the final innovation before the lawsuits began, statistics from weekly fantasy sports play were added. Each week, the game “replayed” a highlight reel featuring play of fantasy teams against each other in fantasy leagues. With the nearly photographic quality of the animation, these fifteen minute highlight shows became as popular as some network broadcasts, appearing on YouTube and other websites. PRO’s popularity rivaled the Super Bowl. Millions of casual gamers held accounts and regularly spent time acting out their professional games; tens of thousands were heavy users. Advertisers vied to purchase billboard space and naming rights to the virtual arenas and stadiums. Only the leagues, unions, players, networks, and videogame publishers objected.

1. INTRODUCTION The development of law within and about online games, virtual worlds, and augmented reality will remain a complex and conceptually challenging exercise for years to come. In many respects, the commercial and social interactions within these environments are essentially the same as those interactions conducted face-to-face or over less engrossing technologies. In certain respects, however, the immersive nature of these virtual worlds redefines the nature of the experience. 2   See infra, note 5 and accompanying text. The description used above distinguishes between avatars, which are individuated representations of the computer user, and characters, which may be substantially the same identity for every computer user interacting with that aspect of the game. Like Mickey Mouse, Mario and Kirby are characters in a game rather than avatars representing the person operating the controls.

BARFIELD PRINT.indd 514

13/11/2018 11:53

Reordering the chaos of the virtual arena  515 Because virtual worlds mimic their brick and mortar counterparts, they exhibit commercial attributes unlike those of plays, television shows, or motion pictures. At the same time, the many-to-many communications channels create additional complexity because there may not be a single individual or entity responsible for all the content.3 To the extent that there is commerce conducted within a medium, the historic separation between commercial conduct and expressive speech can still be used, but its role must be updated. In the first instance, such legal line drawing will necessarily be done with crude tools, so this chapter suggests that, just as the theatre and motion picture industries turned to collective bargaining agreements to provide a more refined set of rules for professional content development, so the entertainment content created in virtual worlds will benefit from similar collective bargaining solutions to legally difficult conundrums. Section 2 of this chapter provides an overview of virtual worlds and the legal framework for the regulation of content ownership. Section 3 addresses the tension between the speech and property rights associated with the participants in this rapidly evolving art form, identifying what the law suggests and how it should evolve through case law and legislation. Section 4 suggests the steps that can be taken through private ordering collective bargaining arrangements to further clarify the protections for professionals associated with this developing new medium.

2. CONTENT OWNERSHIP INSIDE COMPUTER GAMES AND VIRTUAL WORLDS 2.1 Origins Since the advent of Pong,4 computer gamers have been searching for increasingly ­realistic—or at least photorealistic—experiences with their computer generated content. As technology has allowed for ever more realistic images, sound, and even tactile response, the gaming experience has grown far beyond high-score lists to encompass immersive, interactive environments. The “virtual environment is an interactive computer simulation which lets its participants see, hear, use, and even modify the simulated objects in the computer-generated environment.”5 As described by the plaintiff in a legal action against Second Life, “many people ‘are now living large portions of their lives, forming friendships with others, building and acquiring virtual property, forming contracts, substantial 3   See Eric T. Gerson, Video Game Violence and the Technology of the Future, 76 Brook. L. Rev. 1121, 1142 (2011) (“given its potential for full-immersion gaming, there is no question that, once perfected, VR will become a staple of the gaming industry. The theoretical implications of widespread implementation are staggering”). 4   “Pong, while not the first videogame, was the first coin-op arcade game and the first mainstream videogame that was available to almost everyone. Pong was the impetus for the development of the videogaming industry, almost single-handedly creating both the home and the arcade videogame markets.” IGN.com, Pong, www.ign.com/wikis/history-of-video-game-consoles/Pong. 5   Woodrow Barfield, Intellectual Property Rights in Virtual Environments: Considering the Rights of Owners, Programmers and Virtual Avatars, 39 Akron L. Rev. 649 (2006) (footnote omitted).

BARFIELD PRINT.indd 515

13/11/2018 11:53

516  Research handbook on the law of virtual and augmented reality business relationships and forming social organizations’ in virtual worlds such as Second Life.”6 Attributes of the genre include the ability to involve a very large number of simultaneous participants online through the internet or other networking systems, and the ability for the participant to take on a character or role in the game. As some of these gaming environments involve millions of players, they are now known as “massively multi-player online role-playing games” (MMORPG).7 The user-created characters in these worlds are referred to as “avatars,” virtual representations of the players.8 The online roleplaying games trace their lineage to Dungeons & Dragons and the genre of tabletop roleplaying games in which participants would create characters who would navigate in the fictional worlds created and managed by dungeon masters.9 In Dungeons & Dragons, the players would give themselves the attributes of wizards, warriors, trolls, elves, and similar mythical characters. The interaction of the players and their characters would last in story arcs that could run for indefinite periods. There may even be some characters still in use from 1974, when the rules were first published.10 Character attributes were kept on note cards and interactions mediated through polyhedral dice. The genre moved to the online environment and added games such as World of Warcraft, Lord of the Rings Online: Shadows of Angmar, Warhammer Online, EVE Online, Anarchy Online, RuneScape, Rappelz, and Shadow of Legend.11 The genre also includes science fiction worlds such as The Matrix Online and Star Wars Galaxies.12 Professor Erez Reuveni suggests two attributes that define MMORPGs and virtual worlds, separating them from classic computer games or their tabletop predecessors.13 “Unlike traditional computer games . . . virtual worlds are persistent and exist independently of any individual’s presence. Virtual worlds exist in real time even after a specific player logs off, and a person’s actions can permanently shape the virtual world.”14 While this is a literary distinction, the persistence and literary independence may also provide a framework for treating certain aspects of these works as distinct from novels, computer games, or motion pictures.   Bragg v. Linden Research, Inc., 487 F. Supp. 2d 593, 595 (E.D. Pa. 2007).   See Barfield, supra note 5, at 650; Cory Ondrejka, Escaping the Gilded Cage: User Created Content and Building the Metaverse, 49 N.Y.L. Sch. L. Rev. 81 (2004); Caroline Bradely & A. Michael Froomkin, Virtual Worlds, Real Rules, 49 N.Y.L. Sch. L. Rev. 103, 121 (2004).  8   Bragg, 487 F. Supp. 2d at 595. The Bragg court also described avatars, explaining that “since the advent of computers, however, ‘avatar’ is also used to refer to an Internet user’s virtual representation of herself in a computer game, in an Internet chat room, or in other Internet fora.” Id at 595 n.3.  9   Allen Rausch, Magic & Memories: The Complete History of Dungeons & Dragons, GameSpy. com (Aug. 16, 2004), http://pc.gamespy.com/articles/538/538262p1.html. 10   Id. 11   All of these games can be found at www.gamespy.com (last visited July 9, 2017). 12   Traveller was the science fiction roleplaying counterpart to Dungeons & Dragons created by Marc Miller for Game Designers’ Workshop in 1977. See Traveller (role-playing game), Wikipedia, https://en.wikipedia.org/wiki/Traveller_(role-playing_game). 13   Erez Reuveni, On Virtual Worlds: Copyright and Contract Law at the Dawn of the Virtual Age, 82 Ind. L.J. 261 (2007). 14   Id at 265.  6  7

BARFIELD PRINT.indd 516

13/11/2018 11:53

Reordering the chaos of the virtual arena  517 Online roleplaying games’ persistence and literary independence have spawned an entirely new genre from the fantasy worlds. In Second Life, the environment is a fantasy alternative to modern reality, with businesses, lounges, universities, and other brick and mortar equivalents. These environments are “near worlds,” or alternate realities set in the present with only modest changes from the world around us: On the surface, Second Life is an online place somewhat similar to a Sims computer game, with buildings and roads and a population of avatars – cartoonish representations of users. Anyone can join, and users chat with each other, build houses, start businesses, go to concerts, and otherwise hang out together. The site is winning the attention of Internet heavy hitters . . . who believe Second Life—or something like it—will evolve into a way to use the Net that is more like the way humans use the real world. In that way, it could mark a next great leap in the Net’s accessibility.15

Although the MMORPG environments, with their fantastical gameplay, have captured the primary focus of the public in the past decade, both gaming and near world experiences will continue to be available. Largely as a result of the general deployment of virtual reality goggles, interest in virtual reality has reemerged. Linden Lab, the owners of Second Life, are launching “Sansar,” a platform for developers to create virtual reality spaces for use with virtual reality (“VR”) goggles.16 Social media, which has surpassed virtual reality for online social engagement may also be moving toward virtual reality. In 2017, “Facebook announced Facebook Spaces, which allows users to share photos, videos, drawings and more in a VR realm populated by cartoon avatars.”17 The growth of virtual worlds is coming from a tremendous curiosity about these opportunities and a dramatic shift in spending of the entertainment dollar: Video game revenue is expected to rise from $17.0 billion to $20.3 billion by 2020 (3.6 percent CAGR). Despite the high price, the quality of the experience on virtual reality devices (for games, interactive entertainment, and VR video) is expected to draw more consumers into trying them, setting the stage for the real sales drive in 2017 and 2018. In contrast, VR on mobile devices will likely go mainstream in 2016 with cheap headsets that can be slotted into a phone.18

This projected revenue for video games makes this segment comparable to radio and music industries.19 Calculated in this fashion, the videogame industry is discrete from Internet access, valued at $181.7 billion, with internet advertising projected to climb to

15   Kevin Maney, The King of Alter Egos Is Surprisingly Humble Guy, Creator of Second Life’s Goal? Just to Reach People, USA Today Feb. 5, 2007, at B1. 16   Ben Lang, “Sansar” Will Open to All in First Half of 2017 with a New Approach to Virtual Worlds, Road To VR, www.roadtovr.com/sansar-release-date-preview-virtual-reality-liden-lab/ (last visited July 9, 2017). 17   Edward C. Baig, Facebook Spaces Is Where Real and Virtual Worlds Collide, USA Today (May 1, 2017), www.usatoday.com/story/tech/columnist/baig/2017/05/01/facebook-spaces-social-virual-rea​ lity-friends-review/101019016/. 18  PwC, PwC’s Entertainment & Media Outlook Forecasts U.S. Industry Spending to Reach $720 Billion by 2020, PR Newswire (June 8, 2016), www.prnewswire.com/news-releases/pwcs-­entertainm​ ent--media-outlook-forecasts-us-industry-spending-to-reach-720-billion-by-2020-300281287.html. 19   See id (“Radio is expected to rise from $21.4 billion to $23.1 billion in 2020 . . . Music revenue is expected to increase from $15.2 billion to $18.0 billion in 2020”).

BARFIELD PRINT.indd 517

13/11/2018 11:53

518  Research handbook on the law of virtual and augmented reality $93.5 billion by 2020.20 If VR devices drop in price and increase in popularity as anticipated, the virtual world and MMORG environment could potentially upend traditional media and education. The benefit of persistence, literary independence, and interactivity results in an ongoing user participation, which requires consumers to pay to remain subscribed. Moreover, the virtual environments create opportunity for advertisers to place goods and services in front of consumers in a variety of methods—which will fuel advertiser spending and consumer behavior. Linked through social media platforms, the public will be engaged in experiencing these constantly changing virtual worlds and interacting with friends and fellow players in a self-perpetuating, potentially addictive cycle. In addition to offering a highly interactive entertainment experience, the persistence and literary independence of the medium provides an ideal environment in which to place advertising and products.21 Even in the medium of an interactive, virtual world, the distribution of producercreated content will continue to coexist with that of participant-created content. The literary independence does not require that content come only from participants, so it is likely that producer-created content will play a significant role in many of the games and worlds popular within the genre.22 As a variation of traditional theatre, the coexistence might manifest as the ability for both professional and amateur actors to utilize avatars to perform scripted productions. The entire canon of Shakespeare’s plays will eventually be available performed in animation—with avatar performances by amateur and professional acting companies. Moreover, the choice of avatar will undoubtedly involve the public’s affinity for the stars of the moment. What better for an amateur actor than to portray the avatar of a famous performer? The aspiring professional actress will be able to shine with a nod to the femme fatales of the past.23 In the realm of sports, a simple version of this opportunity regularly takes place. Through the use of game controllers, players control professional athletes competing in all major professional sports. Madden Football, a game licensed by the National Football League and Players, Inc.—the for-profit arm of the National Football League Players Association24—is immensely popular.25 The football play calling, player control and   Id.   The growth of the internet, social media, virtual worlds, and gaming have not necessarily displaced traditional television and video, but those media are changing as well. See id (“TV and video is expected to rise from $121.4 billion to $124.2 billion in 2020 . . . The continued growth of video on demand (VOD) and over-the-top (OTT) services is putting pressure on the ‘theatrical window’ period traditionally enjoyed by cinemas. In fact, electronic home video sales ($11.2 billion . . .) eclipsed box office sales ($10.3 billion . . . in 2015”). 22   Because producer-created content can be used to set professional norms through collective bargaining agreements, the presence of such content may be critical to the development of the medium. See infra section 3. 23   I imagine as one such production a restaging of King Lear played by James Earl Jones with a Darth Vader avatar and avatars from within the Star Wars cavalcade. Jar Jar Binks or Yoda can play the Fool and General (Princess) Leia Organa serves well as Cordelia. Rey and Jyn Erso may now join Queen Padmé Amidala to compete for the casting of Regan and Goneril. 24   Players, Inc., NFL Players Inc., www.nflpa.com/players. 25   Video Game Sales Wiki, http://vgsales.wikia.com/wiki/Madden_NFL (“[M]adden NFL is EA Sports’ biggest franchise in the United States. The game has consistently been a best seller and is often played by NFL players and celebrities”). 20 21

BARFIELD PRINT.indd 518

13/11/2018 11:53

Reordering the chaos of the virtual arena  519 immersive graphics make Madden Football only slightly less immersive than the virtual worlds or online roleplaying games.26 From an economic and social standpoint, both law and industry must help determine whether the professional athletes and professional actors are entitled to compensation for being utilized as part of this entertainment product. Should an author have an unbridled right to excerpt celebrities’ identities from real life when creating upon a digital, interactive canvas? The athletes certainly care. “Many professional athletes have grown up playing sports video games. Players have even lobbied game companies to improve their digital representations.”27 Such economic interests create powerful incentives to maximize ownership of the content within the computer games and virtual worlds as well as protect the limitations on such content ownership. 2.2  First Amendment Protection for Computer Games and Virtual Worlds The question of content ownership within computer games and virtual worlds assumes that the games and worlds are authorial, creative content, benefiting from the full panoply of legal protections. It has long been held that computer programs are literary works, protected by copyright laws.28 Beyond copyright, the law must reflect the changed paradigm of computer games and virtual worlds. While the makers of Pong were espousing no point of view with their square ball or simple paddles, modern games can be violent, sexy, propagandistic, satirical or politically astute.29 At some point, the authors of these games tapped them for their expressive ability. The law has slowly followed suit. The tardiness of the law to recognize the legitimate expressive interests of entertainment media is not a new concern. Since at least 1915, the question of the applicability of First Amendment rights has been reassessed by the courts with the introduction of each

26   See generally Christopher Paul and Jeffrey Philpott, The Rise and Fall of “Cardboard Tube Samurai,” in Online Gaming in Context: The Social and Cultural Significance of Online Games, 185–6 (Garry Crawford et al eds, Routledge 2011). 27  Schiesel, With Famed Players, Game Takes on Madden’s Turf, N.Y. Times, Sept. 17, 2007, at C3. 28   See Williams Elecs., Inc. v. Artic Int’l, Inc., 685 F.2d 870 (3d Cir. 1982). The House Report on the 1976 Act contains the following statement: “The term ‘literary works’ does not connote any criterion of literary merit or qualitative value: it includes catalogs, directories, and similar factual, reference, or instructional works and compilations of data. It also includes computer data bases and computer programs to the extent that they incorporate authorship in the programmer’s expression of original ideas, as distinguished from the ideas themselves.” Id at 875 n.4 quoting H.R. Rep. No 94-1476, 94th Cong., 2d Sess. 54 (1976). Cf. Midway Mfg. Co. v. Dirkschneider, 543 F. Supp. 466, 479 (D. Neb.1981) (copyright infringement involving Pac-Man, Galatian, and Rally-X); Apple Comput., Inc. v. Franklin Comput. Corp., 714 F.2d 1240 (3d Cir. 1983); Arthur Miller, Copyright Protection for Computer Programs, Databases, and Computer Generated Works: Is Anything New Since CONTU? 106 Harv. L. Rev. 977, 979 (1993); Pamela Samuelson, CONTU Revisited: The Case against Copyright Protection for Computer Programs in Machine-Readable Form, Duke L.J. 663 (1984). 29   See e.g., Chris Morris, Your Tax Dollars at Play: U.S. Army Gets into the Gaming Business. You’re Paying for It, CNN Money (June 3, 2002, 2:17 PM), http://money.cnn.com/2002/05/31/ commentary/game_over/column_gaming (describing computer game recruiting tools).

BARFIELD PRINT.indd 519

13/11/2018 11:53

520  Research handbook on the law of virtual and augmented reality medium. In Mutual Film Corp. v. Industrial Commission of Ohio (“The Mutual”)30 a filmmaker facing a censor board challenged the state law claiming violations of free speech rights,31 as well as interference with interstate commerce.32 In this case,33 free speech claims were brought on both the state level,34 and the federal one.35 Because the Supreme Court did not apply the First Amendment to the states until 1925,36 the federal claim was exceptionally weak and technically beyond the power of the Court.37 Looking at the state speech protection, the Court endeavored to assess the extent to which silent movies were protected as “speech, writing, or printing.” Are moving pictures within the principle, as it is contended they are? They, indeed, may be mediums of thought, but so are many things. So is the theatre, the circus, and all other shows and spectacles, and their performances may be thus brought by the like reasoning under the same immunity from repression or supervision as the public press,—made the same agencies of civil liberty . . .   We immediately feel that the argument is wrong or strained which extends the guaranties of free opinion and speech to the multitudinous shows which are advertised on the bill-boards of our cities and towns . . . and which seeks to bring motion pictures and other spectacles into practical and legal similitude to a free press and liberty of opinion. The judicial sense supporting the common sense of the country is against the contention.38

The Court took more than three decades to reverse this position regarding such “spectacles” and other entertainment.39 In 1948, the Court articulated a new philosophy regarding the role of the First Amendment: We do not accede to appellee’s suggestion that the constitutional protection for a free press applies only to the exposition of ideas. The line between the informing and the entertaining is too

  236 U.S. 230 (1915).   Id at 239–40. The state law provision in question read as follows: “Only such films as are in the judgment and discretion of the board of censors of a moral, educational or amusing and harmless character shall be passed and approved by such board.” Id at 240. 32   Jon M. Garon, Entertainment Law, 76 Tul. L. Rev. 559, 637–8 (2002) (“The interstate commerce argument was, in many ways, the more practical of the two arguments. Films were admittedly shipped in interstate commerce; the exhibitors often owned or were members of consortia that crossed state lines. Interstate commerce was a justiciable issue for the federal courts.”). 33   See John Wertheimer, Mutual Film Reviewed: The Movies, Censorship, and Free Speech in Progressive America, 37 Am. J. Legal Hist. 158, 159 (1993). (The term “The Mutual” represented both the company and case.) 34   The Ohio Constitution also has a strong free speech clause: “Every citizen may freely speak, write, and publish his sentiments on all subjects, being responsible for the abuse of the right; and no law shall be passed to restrain or abridge the liberty of speech, or of the press. In all criminal prosecutions for libel, the truth may be given in evidence to the jury, and if it shall appear to the jury that the matter charged as libelous is true, and was published with good motives, and for justifiable ends, the party shall be acquitted.” Ohio Const. art. I, § 11. 35   Mut. Film Corp. v. Indus. Comm’n of Ohio, 236 U.S. 230 (1915). 36   Gitlow v. New York, 268 U.S. 652 (1925). 37   Mut. Film Corp., 236 U.S. at 243. 38   Id at 243–4; see also Donald Lively, Modern Communications Law 10 (1991). 39   Joseph Burstyn, Inc. v. Wilson, Comm’r of Educ. of N.Y., 343 U.S. 495, 502 (1952); Lively, supra note 38, at 10. 30 31

BARFIELD PRINT.indd 520

13/11/2018 11:53

Reordering the chaos of the virtual arena  521 elusive for the protection of that basic right. Everyone is familiar with instances of propaganda through fiction. What is one man’s amusement, teaches another’s doctrine.40

Relying on this expansive language, the Court eroded limitations on the First Amendment based on medium and in 1952, the Court reversed the Mutual decision and protected motion pictures despite their largely commercial nature and their proclivity for greater evil—both accusations hurled today at computer games and virtual worlds: It is urged that motion pictures do not fall within the First Amendment’s aegis because their production, distribution, and exhibition is a large-scale business conducted for private profit. We cannot agree. That books, newspapers, and magazines are published and sold for profit does not prevent them from being a form of expression whose liberty is safeguarded by the First Amendment. We fail to see why operation for profit should have any different effect in the case of motion pictures.   It is further urged that motion pictures possess a greater capacity for evil, particularly among the youth of a community, than other modes of expression. Even if one were to accept this hypothesis, it does not follow that motion pictures should be disqualified from First Amendment protection. If there be capacity for evil it may be relevant in determining the permissible scope of community control, but it does not authorize substantially unbridled censorship such as we have here.41

Despite the Court’s explicit rejection of the regulatory rationales in the context of motion pictures, modern legislatures attempted to regulate video games again this century.42 The municipal bodies and state legislatures attempted to ban or restrict certain aspects of video games, suggesting that the nature of the video game industry makes the content unworthy of constitutional protection or that these diversions’ attractiveness to youth makes censorship more appropriate.43 They return to the same themes espoused in The Mutual and Joseph Burstyn: “Several decades ago, the Supreme Court ‘recognized that some believe motion pictures possess a greater capacity for evil, particularly among the youth of a community, than other modes of expression.’”44 The US Supreme Court answered the question in Brown v. Entertainment Merchants Ass’n,45 providing that even ultraviolent video games sold to minors were included in broad First Amendment protection.46 “Like the protected books, plays, and movies that   Winters v. New York, 333 U.S. 507, 510 (1948).   Burstyn, 343 U.S. at 501–2 (citing Grosjean v. Am. Press Co., 297 U.S. 233 (1936); Thomas v. Collins, 323 U.S. 516, 531 (1945)) (footnotes omitted). 42   Video Software Dealers Ass’n v. Schwarzenegger, No C-05-04188 (RMW), 2007 U.S. Dist. LEXIS 57472 at 25 (N.D. Cal. Aug. 6, 2007). 43   See, e.g., La. Stat. Ann. § 14:91.14 (June 15, 2006) (criminalizing video games that appeal to a minor’s morbid interest in violence); St. Louis County, Mo., Ordinance 20,193 (Oct. 26, 2000); Indianapolis and Marion County, Ind., Ordinance 72 (July 10, 2000). 44   Video Software Dealers Ass’n v. Schwarzenegger, 2007 U.S. Dist. LEXIS 57472 (D. Cal. 2007) at 9 n.2 quoting Interstate Circuit, Inc. v. City of Dallas, 390 U.S. 676, 690 (1968). 45   131 S. Ct. 2729 (2011). 46   Id at 2734–5. Prior to the Supreme Court decision, there were a number of lower court rulings, including Interactive Dig. Software Ass’n v. St. Louis Cty., 329 F.3d 954 (8th Cir. 2003); Am. Amusement Mach. Ass’n v. Kendrick, 244 F.3d 572 (7th Cir. 2001); Video Software Dealers Ass’n v. Schwarzenegger, 2007 U.S. Dist. LEXIS 57472 (D. Cal. 2007); Entm’t Software Ass’n v. Foti, 451 F. Supp. 2d 823 (M.D. La. 2006); Entm’t Software Ass’n v. Hatch, 443 F. Supp. 2d 1065 40 41

BARFIELD PRINT.indd 521

13/11/2018 11:53

522  Research handbook on the law of virtual and augmented reality preceded them, video games communicate ideas—and even social messages—through many familiar literary devices (such as characters, dialogue, plot, and music) and through features distinctive to the medium (such as the player’s interaction with the virtual world). That suffices to confer First Amendment protection.”47 The Supreme Court used the Brown decision to put an exclamation point on its prior decision “that new categories of unprotected speech may not be added to the list by a legislature that concludes certain speech is too harmful to be tolerated.”48 The Brown decision was a bold statement regarding the future of the First Amendment. The Supreme Court rejected the concept that ultraviolent content could be treated as a new category of unprotected speech.49 The Supreme Court also refused to entertain the possibility that content could impact minors and adults in different ways.50 The Supreme Court joined a series of lower courts that have been steadfast in their refusal to allow prohibitions or criminal sanctions for the sale of ultraviolent or sexually explicit video games, even to minors.51 In correctly refusing to make such categorical exceptions, however, the decisions may be going too far in the opposite direction, cloaking the speech aspects of the interactive games with too broad a penumbra. The Brown decision states that “California’s argument would fare better if there were a longstanding tradition in this country of specially restricting children’s access to depictions of violence, but there is none.”52 Despite this judicial assertion, the attempt to create additional categories of speech that can be regulated with regard to minors is certainly not new. Broadcast regulation has added mandatory children’s programming,53 technological blocking measures,54 viewer ratings,55 and internet privacy protections.56 So

(D.  Minn. 2006); Entm’t Software Ass’n v. Granholm, 426 F. Supp. 2d 646 (E.D. Mich. 2006); Entm’t Software Ass’n v. Blagojevich, 404 F. Supp. 2d 1051 (E.D. Ill. 2005); and Video Software Dealers Ass’n v. Maleng, 325 F. Supp. 2d 1180 (W.D. Wash. 2004). 47   Brown, 131 S. Ct. at 2733. 48   United States v. Stevens, 559 U.S. 460, ___, 130 S. Ct. 1577, 1582–1583, 176 L.Ed.2d 435 (2010). 49   Id at 2735. 50   Id (“Because speech about violence is not obscene, it is of no consequence that California’s statute mimics the New York statute regulating obscenity-for-minors that we upheld in Ginsberg v. New York, 390 U.S. 629, 88 S.Ct. 1274, 20 L. Ed. 2d 195 (1968)”). But see FCC v. Pacifica Found., 438 U.S. 726, 749–50 (1978) (upholding indecent broadcasts because of the impact on minors who may be listening); Pub. L. No 101-437, 104 Stat. 996 (codified in scattered sections of 47 U.S.C.). See Cass R. Sunstein, Television and the Public Interest, 88 Calif. L. Rev. 499 (2000). 51   See Christian Genetski, Brown v. Ema/Esa: U.S. Supreme Court Stops California From Playing Games with The First Amendment, 15 SMU Sci. & Tech. L. Rev. 135, 137, 143 (2012) (“Leading up to the Brown case, the industry was 13-0 in fighting back challenges; none of these laws took effect because all were defeated in the courts . . . Including the first thirteen unsuccessful cases involving state regulation of the video game industry, some states have spent over $3 million in pursuit of these efforts.”). 52   Brown v. Entm’t Merch. Ass’n, 131 S. Ct. 2729, 2736 (2011). 53  Genetski, supra note 51 at 137. 54  47 U.S.C. 303(a) (2007). 55   Telecommunications Act of 1996, Pub. L. No. 104-104, 110 Stat. 56 (1996); 47 U.S.C. 303(w) (2007). 56   Children’s Online Privacy Protection Act of 1998, 15 U.S.C. §§ 6501–3 (2007).

BARFIELD PRINT.indd 522

13/11/2018 11:53

Reordering the chaos of the virtual arena  523 much for “no law”—Congress has been quite active in balancing the interests of speakers and their minor audiences.57 Of course, as with the attempts to regulate sexually explicit content, any such laws must avoid creating an unlawful burden on the production of protected speech for adults when building these protections for minors.58 But, given the wide-ranging nature of the content being developed (and described presently), content creators should recognize that the battle to protect the public from the most excessive offensive content will continue, despite the absolutist approach of the Brown court.59 As virtual-world and augmented reality technology merge the real world with the graphical and literal one, new questions will emerge regarding the secondary effects of the content that will again raise questions of regulation.60 Secondary effects of zoning are permitted to restrict speech while secondary effects on the audience are not.61 Immersive technologies may force legislators and courts to address this dichotomy once again. 2.3 Copyright’s First Amendment Accommodations through the Idea/Expression Dichotomy and Fair Use A second balancing occurs between First Amendment rights of the publisher and the intellectual property rights of any third parties who may have had some of their work exploited without permission. Copyright holders, trademark owners, athletes, celebrities, and performers holding publicity rights all seek to limit what publishers attempt to do with their proprietary content. The balancing of interests for First Amendment protection has very different 57   But see Ashcroft v. ACLU, 542 U.S. 656 (U.S. 2004) (granting preliminary injunction against Children’s Online Protection Act of 1998); ACLU v. Gonzales, 478 F. Supp. 2d 775 (D. Pa. 2007) (permanent injunction of same). 58   See Ashcroft v. ACLU, 542 U.S. 656, 677 (“the term ‘less restrictive alternative’ is a comparative term. An ‘alternative’ is ‘less restrictive’ only if it will work less First Amendment harm than the statute itself, while at the same time similarly furthering the ‘compelling’ interest that prompted Congress to enact the statute.”). 59   See, e.g., Ashutosh Bhagwat, In Defense of Content Regulation, 102 Iowa L. Rev. 1427, 1430 (2017) (“it is time to rethink our hostility to all forms of content regulation, and to consider whether a more nuanced approach is required”). 60   See David L. Hudson, Jr., The Secondary Effects Doctrine: “The Evisceration of First Amendment Freedoms,” 37 Washburn L.J. 55 (1997); Bhagwat, supra note 59 at 1442. 61   See Young v. Am. Mini Theatres, Inc., 427 U.S. 50 (1976) (“the zoning ordinances requiring that adult motion picture theaters not be located within 1,000 feet of two other regulated uses does not violate the Equal Protection Clause of the Fourteenth Amendment”); City of Renton v. Playtime Theatres, Inc., 475 U.S. 41 (1986) (“the Renton ordinance represents a valid governmental response to the ‘admittedly serious problems’ created by adult theaters”); City of Los Angeles v. Alameda Books, Inc., 535 U.S. 425 (2002) (“The Constitution does not prevent those communities that wish to do so from regulating, or indeed entirely suppressing, the business of pandering sex”: Scalia, concurring); Cohen v. California, 403 U.S. 15, 26 (1971) (“Fuck the Draft” permitted despite the breach of decorum and discomfort felt by those in view of the clothing); Snyder v. Phelps, 562 U.S. 443 (2011) (protestors allowed to march at a funeral despite emotional discomfort caused to the family of deceased). But see Madsen v. Women’s Health Ctr., Inc., 512 U.S. 753 (1994) (permitting a buffer zone between protestors and the entrance to an abortion clinic).

BARFIELD PRINT.indd 523

13/11/2018 11:53

524  Research handbook on the law of virtual and augmented reality a­ ssumptions depending on the nature of the rights at stake. Since computer games and virtual worlds involve copyright, trademarks, and publicity rights emphasized in this chapter, a brief survey of each helps to illustrate the variety of tests used to balance these interests. In addressing the limitations placed on a copyright holder’s interest by the First Amendment, the Supreme Court had historically focused on the “idea/expression dichotomy,”62 protecting the public’s right to use the idea expressed but not the particular expression owned by the copyright holder.63 “[T]he democratic dialogue—a self-governing people’s participation in the marketplace of ideas—is adequately served if the public has access to an author’s ideas, and such loss to the dialogue as results from inaccessibility to an author’s “expression” is counterbalanced by the greater public interest in the copyright system.”64 In Harper & Row Publishers, Inc. v. Nation Enterprises,65 the Supreme Court endorsed this general approach. “[C]opyright’s idea/expression dichotomy ‘[strikes] a definitional balance between the First Amendment and the Copyright Act by permitting free communication of facts while still protecting an author’s expression.’”66 “Due to this distinction, every idea, theory, and fact in a copyrighted work becomes instantly available for public exploitation at the moment of publication.”67 Dr Michael D. Birnhack, writing in the Nimmer copyright treatise, raised questions regarding the legitimacy of overreliance on the idea/expression dichotomy: Consider the photographs from the Vietnam War of the My Lai massacre. Here is an instance where the visual impact of a graphic work made a unique contribution to an enlightened democratic dialogue. No amount of words describing the “idea” of the massacre could substitute for the public insight gained through the photographs. The photographic expression, not merely the idea, became essential if the public was to fully understand what occurred in that tragic episode. It would be intolerable if the public’s comprehension of the full meaning of My Lai could be censored by the copyright owner of the photographs. Here it would seem that the speech interest outweighs the copyright interest.68

62   H.R. Rep. No 94-1476, at 57 (1976), as reprinted in 1976 U.S.C.C.A.N. 5659, 5670 (“Section 102(b) in no way enlarges or contracts the scope of copyright protection under the present law. Its purpose is to restate, in the context of the new single Federal system of copyright, that the basic dichotomy between expression and idea remains unchanged”); Comput. Assocs. Int’l Inc. v. Altai, Inc., 982 F.2d 693, 703 (2d Cir. 1992) (“It is a fundamental principle of copyright law that a copyright does not protect an idea, but only the expression of the idea”). 63   See, e.g., 17 U.S.C. § 102(b). “In no case does copyright protection for an original work of authorship extend to any idea, procedure, process, system, method of operation, concept, principle, or discovery, regardless of the form in which it is described, explained, illustrated, or embodied in such work.” Id. 64  4-19E Nimmer on Copyright § 19E.03 [A] [2]; Suntrust Bank v. Houghton Mifflin Co., 268 F.3d 1257, 1264 (11th Cir. 2001). 65   471 U.S. 539 (1985). 66   Id at 556 quoting Circuit Ct. opinion below, Harper & Row, Publishers v. Nation Enters., 723 F.2d 195, 203 (1983). 67   Eldred v. Ashcroft, 537 U.S. 186, 219 (U.S. 2003). See Feist Publ’ns, Inc. v. Rural Tel. Serv. Co., 499 U.S. 340, 349–50 (1991). 68  4-19E Nimmer on Copyright § 19E.03 [A] [2].

BARFIELD PRINT.indd 524

13/11/2018 11:53

Reordering the chaos of the virtual arena  525 Despite the importance of visually powerful images to the public debate, the publishers of most photographs retain the ability to demand payment for use of the copyrighted work.69 Such an image only becomes iconic through public exposure and adoption within the public’s consciousness. Instead of using the First Amendment to create an additional, constitutional limitation on the rights of copyright holders, in attempts to ensure that the public’s access to socially or culturally important aspects of expressive works are not unduly restricted by the copyright owners,70 courts have increasingly turned to the doctrine of fair use.71 The Supreme Court has made this increased constitutional importance of fair use explicit in Eldred v. Ashcroft.72 There the Court raised the implicit protections of fair use to the level of constitutional importance. “[T]he ‘fair use’ defense allows the public to use not only facts and ideas contained in a copyrighted work, but also expression itself in certain circumstances . . . [T]he fair use defense affords considerable ‘latitude for scholarship and comment,’ and even for parody.”73 The Supreme Court has reaffirmed and expanded on this First Amendment ­accommodation for copyright.74 “The ‘traditional contours’ of copyright protection, i.e.,   Cf. Rosemont Enters., Inc. v. Random House, Inc., 366 F.2d 303 (2d Cir. 1966) (attempted use of copyright ownership to stop unauthorized biography of Howard Hughes not permitted); N.Y. Times Co. v. United States, 403 U.S. 713 (1971) (the “Pentagon Papers” case). In concurrence, Justice Brennen expressed the lack of any need to balance copyright interests. “The Government is not asserting an interest in the particular form of words chosen in the documents, but is seeking to suppress the ideas expressed therein.” Id at 727 note. But see Salinger v. Random House, Inc., 811 F.2d 90 (2d Cir. 1987) (infringement for publication of unpublished letters in biography resulting in less access to such unpublished resources); New Era Publ’ns Int’l v. Henry Holt & Co., 873 F.2d 576 (2d Cir. 1989) (same). In rejecting the lower court’s rationale for refusing an injunction, the Second Circuit commented that “[t]he public would not necessarily be deprived of an ‘interesting and valuable historical study,’ but only of an infringing one.” Id, quoting New Era Publ’ns Int’l, ApS v. Henry Holt and Co., 695 F. Supp. 1493, 1528 (S.D.N.Y. 1988). 70   See Michael D. Birnhack, Copyright Law and Free Speech after Eldred v. Ashcroft, 76 S. Cal. L. Rev. 1275 (2003). 71  17 U.S.C. §107. Notwithstanding the provisions of sections 106 and 106A, the fair use of a copyrighted work,   including such use by reproduction in copies or phonorecords or by any other means specified by that section, for purposes such as criticism, comment, news reporting, teaching (including multiple copies for classroom use), scholarship, or research, is not an infringement of copyright. In determining whether the use made of a work in any particular case is a fair use the factors to be considered shall include—   (1) the purpose and character of the use, including whether such use is of a commercial nature or is for nonprofit educational purposes;   (2) the nature of the copyrighted work;   (3) the amount and substantiality of the portion used in relation to the copyrighted work as a whole; and   (4) the effect of the use upon the potential market for or value of the copyrighted work.   The fact that a work is unpublished shall not itself bar a finding of fair use if such finding is made upon consideration of all the above factors. Id. 69

  537 U.S. 186 (2003).   Id at 219–20, quoting 17 U.S.C. § 107 (2000), Harper & Row Publishers, Inc. v. Nation Enters., 471 U.S. 539, 560 (1985) (scholarship and comment); citing Campbell v. Acuff-Rose Music, Inc., 510 U.S. 569 (1994) (parody). 74   Golan v. Holder, 565 U.S. 302 (2012). 72 73

BARFIELD PRINT.indd 525

13/11/2018 11:53

526  Research handbook on the law of virtual and augmented reality the ‘idea/expression dichotomy’ and the ‘fair use’ defense, moreover, serve as ‘built-in First Amendment accommodations.’”75 Although fair use has been statutorily codified into four prongs, courts have been quite adept at reviewing the significance of the usage and the lack of legitimate access to the work as part of the analysis. For example, in refusing to enjoin the unauthorized sequel to Gone with the Wind, the Eleventh Circuit looked to the harm caused to the well-established classic novel against the public’s interest in a unique approach to the well-established novel.76 The case closest to the My Lai massacre example comes from the famous Zapruder film which serendipitously captured the assassination of John F. Kennedy.77 The owner of the copyright refused to license the images for use in the defendant’s book. In response, the defendant used his access to the photographs to steal copies of the images for use in his own work. The district court acknowledged the misconduct of the defendant but focused on the elastic nature of the fair use doctrine, drawing the following conclusion: There is a public interest in having the fullest information available on the murder of President Kennedy. Thompson did serious work on the subject and has a theory entitled to public consideration . . ..[T]he Book is not bought because it contained the Zapruder pictures; the Book is bought because of the theory of Thompson and its explanation, supported by Zapruder pictures.   There seems little, if any, injury to plaintiff, the copyright owner. There is no competition between plaintiff and defendants. Plaintiff does not sell the Zapruder pictures as such and no market for the copyrighted work appears to be affected. Defendants do not publish a magazine. There are projects for use by plaintiff of the film in the future as a motion picture or in books, but the effect of the use of certain frames in the Book on such projects is speculative. It seems more reasonable to speculate that the Book would, if anything, enhance the value of the copyrighted work; it is difficult to see any decrease in its value.78

Such an analysis provides an excellent solution to the problem posed by the My Lai photographs. In the situation where a license is available for purchase, then the presumption is that a publisher seeking to use the work should secure that license. Where the content is not available and used in a transformative manner,79 such as to propound a new idea or theory, then such use should be worthy of fair use protection.80 A similar situation occurred when an unauthorized use of an artist’s evocative and graphic imagery in brochures objecting to “obscene art” was held a fair use.81 Fundraising brochures objecting to “‘offensive’ and ‘blasphemous’ art by the National Endowment   Id at 305.   Suntrust Bank v. Houghton Mifflin Co., 268 F.3d 1257, 1264 (11th Cir. 2001). 77   Time Inc. v. Bernard Geis Assocs., 293 F. Supp. 130 (S.D.N.Y. 1968). 78   Id at 146. 79   See Campbell v. Acuff-Rose Music, Inc., 510 U.S. 569, 579 (1994) (a more transformative use will weigh heavily towards a finding of fair use because the use “adds something new, with a further purpose or different character, altering the first with new expression, meaning, or message . . . The more transformative the new work, the less will be the significance of other factors, like commercialism, that may weigh against a finding of fair use”). 80   Compare L.A. News Serv. v. KCAL-TV Channel 9, 108 F.3d 1119, 1122 (9th Cir. 1997) (rebroadcast of “Beating of Reginald Denny” not transformative), with L.A. News Serv. v. CBS Broad., Inc., 305 F.3d 924 (9th Cir. 2002) (short use of same clip in montage a transformative use). 81   Wojnarowicz v. Am. Family Ass’n, 745 F. Supp. 130, 142 (D.N.Y. 1990). 75 76

BARFIELD PRINT.indd 526

13/11/2018 11:53

Reordering the chaos of the virtual arena  527 for the Arts” used cropped images of David Wojnarowicz’s works in an attempt to shock, offend, and motivate donors to object to the government-funded art.82 In such a situation, it is highly unlikely that the victim of such a campaign would voluntarily license his work to the proselytizer, so fair use provides a balancing of the interests of the artist and the organization objecting to the work. If there were any example where the First Amendment could be used as an independent tool granting access to the content, the political nature of the debate over the National Endowment of the Arts and the concomitant refusal by Wojnarowicz to allow such use could suggest that the First Amendment would be an independent ground to balance the interests. The court had no need to turn to an independent analysis using the First Amendment, however, because the elastic nature of the fair use doctrine provides more precedent and a more narrowly balanced framework than would a resort to constitutional first principles. Copyright fair use makes explicit that “the effect of the use upon the potential market for or value of the copyrighted work”83 must be taken into account in determining the scope of fair use. The inclusion of this aspect of the fair use test creates an economic incentive to license copyrighted works for exploitation. Copyright holders are encouraged to license their works because of precedent that gives economic protection to the market-based licensing while generally granting broader fair use rights to third parties in situations where the copyright holder has withheld licensing opportunities. In situations where the copyright holder refuses to license, the equities tend to shift towards fair use whereas in cases where the third party refuses to seek the license or pay the fee, the equities favor the rights owner. Through the use of both the idea/expression dichotomy and fair use, copyright holders are free to exploit the exclusive rights of their work without unduly limiting the expressive rights of others. Though the copyright owners occasionally object to others’ use of their rights, the system works to reward innovation, protect editorial integrity, and promote the marketplace of ideas. These concepts, in turn, have informed publicity rights jurisprudence. 2.4  Publicity Rights and General Applicability Publicity rights are the most recent of the intellectual property rights to be recognized under the law,84 and have evolved into an important economic resource for their creators and a tool to manage the editorial and commercial integrity of the goods associated with   Id at 133.  17 U.S.C. §107(4). 84   Kristina M. Sesek, Twitter or Tweeter: Who Should Be Liable for a Right of Publicity Violation under the DCA? 15 Marq. Intell. Prop. L. Rev. 237, 240 (2011). The notion of the right of publicity arose in the American legal community as a result of three pivotal law review articles. It was not until 63 years after the publication of these articles, in 1953, that Judge Frank of the Second Circuit Court of Appeals coined the term “right of publicity.” The Supreme Court recognized the right of publicity in 1977, as it noted the following different motivations behind the right of privacy and the right of publicity: economic interests drive the right of publicity, while privacy protections drive the right of privacy. Id, quoting Haelan Labs., Inc., v. Topps Chewing Gum, Inc., 202 F.2d 866, 868 (2d Cir. 1953), and citing Zacchini v. Scripps-Howard Broad. Co., 433 82 83

BARFIELD PRINT.indd 527

13/11/2018 11:53

528  Research handbook on the law of virtual and augmented reality the celebrity exercising the right. As such, protections of the economic and associative rights are similar to those of copyright and trademark. “The right of publicity is an intellectual property right of recent origin which has been defined as the inherent right of every human being to control the commercial use of his or identity.”85 Athletes, actors, and other celebrities expend considerable effort to maximize the potential for licensing, commercial exploitation, and identities. Publishers such as EA Sports invest heavily in these relationships so that their titles achieve dominant market presence.86 But the economic interest only has market value if the legal interests are respected by third parties and enforced by the courts or other relevant bodies. The protectible identity of a person is broadly defined: “[T]he Right of Publicity . . . is generally understood to encompass any personal attribute that identifies a particular person.”87 California statutory law lists one’s “name, voice, signature, photograph, or likeness” as the attributes of identity.88 While some states have yet to address the cause of action, most states provide a statutory cause of action, a common law cause of action, or both.89 California’s common law articulation is typical: “A cause of action for common law misappropriation of a plaintiff’s name or likeness may be pled by alleging: (1) the defendant’s use of the plaintiff’s identity; (2) the appropriation of plaintiff’s name or likeness to defendant’s advantage, commercially or otherwise; (3) lack of consent; and (4) resulting injury.”90 The right of publicity is distinct from either an inchoate right to be free from defamation or an invasion of privacy.91 One’s name or appearance in public does not give rise to a publicity rights claim. “It is only when the publicity is given for the purpose of appropriating to the defendant’s benefit the commercial or other values associated with the name or the likeness that the right of privacy is invaded.”92 There is some ­inconsistency on whether publicity rights are limited to those individuals who have exploited their U.S. 562, 576 (1977). See also Michael Madow, Private Ownership of Public Image: Popular Culture and Publicity Rights, 81 Cal. L. Rev. 125 (1993). 85   ETW Corp. v. Jireh Publ’g, Inc., 332 F.3d 915, 928 (6th Cir. 2003). 86   See, e.g., Kevin Gemmell, Video Scouting—For Sports Fans and Video Gamers—Even NFL Players—It’s a Mad, Mad, Mad, Mad, “Madden 2007” World, Where Graphics Are Good and a Team’s Playbook Is as Real as It Gets, S.D. Union-Trib., Sept. 3, 2006, at C (describing the relationship between EA and the NFL in creating the video game Madden). 87   W. Mack Webner & Leigh Ann Lindquist, Transformation: The Bright Line Between Commercial Publicity Rights and the First Amendment, 37 Akron L. Rev. 171 (2004). 88   Cal. Civ. Code § 3344 (2007): “(a) Any person who knowingly uses another’s name, voice, signature, photograph, or likeness, in any manner, on or in products, merchandise, or goods, or for purposes of advertising or selling, or soliciting purchases of, products, merchandise, goods or services, without such person’s prior consent, or, in the case of a minor, the prior consent of his parent or legal guardian, shall be liable for any damages sustained by the person or persons injured as a result thereof.” Id. 89   See generally 1–2 Gilson on Trademarks § 2.16. 90   Montana v. San Jose Mercury News, Inc., 40 Cal. Rptr. 2d 639, 640 (Ct. App. 1995). Cf Allison v. Vintage Sports Plaques, 136 F.3d 1443 (11th Cir. 1998). Another useful articulation of the elements which comprise the infringement of the right are derived from the Restatement (Third) of Unfair Competition. “(1) That defendant used plaintiff’s name as a symbol of his identity (2) without consent (3) and with the intent to obtain a commercial advantage.” Doe v. TCI Cablevision, 110 S.W.3d 363, 369 (Mo. 2003). 91  1–2 Gilson on Trademarks § 2.16. 92   Id.

BARFIELD PRINT.indd 528

13/11/2018 11:53

Reordering the chaos of the virtual arena  529 rights commercially—so-called celebrities—or whether the rights are protected for all individuals.93 In both California and New York, for example, there is no distinction based on the prior use.94 Although the history of publicity rights in the United States can properly be traced back to 1905,95 the modern age of licensing is better identified with the victory by Topps Chewing Gum, Inc., that allowed it to enforce exclusive rights in professional baseball players’ identity interests to help it sell gum by packaging it with baseball cards.96 Topps began a vigorous program of signing individual standard form contracts with thousands of both major and minor league baseball players. Under these agreements, each player granted Topps an exclusive right to publish his name, picture, signature, and biographical sketch “to be sold either alone or in combination with chewing gum, candy and confection or any of them.” As consideration, Topps guaranteed the player a lump-sum payment of $ 125.00 for each season in which either his picture was used or the player was an active member of a major league club. These contracts ran until Topps had made five years of payments to the individual player.97

The nature of the rights for baseball players trying to control their economic interest for playing cards is not dissimilar to those same athletes and actors claiming a right to control their identities when used for commercial purposes in computer games and virtual worlds. As the right developed, it needed (and still needs) to be untangled from other personal tort claims, such as defamation, false light invasion of privacy, unfair competition, and misrepresentation.98   Id.   Cal. Civ. Code § 3344 (2007); N.Y. Civ. Rights Law §§ 50–51 (McKinney 2002). 95   Pavesich v. New England Life Ins. Co., 50 S.E. 68 (Ga. 1905) (the first case to recognize the common law right of publicity in the United States). The common law doctrine was developed from the seminal law review article which argued that a common law right of publicity had long existed at common law under various names. See Samuel D. Warren & Louis D. Brandeis, The Right to Privacy, 4 Harv. L. Rev. 193 (1890). The common law right in New York was rejected by the Court of Appeals. See Roberson v. Rochester Folding Box Co., 64 N.E. 442 (N.Y. 1902). New York adopted its Civil Rights Law to reverse the court decision. 96   Haelan Labs., Inc. v. Topps Chewing Gum, Inc., 202 F.2d 866 (2d Cir. 1953). See also Fleer Corp. v. Topps Chewing Gum, 658 F.2d 139, 142 (3d Cir. 1981) (reviewing history of Topps licensing history). 97   Fleer Corp. v. Topps Chewing Gum, 658 F.2d 139, 142 (3d Cir. 1981). 98   Haelan Labs., Inc. v. Topps Chewing Gum, Inc., 202 F.2d at 868: 93 94

[A] man has a right in the publicity value of his photograph, i.e., the right to grant the exclusive privilege of publishing his picture, and that such a grant may validly be made “in gross,” i.e., without an accompanying transfer of a business or of anything else. Whether it be labelled a “property” right is immaterial; for here, as often elsewhere, the tag “property” simply symbolizes the fact that courts enforce a claim which has pecuniary worth.   This right might be called a “right of publicity.” For it is common knowledge that many prominent persons (especially actors and ball-players), far from having their feelings bruised through public exposure of their likenesses, would feel sorely deprived if they no longer received money for authorizing advertisements, popularizing their countenances, displayed in newspapers, magazines, busses, trains and subways. This right of publicity would usually yield them no money unless it could be made the subject of an exclusive grant which barred any other advertiser from using their pictures. Id.

BARFIELD PRINT.indd 529

13/11/2018 11:53

530  Research handbook on the law of virtual and augmented reality In Zacchini v. Scripps-Howard Broadcasting, Co.,99 the Supreme Court provided important recognition that publicity rights are a legitimate economic interest which can be recognized by states.100 [T]he State’s interest in permitting a “right of publicity” is in protecting the proprietary interest of the individual in his act in part to encourage such entertainment . . . [T]he State’s interest is closely analogous to the goals of patent and copyright law, focusing on the right of the individual to reap the reward of his endeavors and having little to do with protecting feelings or reputation.101

The Zacchini decision unequivocally established that state publicity rights were not automatically preempted by the First Amendment.102 The Court did not, however, wrestle with the challenge of the correct balancing test between these competing interests. Moreover, Zacchini was a somewhat anomalous example of a publicity rights claim because the real objection was the broadcast of Zacchini’s entire act rather than his appearance on television.103 Unlike Zacchini, in Topps and other cases articulating publicity rights, the primary underlying assumption emphasized the commercial rights in one’s likeness—that a person owns his or her identity for endorsing or advertising a product.104 The protection of this interest necessarily includes a protection of reputation since the endorsement of a poorly manufactured or dangerous product would directly harm the reputation of the person giving the endorsement.105 This should not blur the parameters of the right, however, but merely highlight the importance of protecting the use of the right as commercially important. Courts and federal agencies have expanded the notion of commercial endorsement, however, to include more creative ways of taking unfair advantage of others.106 The most important change to commercial endorsement resulted from the 2009 FTC Endorsement Guidelines promulgated under Section 5 of the Federal Trade Commission Act (“FTC Act”) to offer “guidance [for] the public in conducting its affairs in conformity with legal requirements . . . [regarding] the use of endorsements and testimonials in advertising.”107   433 U.S. 562 (1977).   Id at 572. 101   Id at 573. 102   Id. 103   Id at 570 (Time, Inc. v. Hill, 385 U.S. 374 (1967) applied the New York Times “actual malice” standard of knowingly false or reckless disregard of the truth, but even an accurate depiction “does not mandate a media privilege to televise a performer’s entire act without his consent”). 104   See, e.g., Wendt v. Host Int’l, Inc., 125 F.3d 806, 811 (9th Cir. 1997) (use of likeness as animatronic puppets in bars); Abdul-Jabbar v. Gen. Motors Corp., 85 F.3d 407, 416 (9th Cir. 1996) (use of prior name for car commercial); Carson v. Here’s Johnny Portable Toilets, Inc., 698 F.2d 831, 835–6 (6th Cir. 1983) (use of slogan for commode sales). 105   See, e.g., Stephanie Strom, A Sweetheart Becomes Suspect; Looking Behind Those Kathie Lee Labels, N.Y. Times, June 27, 1996, at D1. 106   See, e.g., FTC Guides Concerning the Use of Endorsements and Testimonials in Advertising, 74 Fed. Reg. 53,124, 53,125 (Oct. 15, 2009) (codified at 16 C.F.R. pt. 255); Nemani v. St. Louis Univ., 33 S.W.3d 184, 185 (Mo. 2000) (en banc) (using a researcher’s name in a grant application). 107   16 C.F.R. § 255.0(a) (2016). See also Jon M. Garon, Commercializing the Digital Canvas: Renewing Rights of Attribution for Artists, Authors, and Performers, 1 Tex. A&M L. Rev. 837, 854  99 100

BARFIELD PRINT.indd 530

13/11/2018 11:53

Reordering the chaos of the virtual arena  531 The FTC Endorsement Guidelines require “the honest opinions, findings, beliefs, or experience of the endorser.”108 The FTC Endorsement Guidelines also include a content requirement of truthfulness, requiring that “an endorsement may not convey any express or implied representation that would be deceptive if made directly by the advertiser.”109 In this way, the policy reflects the First Amendment that does not reach false commercial speech.110 Unfortunately there is no bright line distinguishing commercial activities of product sales, advertising billboards, and commercial advertising from expressive communications in newspapers, magazines, plays, films, or books. A photograph of a model wearing a provocative garment would be a commercial publication if included in an advertisement to sell the item but it would be editorial if selected by the publisher to illustrate a journalist’s story.111 The photograph has no intrinsic commercial or expressive form; the form is based on its particular use. With the use of commercial endorsements in social media and sales of goods through virtual worlds, parties continue to seek additional ways of exploiting the goodwill associated with celebrities for products unrelated to those celebrities. The distinction may be critically important to understand the rights of third parties to exploit celebrity identities: whether the use of a person’s name and identity is “expressive,” in which case it is fully protected, or “commercial,” in which case it is generally not protected. For instance, the use of a person’s identity in news, entertainment, and creative works for the purpose of communicating information or expressive ideas about that person is protected “expressive” speech. On the other hand, the use of a person’s identity for purely commercial purposes, like advertising goods or services or the use of a person’s name or likeness on merchandise, is rarely protected.112

The distinctions between commercial and expressive use has proven quite difficult to separate. There is wide disagreement on the interpretation and efficacy of the laws across jurisdictions.113 Because publicity rights are creatures of state law, both the articulation of

(2014); Jon M. Garon, Beyond the First Amendment: Shaping the Contours of Commercial Speech in Video Games, Virtual Worlds, and Social Media, 2012 Utah L. Rev. 607. 108   16 C.F.R. § 255.1(a) (2016). 109   Id. 110   See Cent. Hudson Gas & Elec. Corp. v. Pub. Serv. Comm’n, 447 U.S. 557, 563 (1980) (“there can be no constitutional objection to the suppression of commercial messages that do not accurately inform the public about lawful activity”); Va. State Bd. of Pharmacy v. Va. Citizens Consumer Council, Inc., 425 U.S. 748, 771 (1976) (“Untruthful speech, commercial or otherwise, has never been protected for its own sake”). See also Shafiel A. Karim, “Clear and Conspicuous” Disclosures between Celebrity Endorsers and Advertisers on Social Media Websites, 25 Competition: J. Anti., UCL & Privacy Sec. St. B. Cal. 172, 191 (2016). 111   See Stephano v. News Grp. Publ’ns, Inc., 474 N.E.2d 580, 584–5 (N.Y. 1984) (New York’s publicity law statute “does not define trade or advertising purposes. However, the courts have consistently held, from the time of its enactment, that these terms should not be construed to apply to publications concerning newsworthy events or matters of public interest”); Stewart v. Rolling Stone LLC, 105 Cal. Rptr. 3d 98, 104 (Ct. App. 2010), as modified on denial of reh’g (Feb. 24, 2010) (Rolling Stone magazine’s placement of a foldout on “indie rock” not commercial speech merely because the foldout was framed by advertising). 112   Doe v. TCI Cablevision, 110 S.W.3d 363, 373 (Mo. 2003) (citations omitted). 113   C.B.C. Distrib. & Mktg., Inc. v. Major League Baseball Advanced Media, L.P., 443 F. Supp. 2d 1077, 1084 (D. Mo. 2006) (discussing the right of publicity in various states).

BARFIELD PRINT.indd 531

13/11/2018 11:53

532  Research handbook on the law of virtual and augmented reality the specific rights and limitations of the interest vary a great deal from state to state.114 In contrast, the concern regarding expressive use of identity is the extent to which it impinges on federal constitutional rights of free speech. 2.5  Publicity Rights—Accommodations for the First Amendment The Zacchini decision unequivocally established that state publicity rights were not automatically preempted by the First Amendment and were sufficiently flexible to protect the commercial interest of the rightsholder.115 The Court did not, however, wrestle with the challenge of the correct balancing test between these competing interests. Instead, a variety of state and federal courts have evolved inconsistent and sometimes incompatible approaches to create the appropriate balancing of these interests. Despite a good deal of dicta from various courts looking to base their decisions on doctrine common across jurisdictions, three fairly distinct lines of decisions have developed. There are potentially seven or more different legal tests used to balance the free speech interests of publishers with the property interests of individuals, including the “actual malice” test borrowed from defamation law;116 the “transformative use test” borrowed from copyright fair use;117 the “Rogers Test,” borrowed from nominative trademark fair use;118 the “public interest” test;119 the “predominant use” test;120 the “adequate alternatives” test;121 ad hoc balancing;122 or a First Amendment bar to publicity rights in expressive works.123 In many ways, the national standard should be the actual malice test of New York

114   See, e.g., George P. Smith II, The Extent of Protection of the Individual’s Personality Against Commercial Use: Toward A New Property Right, 54 S.C. L. Rev. 1, 29 (2002). 115   Zacchini v. Scripps-Howard Broad. Co., 433 U.S. 562, 570 (1977). 116   Time, Inc. v. Hill, 385 U.S. 374, 397 (1967) (in interpreting the New York publicity statute the Supreme Court applied the actual malice standard for the use of a person’s identity, establishing that “a verdict of liability could be predicated only on a finding of knowing or reckless falsity in the publication”); Hoffman v. Capital Cities/ABC, Inc., 255 F.3d 1180 (9th Cir. 2001). 117   Comedy III Prods., Inc. v. Gary Saderup, Inc., 25 Cal. 4th 387, 21 P.3d 797 (Cal. 2001). 118   Rogers v. Grimaldi, 875 F.2d 994 (2d Cir. 1989). 119   Hilton v. Hallmark Cards, 599 F.3d 894, 912 (9th Cir. 2010). 120   Doe v. TCI Cablevision, 110 S.W.3d 363 (Mo. 2003). 121   The test has not been adopted for publicity rights, but it has been used for property and trademark litigation. See Lloyd Corp. v. Tanner, 407 U.S. 551 (1972) (rights of property owners to restrict handbill distribution); Mut. of Omaha Ins. Co. v. Novak, 836 F.2d 397, 402–3 (8th Cir. 1987) (“Mutant of Omaha,” parody of Mutual of Omaha’s logo, constituted trademark infringement); Dall. Cowboys Cheerleaders, Inc. v. Pussycat Cinema, Ltd., 604 F.2d 200, 206 (2d Cir. 1979) (trademarked uniforms worn by Dallas Cowboy Cheerleader were infringed when used as costumes in the adult film Debbie Does Dallas). 122   Guglielmi v. Spelling-Goldberg Prods., 25 Cal. 3d 860 (1979) (fictionalization of one’s life story protected by First Amendment); Lugosi v. Universal Pictures, 603 P.2d 425 (Cal. 1979) (in a dispute over the exploitation of Dracula, the actor’s likeness may be protected from film studio’s unauthorized commercialization of character). See also Cardtoons, L.C. v. Major League Baseball Players Ass’n, 95 F.3d 959, 966 (10th Cir. 1996) (if both publicity rights and First Amendment rights are “at stake, we proceed to a final determination of the relative importance of those rights in the context of [the] case”). 123   Tyne v. Time Warner Entm’t Co., 901 So. 2d 802 (Fla. 2005).

BARFIELD PRINT.indd 532

13/11/2018 11:53

Reordering the chaos of the virtual arena  533 Times Co. v. Sullivan,124 simply because it is the test adopted by the Supreme Court to address the use of publicity rights in expressive works. In Time, Inc. v. Hill,125 the actual malice test was used to remand a verdict against a magazine for publishing a false (but not defamatory) magazine story about the plaintiff that was later used to create a play about the plaintiff’s ordeal as a kidnapping victim. Since Brown established that video games have full First Amendment protection, video games should have the same treatment as magazine articles and plays. The problem with applying Time, Inc. v. Hill is that the actual malice test is not particularly clear or relevant.126 The test provides little information about the use of a person’s identity as a depiction on a t-shirt or in a science fiction world. Worse, because the focus turns on truthfulness rather than commercial exploitation, a mere disclaimer that the contents are parody or otherwise not intended to be true could be sufficient to overcome liability for highly commercial and inaccurate content.127 Because of the poor fit, the Supreme Court held in Zacchini that “the ‘actual malice’ standard does not apply to the tort of appropriation of a right of publicity.”128 Despite the Supreme Court moving video games into the same orbit covered by Time, Inc., the Zacchini decision separates out the publicity rights from the inchoate rights of privacy and defamation. Following this approach, the next most popular analogy has been to copyright law and the fair use accommodation afforded to the First Amendment. In Comedy III Prods., Inc. v. Gary Saderup, Inc.,129 the California Supreme Court created the “transformative use test” by incorporating the concept of transformative fair use from copyright law.130 The California Supreme Court created as a framework through which the rights of publicity and speech could be balanced, describing it as “at the heart of any judicial attempt to square the right of publicity with the First Amendment.”131 The California Supreme Court attempted to build upon the copyright fair use doctrine as a method of separating fair use from protected speech. As it explained, “both the First Amendment and copyright law have a common goal of encouragement of free expression and creativity.”132 Having identified fair use as the framework, the California court set out to create a rule under which courts and parties could operate: When artistic expression takes the form of a literal depiction or imitation of a celebrity for commercial gain, directly trespassing on the right of publicity without adding significant expression beyond that trespass, the state law interest in protecting the fruits of artistic labor outweighs the expressive interests of the imitative artist. On the other hand, when a work contains significant 124   N.Y. Times Co. v. Sullivan, 376 U.S. 254, 280 (1964) (“‘actual malice’—that is, with knowledge that it was false or with reckless disregard of whether it was false or not”). 125   Time, Inc. v. Hill, 385 U.S. 374 (1967). 126   See, e.g., Spahn v. Julian Messner, Inc., 221 N.E.2d 543 (N.Y. 1966) (actual malice standard did not preclude a finding of damages for an unauthorized depiction of a sports celebrity in a children’s book). 127   See Hustler Magazine, Inc. v. Falwell, 485 U.S. 46, 52 (1988) (“even though falsehoods have little value in and of themselves, they are nevertheless inevitable in free debate”). 128   Hustler, 485 U.S. at 52. 129   21 P.3d 797 (Cal. 2001). 130   Comedy, 21 P.3d at 808. 131   Id. 132   Id. at 404–5 (internal citations omitted).

BARFIELD PRINT.indd 533

13/11/2018 11:53

534  Research handbook on the law of virtual and augmented reality transformative elements, it is not only especially worthy of First Amendment protection, but it is also less likely to interfere with the economic interest protected by the right of publicity. As has been observed, works of parody or other distortions of the celebrity figure are not, from the celebrity fan’s viewpoint, good substitutes for conventional depictions of the celebrity and therefore do not generally threaten markets for celebrity memorabilia that the right of publicity is designed to protect.133

The court in Comedy III recognized that it was not articulating a single new legislative decree for publicity rights. Instead, the court tried to articulate its reasoning in a variety of ways. “Another way of stating the inquiry is whether the celebrity likeness is one of the ‘raw materials’ from which an original work is synthesized, or whether the depiction or imitation of the celebrity is the very sum and substance of the work in question.”134 This formulation emphasizes the relevance of the publicity rights to the output of the new author or artist. Incorporation of the identity interests remains essential for biographies, news reporting, social commentary, and many other forms of protected speech.135 Perhaps because the case was dealing with parody, in Comedy III the court chose not to incorporate the entirety of fair use, but instead to incorporate various prongs of the statutory formulation.136 Parody cases may be where the tensions run the highest rather than where the fair use is most robust. When asking “whether a product containing a celebrity’s likeness is so transformed that it has become primarily the defendant’s own expression rather than the celebrity’s likeness,”137 parodies of the celebrities are often the form of fair use that is most like the celebrity’s likeness and least transformed. Influenced by the particular use before it, the court rejected incorporation of the fourth statutory fair use element, “the effect of the use upon the potential market for or value of the copyrighted work.”138 The limitation should be revisited because in many cases outside of parody, such as biography, news reporting, or commercial endorsement, the potential market would be dispositive. One can never exploit the potential market for an unauthorized biography about oneself, nor should one be able to be compensated for news reporting about oneself. In contrast, commercial endorsements of products or services have a robust market that should never give rise to claims that the endorsement was a transformative use of the celebrity’s identity interests. Had the California courts adopted   Id.   25 Cal. 4th at 406, 21 P.3d at 809. 135   The Court’s own catalog of copyright examples is highly instructive:   We emphasize that the transformative elements or creative contributions that require First Amendment protection are not confined to parody and can take many forms, from factual reporting (see, e.g., Rosemont Enterprises, Inc. v. Random House, Inc. (1968) 58 Misc.2d 1 [294 N.Y.S.2d 122, 129], affd. mem. (1969) 32 A.D.2d 892 [301 N.Y.S.2d 948]) to fictionalized portrayal (Guglielmi, 25 Cal. 3d at pp. 871–2; see also Parks v. Laface Records (E.D.Mich. 1999) 76 F. Supp.2d 775, 779–782 [use of civil rights figure Rosa Parks in song title is protected expression]), from heavy-handed lampooning (see Hustler Magazine v. Falwell (1988) 485 U.S. 46 [108 S. Ct. 876, 99 L. Ed. 2d 41]) to subtle social criticism (see Coplans et al., Andy Warhol (1970) pp. 50–52 [explaining Warhol’s celebrity portraits as a critique of the celebrity phenomenon]). Id. 133 134

  Id.   Id. 138  17 U.S.C. § 107(4). 136 137

BARFIELD PRINT.indd 534

13/11/2018 11:53

Reordering the chaos of the virtual arena  535 the full fair use test rather than merely its transformative elements, the test would be more consistent with First Amendment goals and commercial protections. Despite the limitations of the transformative test, the California Supreme Court reaffirmed its approach in Winter v. DC Comics.139 Because the work in question was an expressive literary work, in this case a comic book, the Court reiterated two important limitations for publicity rights it espoused in Comedy III: We made two important cautionary observations. First, “the right of publicity cannot, consistent with the First Amendment, be a right to control the celebrity’s image by censoring disagreeable portrayals. Once the celebrity thrusts himself or herself forward into the limelight, the First Amendment dictates that the right to comment on, parody, lampoon, and make other expressive uses of the celebrity image must be given broad scope. The necessary implication of this observation is that the right of publicity is essentially an economic right. What the right of publicity holder possesses is not a right of censorship, but a right to prevent others from misappropriating the economic value generated by the celebrity’s fame through the merchandising of the ‘name, voice, signature, photograph, or likeness’ of the celebrity.” Second, in determining whether the work is transformative, courts are not to be concerned with the quality of the artistic contribution—vulgar forms of expression fully qualify for First Amendment protection. On the other hand, a literal depiction of a celebrity, even if accomplished with great skill, may still be subject to a right of publicity challenge. The inquiry is in a sense more quantitative than qualitative, asking whether the literal and imitative or the creative elements predominate in the work.140

In the comic book, Texas recording artists Johnny and Edgar Winter were depicted as “villainous half-worm, half-human offspring.”141 Given the limitations placed on publicity rights that they not be used to censor disagreeable portrayals or avoid vulgar expression, the two most significant complaints regarding the Winters’ depictions properly became irrelevant. The court also made clear that subtleties of interpretation between parody and satire suggested by the Supreme Court in Campbell were inapplicable in the case of publicity rights.142 In doing so, the California court signaled that a much wider berth should be given for the editorial use of publicity rights than fair use is afforded for copyright law. By determining that the use of the redrawn figures could be used for the comic books, the California court made clear that most insertions of caricatures of celebrities in other literary works would be considered a transformative use protected by the First Amendment. The transformative use test has also been applied to video games. In Hart v. Electronic Arts, Inc.143 and Keller v. Electronic Arts, Inc.,144 the Third Circuit and Ninth Circuit both applied the transformative use test to find that the use of collegiate athletes’ statistics, jersey numbers, height, weight, facial features, and in some cases biographical   30 Cal. 4th 881 (2003).   Id at 889. 141   Id at 886. 142   Id at 891. Cf. Campbell v. Acuff-Rose Music, Inc., 510 U.S. 569, 581 (1994); Dr. Seuss Enters, L.P. v. Penguin Books 109 F.3d 1394 (9th Cir. 1997). 143   717 F.3d 141 (3rd Cir. 2013) (although the transformative test permits the use of one’s likeness as “raw material” in a creative work, such as video games, here the game’s avatars use real player names, stats, hometowns, jersey numbers, height, weight, and even facial features). 144   Keller v. Elec. Arts, Inc., 2010 WL 445190 (Feb. 8, 2010), aff’d In re NCAA StudentAthlete Name & Likeness Licensing Litig., 724 F.3d 1268 (9th Cir. 2013). 139 140

BARFIELD PRINT.indd 535

13/11/2018 11:53

536  Research handbook on the law of virtual and augmented reality information constituted an appropriation of the players’ identities was not sufficiently transformative.145 The transformative use test was previously applied to a video game by an intermediate California court with the opposite result. In Kirby v. Sega of America, Inc.,146 a California appellate court addressed a right of publicity claim in the context of a Sega computer game named “Space Channel 5.”147 Plaintiff Kierin Kirby, lead singer of the retro-funk band Deee-Lite, complained that her identity was used for the game. As described by the court, the similarities were quite difficult to imagine. The court had only modest respect for the publicity claims brought by Kirby, but nonetheless found that she raised triable issues of fact.148 The California Court of Appeals applied Comedy III as interpreted by Winter to find that Kirby was at best the raw material for Ulala, the character in Sega’s game.149 The court noted that “notwithstanding certain similarities, Ulala is more than a mere likeness or literal depiction of Kirby. Ulala contains sufficient expressive content to constitute a ‘transformative work’ under the test articulated by the Supreme Court.”150 Such transformative changes included the anime style drawing, the height of the animated character, and the difference in dance style from Kirby. The concern raised by these cases is not the outcome but the focus of the approach. As applied in Keller, Winter, and Kirby, the copyright fair use test has evolved into an unhelpful transformative test. When described in Kirby, the very act of rendering a live person’s likeness into an avatar was transformative, but this was then rejected in Keller. Describing the depictions in Comedy III, the Kirby court illustrates its understanding of the difference. “The artist who created [the Three Stooges drawings], while highly skilled, contributed nothing other than a trivial variation that transformed the drawings from literal likenesses of the three actors.”151 145   Hart, 717 F.3d at 165; In re NCAA Student-Athlete Name & Likeness Licensing Litig., 724 F.3d 1268, 1271 (9th Cir. 2013) (“In NCAA Football, EA seeks to replicate each school’s entire team as accurately as possible. Every real football player on each team included in the game has a corresponding avatar in the game with the player’s actual jersey number and virtually identical height, weight, build, skin tone, hair color, and home state”). 146   50 Cal. Rptr. 3d 607, (Ct. App. 2006). 147   Id at 609. 148   Id at 613. 149   Id at 615. 150   Id at 616. The similarities which gave rise to a triable issue of fact are not mentioned. Ulala resembles Kirby in certain respects. Certain of Ulala’s characteristics and computer-generated features resemble Kirby’s. Both images are thin, and have similarly shaped eyes and faces, red lips, and red or pink hair. Both wear brightly colored, formfitting clothing, including short skirts and platform shoes in a 1960s retro style. In addition, Ulala’s name is a phonetic variant of “ooh la la,” a phrase often used by Kirby and associated with Kirby. Finally, as the trial court pointed out, both Kirby and Ulala used the phrases “groove,” “meow,” “dee-lish,” and “I won’t give up.” These similarities support Kirby’s contention  her identity was misappropriated. Id. at 613. The Kirby Court’s depiction of White is helpful. It was found that a “[n]onconsensual use of robotic image of celebrity Vanna White, dressed in wig, gown and jewelry regularly worn by White, turning letters on a game show set designed to look like the Wheel of Fortune, constitutes common law appropriation of celebrity’s singular identity.” Id at 613–14 citing White v. Samsung Elecs. Am., Inc., 971 F.2d 1395, 1399 (9th Cir. 1992). 151   Id at 615.

BARFIELD PRINT.indd 536

13/11/2018 11:53

Reordering the chaos of the virtual arena  537 The purpose of incorporating fair use into publicity rights analysis was to assure that legitimate respect was given to the free speech rights of artists and authors who used the identity of the famous in their expressive works. As the test is evolving, however, it is losing sight of its role in creating a critical accommodation for expressive works. A better, and perhaps entirely different, version of the fair use test is found in the Sixth Circuit.152 In ETW Corp. v. Jireh Publishing, Inc.,153 the commercial sports artist Rick Rush painted a montage of golf phenomenon Tiger Woods winning his first Masters Tournament in Augusta, Georgia in 1997.154 The ETW Corp. decision emphasized the Restatement (Third) of Unfair Competition comments in helping to shape the contours of the right. In particular, the Court highlighted “[c]omment d., which articulates a rule analogous to the rule of fair use in copyright law. Under this rule, the substantiality and market effect of the use of the celebrity’s image is analyzed in light of the informational and creative content of the defendant’s use.”155 The Sixth Circuit emphasized the need to adopt broad First Amendment protections for expression to foster a robust creative environment. Quoting Judge Kozinski from his dissent from White v. Samsung,156 the Court noted that “[o]verprotecting intellectual property is as harmful as under protecting it. Creativity is impossible without a rich public domain . . . Intellectual property rights aren’t free: they’re imposed at the expense of future creators and of the public at large.”157 The Kozinski dissent embraces the need to manifest a vibrant public domain and free speech ethos through a careful balance between the rightsholder and the public. The Sixth Circuit has attempted to weave these various threads together in a more careful and nuanced manner than other jurisdictions, incorporating rather than rejecting the cases that have been decided in other jurisdictions.158 By incorporating all the disparate sources of law into the analysis of Rush’s artwork depicting Woods’ victory at Augusta, the Sixth Circuit sought to bring balance and harmony back to the law of publicity rights. The court went on to discuss the highly creative and transformative work created by Rush, which met the transformative test set out in Comedy III, and allowed for a finding that no publicity rights were implicated. In contrast to the transformative, fair use analysis adopted by California, an interpretation of an Oklahoma statute very similar to the California publicity statute did not feel free

  ETW Corp. v. Jireh Publ’g, Inc., 332 F.3d 915 (6th Cir. 2003).   332 F.3d 915 (6th Cir. 2003). 154   Id at 918. The painting is far more than a simple portrait. In the foreground of Rush’s painting are three views of Woods in different poses. In the center, he is completing the swing of a golf club, and on each side he is crouching, lining up and/or observing the progress of a putt. To the left of Woods is his caddy, Mike “Fluff ” Cowan, and to his right is his final round partner’s caddy. Behind these figures is the Augusta National Clubhouse. In a blue background behind the clubhouse are likenesses of famous golfers of the past looking down on Woods. These include Arnold Palmer, Sam Snead, Ben Hogan, Walter Hagen, Bobby Jones, and Jack Nicklaus. Behind them is the Masters leader board. Id. 155   Id at 937. 156   White v. Samsung Elecs. Am., Inc., 971 F.2d 1395 (9th Cir. 1993) rehearing en banc denied, White v. Samsung Elec. Am., Inc., 989 F.2d 1512, 1513–15 (9th Cir. 1993) (Kozinski, dissenting). 157   Id. 158   ETW Corp. v. Jireh Publ’g, Inc., 332 F.3d 915, 936 (6th Cir. 2003) (footnote omitted). 152 153

BARFIELD PRINT.indd 537

13/11/2018 11:53

538  Research handbook on the law of virtual and augmented reality to incorporate fair use doctrine.159 Instead, in Cardtoons, L.C. v. Major League Baseball Players Ass’n,160 the court directly balanced the interest of the First Amendment rights against the benefits of protecting publicity rights.161 As the Tenth Circuit commented, “In resolving the tension between the First Amendment and publicity rights in this case, we find little guidance in cases involving parodies of other forms of intellectual property.”162 Absent statutory reference to fair use, the court found itself unable to incorporate the copyright law into the state’s publicity statute.163 This case instead requires us to directly balance the magnitude of the speech restriction against the asserted governmental interest in protecting the intellectual property right. We thus begin our analysis by examining the importance of Cardtoons’ right to free expression and the consequences of limiting that right. We then weigh those consequences against the effect of infringing on MLBPA’s right of publicity.164

Having set itself the task of determining the value and social utility of publicity rights, the Tenth Circuit opined that neither the economic nor the noneconomic incentives involved are all that compelling as compared to the benefits of parody.165 Summarizing the analysis of the Cardtoons decision, a district court described the balancing as: Economic interests that states seek to promote include the right of an individual to reap the rewards of his or her endeavors and an individual’s right to earn a living. Other motives for creating a publicity right are the desire to provide incentives to encourage a person’s productive activities and to protect consumers from misleading advertising. But major league baseball players are rewarded, and handsomely, too, for their participation in games and can earn additional large sums from endorsements and sponsorship arrangements. Nor is there any danger here that consumers will be misled, because the fantasy baseball games depend on the inclusion of all players and thus cannot create a false impression that some particular player with “star power” is endorsing CBC’s products.   Then there are so-called non-monetary interests that publicity rights are sometimes thought to advance. These include protecting natural rights, rewarding celebrity labors, and avoiding emo-

  Cardtoons, L.C. v. Major League Baseball Players Ass’n, 95 F.3d 959 (10th Cir. 1996).   95 F.3d 959 (10th Cir. 1996). 161   Id at 969. 162   Id at 970. 163   Id. 164   Id at 972. 165   See, e.g., the reliance on Madow, Private Ownership of Public Image: Popular Culture and Publicity Rights, 81 Cal. L. Rev. 125, 128 (1993): 159 160

Because celebrities take on personal meanings to many individuals in the society, the creative appropriation of celebrity images can be an important avenue of individual expression. As one commentator has stated: “Entertainment and sports celebrities are the leading players in our Public Drama. We tell tales, both tall and cautionary, about them. We monitor their comings and goings, their missteps and heartbreaks. We copy their mannerisms, their styles, their modes of conversation and of consumption.” Whether or not celebrities are “the chief agents of moral change in the United States,” they certainly are widely used–far more than are our institutionally anchored elites – to symbolize individual aspirations, group identities and cultural values. Their images are thus important expressive and communicative resources: the peculiar, yet familiar idiom in which we conduct a fair portion of our cultural business and everyday conversation. Id.

BARFIELD PRINT.indd 538

13/11/2018 11:53

Reordering the chaos of the virtual arena  539 tional harm. We do not see that any of these interests are especially relevant here . . . “Publicity rights . . . are meant to protect against the loss of financial gain, not mental anguish.” We see merit in this approach.166

The outcome of this case would have likely been the same under almost any of the tests, but the dismissal of the state’s interest in protecting publicity rights was inconsistent with the Supreme Court decision in Zacchini and utterly dismissive of the legislative power of a state to recognize and regulate the property interests of its citizens. The difficulty caused by the Cardtoons approach is the categorical balancing of free expression against publicity rights rather than judging the particular use of identity interests as weighed against competing but significant governmental interests. Perhaps the best that can be said regarding the Cardtoons decision was the characterization by the California Supreme Court. “While Cardtoons contained dicta calling into question the social value of the right of publicity, its conclusion that works parodying and caricaturing celebrities are protected by the First Amendment appears unassailable in light of the test [adopted in California].”167 The Supreme Court of Missouri also rejected the California transformative fair use test and other jurisdictional precedents.168 The Court instead adopted Mark Lee’s suggestion of a “predominant use test.” It found that the California approach could lead to unintended and undesirable third party use. It rejected Comedy III because “the transformation or fictionalized characterization of a person’s celebrity status is not actionable even if its sole purpose is the commercial use of that person’s name and identity.”169 Instead it sought a more accommodating balance of interests to address expressive works used for commercial exploitation: If a product is being sold that predominantly exploits the commercial value of an individual’s identity, that product should be held to violate the right of publicity and not be protected by the First Amendment, even if there is some “expressive” content in it that might qualify as “speech” in other circumstances. If, on the other hand, the predominant purpose of the product is to make an expressive comment on or about a celebrity, the expressive values could be given greater weight.170

The Missouri court also rejected the Restatement’s “relatedness” approach because it again objected to the categorical nature of the test.171 In the relatedness test, unauthorized parties are free to use another person’s name or identity in a work that is “related to” that person.172 The Restatement provides the simplest test, harkening back to the axiomatic rule that publicity rights protect a person from others exploiting that person’s identity to sell a good or service without permission by stating that “if the name or likeness is used solely to attract attention to a work that is not related to the identified person, the user 166   C.B.C. Distrib. & Mktg. v. Major League Baseball Advanced, L.P., 505 F.3d 818, 824 (8th Cir. 2007), quoting Cardtoons, 95 F.3d at 973–6 (internal citations omitted). 167   Comedy III Prods., Inc. v. Gary Saderup, Inc., 21 P.3d 797, 809 (Cal. 2001). 168   Doe v. TCI Cablevision 110 S.W.3d at 374. 169   Id. 170   Id (quoting Mark S. Lee, Agents of Chaos: Judicial Confusion in Defining the Right of Publicity–Free Speech Interface, 23 Loy. L.A. Ent. L. Rev. 471, 500 (2003)). 171   Id at 373–4. 172   Id. at 373 citing Restatement (Third) Of Unfair Competition § 47 cmt. c.

BARFIELD PRINT.indd 539

13/11/2018 11:53

540  Research handbook on the law of virtual and augmented reality may be subject to liability for a use of the other’s identity in advertising.”173 The relatedness test essentially boils down to a simple question whether the identity was used in an advertisement, in packaging, or on a product.174 If the identity was used in this fashion without permission, it is actionable. Used in almost any other manner, the use does not violate publicity rights.175 The Missouri Supreme Court required a careful balancing approach because it was upholding the finding of liability against a comic book author in a case not too dissimilar from Winter. In Doe v. TCI Cablevision,176 Todd McFarlane created a comic book character using the name of St Louis Blues hockey “enforcer” Tony Twist for a mafia goon and enforcer in the popular Spawn comic.177 Unlike the depictions in Winter, McFarlane used the plaintiff’s name and a reasonable likeness, adding no demonic caricature of Twist.178 Nonetheless, the court upheld a damage award of 15 million dollars. The Missouri Supreme Court explained the result, at least in part, based on the marketing of the comic as hockey merchandise because of the steps taken to market Spawn to hockey audiences.179 As interpreted by the Missouri Supreme Court, the predominant test here creates a substantial risk of liability for authors well beyond the economic protections suggested by Zacchini or other courts.180 The mere use of a name within an expressive comic book whether or not combined with an unrelated likeness would have been sufficiently transformative under Comedy III to avoid liability in California. The use would have also avoided any liability under the relatedness test of the Restatement since the use was not used to sell the comic but was instead related to the common attributes of this hockey player to mafia goons. Given the chilling effect of the damage award, Doe provides a

  Id.   In this simplistic but perhaps preferable approach, mugs, lunch boxes, and posters should be considered commercial products. Photographs, lithographs, and all expressive media—including newspapers, magazines, radio, television, motion pictures, video games, virtual worlds, and plays— would be expressive and outside of concerns regarding publicity rights claims. Cf. Restatement (Third) of Unfair Competition § 47 cmt. c. T-shirts create the most difficult category because they often serve an expressive purpose, but are also often the source of copyright and trademark piracy. Given the First Amendment importance, t-shirts might be best treated as expressive for publicity rights purposes, though still subject to copyright and trademark law protections. Whatever arbitrariness such an approach introduces, it removes the case by case analysis required by present law which may result in a chilling effect and the substantial self-censorship inevitable to avoid potential liability and the costs associated with litigation. 175   Such a formulation does not preclude liability for misrepresenting the endorsement or involvement of a person. False endorsements, inclusion of researchers’ names in grant applications, and other fraudulent activities would still be actionable under various unfair competition doctrines unrelated to publicity rights. 176   110 S.W.3d 363 (Mo. 2003). 177   Id. 178   Id at 366. 179   Id at 371 (“respondents marketed their products directly to hockey fans”). 180   Had McFarlane used the name of Twist in those marketing efforts (as opposed to the mere use of the name of a character within the comic book), then the trademark balancing approach suggested in Rogers v. Grimaldi would have been more appropriate. See Rogers v. Gramaldi infra at note 118 and accompanying text. 173 174

BARFIELD PRINT.indd 540

13/11/2018 11:53

Reordering the chaos of the virtual arena  541 strong illustration regarding the benefits of a narrow, brightline test as suggested by the Restatement. Just as the courts in Keller and Kirby extended Comedy III, the Eighth Circuit has extended Cardtoons in C.B.C. Distribution and Marketing v. Major League Baseball Advanced, L.P.181 In CBC, a fantasy sports operator brought a declarative action asserting its freedom to use “names, nicknames, likenesses, signatures, pictures, playing records, and/or biographical data of each player in an interactive form in connection with its fantasy baseball products.”182 Although no visual images were used in connection with the for-profit fantasy league site, most other attributes of identity were utilized. The Eighth Circuit began the analysis by identifying the high social utility in access to the statistics of America’s pastime: The public has an enduring fascination in the records set by former players and in memorable moments from previous games . . . “The records and statistics remain of interest to the public because they provide context that allows fans to better appreciate (or deprecate) today’s performances.” The . . . “recitation and discussion of factual data concerning the athletic performance of [players on Major League Baseball’s website] command a substantial public interest, and, therefore, is a form of expression due substantial constitutional protection.   In addition, the facts in this case barely, if at all, implicate the interests that states typically intend to vindicate by providing rights of publicity to individuals . . . [M]ajor league baseball players are rewarded, and handsomely, too, for their participation in games and can earn additional large sums from endorsements and sponsorship arrangements.183

The court thus found that the players’ claims were precluded by the First Amendment.184 The court saw little in either the economic or noneconomic interests of the players to offset the public’s interest in access to game statistics. Compare this result to a similar case decided in 1970. Throughout the 1960s, Negamco’s Major League Baseball and Big League Manager Baseball were sold as sports fantasy games. The two games “employ[ed] the names and professional statistical information such as batting, fielding, earned run and other averages of some 500 to 700 major league baseball players, identified by team, uniform number, playing position and otherwise.”185 The unauthorized use of the information now protected by the First Amendment was barred in the earlier situation. The historical change in the expectation that players control the statistics of their play for purposes of fantasy games dictates the outcome of the litigation. If the court assumes that such an ownership right exists, as it did in Uhlaender v. Henricksen,186 then a game manufacturer cannot exploit that right without permission. If the court assumes that no such right exists, as it did in CBC, then the game manufacturer is free to sell the game. No

  505 F.3d 818 (8th Cir. 2007).   Id at 823. 183   Id at 823–4 quoting Gionfriddo v. Major League Baseball, 114 Cal. Rptr. 2d 307 (Ct. App. 2001). 184   CBC, 505 F,3d at 823–4. 185   Uhlaender v. Henricksen, 316 F. Supp. 1277, 1278 (D. Minn. 1970) (“SCIENTIFICALLY COMPUTED Players are rated in every phase of baseball play. Each pitcher is different and each batter is different. You manage 520 big time players. Your strategy affects the outcome of every game. This game is Big, Colorful and True. 220 pitchers and 300 fielders are included”). 186   Uhlaender 316 F. Supp. at 1278. 181 182

BARFIELD PRINT.indd 541

13/11/2018 11:53

542  Research handbook on the law of virtual and augmented reality empirical justification is available to inform this assumption. It is an essentially arbitrary starting point, normatively based. Again, a reflection on copyright fair use and the Restatement (Third) of Unfair Competition can provide a more nuanced approach.187 Without challenging the finding that the public has a high interest in the baseball statistics, the assertion does not support a finding that the statistics should be controlled by the players or be available free to the public. Free access should reduce the costs to the public, but it may also reduce the reliability of the data gathered. The lack of ownership also means the lack of any quality control from the players. Copyright fair use, in contrast, would allow the court to assess the effect on the market. To the extent that the fantasy statistics are fully available through licensed channels to the public, the public need should be satiated. To the extent that the transaction costs involved in collecting the data are too high to license the information, fair use should provide an adequate remedy. Instead, the Eighth Circuit has made a value judgment regarding the publicity rights of the players as used in an unlicensed commercial fantasy sports business. This is perhaps yet another consequence of the California Supreme Court’s rejection of portions of the fair use test in Comedy III. The Restatement’s relatedness test may provide a better explication regarding the shift of player statistics than is suggested by CBC from the position originally in Uhlaender. Board games (Uhlaender) and video games (CBC) are both products. But websites are content, much closer to newspapers and television than to lunch boxes or billboards. As websites and simple ball and paddle videogames converged into virtual worlds, the line was crossed and the works became wholly protected speech. Player data can no longer be protected as expression because the data has become the facts and ideas necessary for communication about the sport. Fantasy leagues are merely another method of that expression. It is the role of the content which has changed, making CBC correct for its age just as Uhlaender was correct in its era. Taken together, Kirby and CBC reflect a strong swing away from protecting the identity interests of celebrities and toward free speech interests of producers for games and virtual worlds. Keller and Hart reflect and equally swing in the other direction. The inconsistency invariably leads to confusion and an increase in transaction costs as parties struggle to understand what can be used from the public domain and what must be licensed from property holders. 2.6  Intersections with Trademark Fair Use and First Amendment Protection In almost every publicity rights case, the plaintiffs are also able to bring a cause of action for either a registered trademark,188 or an unregistered mark, under § 43(a) of the Lanham Act.189 Tiger Woods, for example, has a registered trademark in both “Woods” and “Tiger 187   It is only an analogy, since copyright does not provide protection for any of the individual statistics and the players have not copyrighted work embodying the order selection and arrangement of the statistics in any copyrighted fashion. See 17 U.S.C. § 102. 188  15 U.S.C. §§ 1051 et. seq. 189  15 U.S.C. §1125(a). The provision protects against false designation of origin as to the sources of goods and services and protection against the false advertising of goods and services.

BARFIELD PRINT.indd 542

13/11/2018 11:53

Reordering the chaos of the virtual arena  543 Woods,”190 while singer Tom Waits and athlete Kareem Abdul-Jabbar have both raised unfair competition claims using § 43(a).191 Professional athletes in league sports may have registered trademarks in their names, while the professional sports leagues for which they play own the registered trademarks in the team names, logos, and related team indicia. Section 43(a) adds protection to stop a false endorsement, “which is likely to confuse consumers as to the [celebrity’s] sponsorship or approval of the product.”192 Like copyright law and publicity rights, trademark law limits the exclusivity afforded by trademark.193 First, any party who wishes to use trademarks can do so if such use does not create a likelihood of confusion.194 Moreover, as a defense to trademark infringement, trademark law generally recognizes two fair use exceptions and a First Amendment exception to allow third parties to exploit trademarks and trade dress without express permission. Under traditional trademark fair use, the words which are owned as a trademark can be used as fair use when accurately describing a company’s own good or service. For example, although “VCR-2” is a legitimate trademark for Go-Video, the trademark holder cannot stop JVC from using VCR-2 to label the second VCR input jack on the back of its own devices.195 Trademark law was not “meant to deprive commercial speakers of the ordinary utility of descriptive words.”196 The second form of trademark fair use is nominative fair use. A party may properly use the trademark of another to accurately describe the trademark owner’s product. For example, an independent Volkswagen repair mechanic was allowed to use the trademark Volkswagen in its advertising to show the makes and models for which it provides

See, e.g., E.S.S. Entm’t 2000, Inc. v. Rock Star Videos, Inc., 444 F. Supp. 2d 1012, 1027 (D. Cal. 2006). 190   ETW Corp. v. Jireh Publ’g, Inc., 332 F.3d 915, 920–21 (6th Cir. 2003). 191   Abdul-Jabbar v. GMC, 85 F.3d 407, 410 (9th Cir. 1996); Waits v. Frito-Lay, Inc., 978 F.2d 1093, 1108 (9th Cir. 1992); Midler v. Ford Motor Co., 849 F.2d 460, (9th Cir. 1988). 192   Id. 193   See, e.g., ETW Corp., 332 F.3d at 920. As an initial matter, trademark law “provides a defense to an infringement claim where the use of the mark ‘is a use, otherwise than as a mark . . . which is descriptive of and used fairly and in good faith only to describe the goods . . . of such party . . . .’” Id quoting 15 U.S.C. § 1115(b)(4). 194   KP Permanent Make-Up, Inc. v. Lasting Impression I, Inc., 543 U.S. 111, 117 (2004). Additional protections are available for “famous marks,” but such protection is beyond the scope of this article. See 15 U.S.C. § 1125(d). 195   In re Dual-Deck Video Cassette Recorder Antitrust Litig., 11 F.3d 1460, 1467 (9th Cir. 1993). See also KP Permanent Make-Up, Inc. v. Lasting Impression I, Inc., 543 U.S. 111 (2004); Brother Records, Inc. v. Jardine, 318 F.3d 900, 905–6 (9th Cir. 2003) (fair use “applies only to marks that possess both a primary meaning and a secondary meaning—and only when the mark is used in its primary descriptive sense rather than its secondary trademark sense”). 196   KP Permanent Make-Up, Inc., 543 U.S. at 122. On remand, the Ninth Circuit adopted the following set of factors for determining fair use:   Among the relevant factors for consideration by the jury in determining the fairness of the use are the degree of likely confusion, the strength of the trademark, the descriptive nature of the term for the product or service being offered by [the defendant] and the availability of alternate descriptive terms, the extent of the use of the term prior to the registration of the trademark, and any differences among the times and contexts in which [the defendant] has used the term.   KP Permanent Make-Up, Inc. v. Lasting Impression I, Inc., 408 F.3d 596, 609 (9th Cir. 2005).

BARFIELD PRINT.indd 543

13/11/2018 11:53

544  Research handbook on the law of virtual and augmented reality service.197 Such accurate descriptions are allowed “even if the defendant’s ultimate goal is to describe his own product.”198 For purposes of comparing the approach to publicity rights, the tests of both the Ninth and Third Circuits are instructive. Under the Ninth Circuit approach, the following characterization applies: [W]here the defendant uses a trademark to describe the plaintiff’s product rather than its own product, a commercial user is entitled to a nominative fair use defense provided the defendant meets three requirements:

(1) the product or service in question must be one not readily identifiable without use of the trademark; (2) only so much of the mark or marks may be used as is reasonably necessary to identify the product or service; and (3) the user must do nothing that would, in conjunction with the mark, suggest sponsorship or endorsement by the trademark holder.199

The Third Circuit has modified the Ninth Circuit test with its own three-prong analysis:

(1) Is the use of plaintiff’s mark necessary to describe (1) plaintiff’s product or service and (2) defendant’s product or service? (2) Is only so much of the plaintiff’s mark used as is necessary to describe plaintiff’s products or services? (3)  Does the defendant’s conduct or language reflect the true and accurate relationship between plaintiff and defendant’s products or services?200

Perhaps one of the reasons that the Ninth Circuit developed its particular articulation has been the extent to which the trademarks in question were related to identity issues, such as in the cases of the band New Kids on the Block,201 surfer legends including George Downing and Paul Strauch,202 actor George Wendt,203 and others for whom their trademarks were their identity. It may be that whatever variation may result from the somewhat different articulations of the two competing formulations in general, there will be no meaningful difference in cases involving identity or publicity rights. In both categories, the nominative fair use question will turn on whether the use of the celebrity’s identity to properly identify that celebrity suggests sponsorship and endorsement. This also answers the question whether “the true and accurate relationship” is accurate or misstated as well as whether the extent to which the identity was used was only “reasonably necessary.”

  Volkswagenwerk Aktiengesellschaft v. Church, 411 F.2d 350 (9th Cir. 1969).   Horphag Research Ltd. v. Pellegrini, 337 F.3d 1036, 1040–41 (9th Cir. 2003) quoting Cairns v. Franklin Mint Co., 292 F.3d 1139, 1152 (9th Cir. 2002). 199   Downing v. Abercrombie & Fitch, 265 F.3d 994, 1009 (9th Cir. 2001); upon New Kids on the Block v. News Am. Publ’g, Inc., 971 F.2d 302 (9th Cir. 1992). 200   Century 21 Real Estate Corp. v. Lendingtree, Inc., 425 F.3d 211, 229 (3d Cir. 2005). See Graeme B. Dinwoodie & Mark D. Janis, Lessons from the Trademark Use Debate, 92 Iowa L. Rev. 1703 (2007). 201   New Kids on the Block, 971 F.2d 302. 202   Downing, 265 F.3d at 994. 203   Wendt v. Host Int’l, Inc., 125 F.3d 806 (9th Cir.1997). 197 198

BARFIELD PRINT.indd 544

13/11/2018 11:53

Reordering the chaos of the virtual arena  545 A billboard for a miniature golf facility saying “Tiger Woods has never won on our golf course” would be accurate but misleading. While it might be humorous to have a large photograph of Woods in despair as he missed a putt, the use of image and name would strongly suggest sponsorship—even if the text of the ad suggests otherwise. In contrast, a restaurant that posts a wall of photographs of famous celebrities taken as those celebrities are eating inside the restaurant accurately reflects that those celebrities dined in the facility, and their placement inside the restaurant rather than on billboards minimizes any likelihood of consumer confusion that the celebrities are endorsing the restaurant, except to the extent that the individuals depicted had dined there. The differences in formulation may be significant in certain trademark contexts, but the differences should not affect the outcome of cases inside computer games and virtual worlds, particularly for trademarks related to identity interests. There have been examples of the role for nominative fair use involving video games and virtual worlds. In E.S.S. Entertainment 2000, Inc. v. Rock Star Videos, Inc.,204 a district court applied the nominative fair use approach of the Ninth Circuit to a claim for trademark and dress infringement when the plaintiff’s marks “Play Pen” and “Totally Nude” were parodied within the video game Grand Theft Auto: San Andreas.205 The court did not take the simple expedient of dismissing the claim under the theory that the defendant used the plaintiff’s mark within an expressive work or under the theory that plaintiff failed to demonstrate any use of its marks as either a false association or false advertising with a product or service. Instead, the opinion emphasized nominative fair use under trademark law. The court acknowledged the circularity of applying a nominative fair use analysis since “the lack of anything that suggests sponsorship or endorsement—is merely the other side of the likelihood-of-confusion coin.”206 Not surprisingly, nominative fair use did not apply to the satirical use within the video game. The game’s use of the name “Pig Pen” and the trade dress did not attempt to refer to the plaintiff’s products, “but only to describe their own product.”207 Had the court applied the likelihood of confusion analysis rather than applying “the other side of the coin” to require an affirmative defense, it would have had far less difficulty in dismissing the claim because there was no association with any good or service being offered by the plaintiff.208 Partially due to the confusion regarding nominative fair use, trademark law analysis requires an additional or alternative First Amendment analysis when expressive works are involved. The First Amendment limitation on trademark rights attempts to balance the limited property rights of the trademark holder with the public’s interest in an open marketplace of ideas and the speaker’s right to communicate. “When unauthorized use of another’s mark is part of a communicative message and not a source identifier, the

  444 F. Supp. 2d 1012, 1027 (D. Cal. 2006).   Id at 1014. The plaintiff’s disputed the characterization of the depiction as parody. While it may be true, the use could better be said to satirize Los Angeles than to parody the Play Pen’s name as trademark “Totally Nude,” but such distinction, if relevant at all, only affects the scope of copyright fair use under Campbell. See id at 1016. 206   Id at 1029 quoting Brother Records, Inc. v. Jardine, 318 F.3d 900, 909 n. 5 (9th Cir. 2003). 207   E.S.S. Entm’t 2000, Inc., 444 F. Supp. 2d at 1036. 208   Id. 204 205

BARFIELD PRINT.indd 545

13/11/2018 11:53

546  Research handbook on the law of virtual and augmented reality First Amendment is implicated in opposition to the trademark right.”209 Increasingly, courts have recognized that were they “to ignore the expressive value that some marks assume, trademark rights would grow to encroach upon the zone protected by the First Amendment.”210 The leading case articulating the balance of the trademark owners’ rights and the expressive rights in creative works was decided by the Second Circuit in Rogers v. Grimaldi.211 Rogers involved a complaint by the famed actress Ginger Rogers regarding the title of a movie, Ginger and Fred, which referred to Fred Astaire and Ginger Rogers motion pictures as a metaphor, but did not involve or relate to either of the screen legends. Addressing the concern that Rogers’ name was being used for the film’s title, the Second Circuit rejected first a narrow First Amendment exception under the alternative avenues of communication test,212 and then a lower court’s categorical assertion that trademark law was simply inapplicable to works of artistic expression.213 Steering clear of either extreme position, the Court announced a balancing of interests between the trademark holder and the author: We believe that in general the Act should be construed to apply to artistic works only where the public interest in avoiding consumer confusion outweighs the public interest in free expression. In the context of allegedly misleading titles using a celebrity’s name, that balance will normally not support application of the Act unless the title has no artistic relevance to the underlying work whatsoever, or, if it has some artistic relevance, unless the title explicitly misleads as to the source or the content of the work.214

The result of the test has been to require that there be a relationship between the literary content of a work and the trademark used in the title of the work as a condition of providing First Amendment protection for the speech. The use of a title in a work has at least some commercial aspects, which gives rise to the right to protect the public from fraudulent or misleading products and advertising.215 Such an approach is quite consistent with First Amendment jurisprudence that finds little value in commercial speech that is intentionally misleading.216 The Ninth Circuit has taken the same approach, extending Rogers to cases involving the use of the trademark “Barbie” in a song title,217 as well as in parody photographs.218 Many courts have also extended Rogers balancing past the titles of expressive works to the

  Yankee Publ’g, Inc. v. News Am. Publ’g, Inc., 809 F. Supp. 267, 276 (S.D.N.Y. 1992).   MCA, Inc. v. Wilson, 677 F.2d 180, 200 (2d Cir. 1981). 211   875 F.2d 994 (2d Cir. 1989). 212   Id at 998. Cf. Mut. of Omaha Insurance Co. v. Novak, 836 F.2d 397, 402 (8th Cir. 1987). 213   Id. 214   Rogers v. Grimaldi, 875 F.2d 994, 999 (2d Cir. 1989). 215   See id at n. 5 (“This limiting construction would not apply to misleading titles that are confusingly similar to other titles. The public interest in sparing consumers this type of confusion outweighs the slight public interest in permitting authors to use such titles”). 216   See Peel v. Attorney Disciplinary Comm’n, 496 U.S. 91 (1990); In re R.M.J., 455 U.S. 191, 203 (1982); Cent. Hudson Gas & Elec. Corp. v. Pub. Serv. Comm’n, 447 U.S. 557 (1980). 217   Mattel, Inc. v. MCA Records, 296 F.3d 894 (9th Cir. Cal. 2002) (song title Barbie Girl protected speech). 218   Inc. v. Walking Mt. Prods., 353 F.3d 792 (9th Cir. 2003). 209 210

BARFIELD PRINT.indd 546

13/11/2018 11:53

Reordering the chaos of the virtual arena  547 covers of books, as well as to trademarks used within the content of the literary work.219 As the courts have suggested, “in deciding the reach of the Lanham Act in any case where an expressive work is alleged to infringe a trademark, it is appropriate to weigh the public interest in free expression against the public interest in avoiding consumer confusion.”220 The Sixth Circuit took a similar approach for the trademark aspects of Tiger Woods’ name when used in conjunction with Rush’s painting of the golfing victory at Augusta.221 The Rogers approach thus posits two reasons to restrict the use of a trademark or celebrity name. Either the mark has no artistic relevance to the underlying work or the use of the mark “explicitly misleads as to the source or the content of the work.”222 Returning to the marks in Grand Theft Auto: San Andreas, the use of plaintiff’s marks— Pig Pen, Totally Nude, and nude figure trade dress223—have relevance to a stylized Los Angeles streetscape in which the marks appear, and the marks do not suggest anything to mislead the players of the game as to the source or content of the game. With regard to the source of the game, the court pointed out that “when First Amendment interests are implicated, the Rogers ‘explicitly misleading’ standard applies, not the traditional ‘likelihood of confusion’ test.”224 The tests regarding consumer expectations are quite important for adjudicating the ability of artists and authors to use trademarks within expressive works. Given the commercialized nature of celebrity endorsements, general audiences may tend to assume that any use of a brand name, celebrity identity, or other mark is done only with permission. There is little social utility, however, in allowing normative expectation to reduce the range of free speech to authors and artists. When focusing on the content within an expressive work, audiences are not harmed by the confusion. The audience member’s choice to read, watch, or play has already been made. This was a key factor in the Rogers court focusing on the title rather than on the content within the movie. The expressive use within the movie should have even greater First Amendment protection than the title. On the other hand, even protected speech has little utility when it is written to be explicitly misleading as to the source. A video game that claims to be the “approved and official game or web page” has little First Amendment protection if it is made without consent and is trying to hold itself out in a fraudulent manner to the public.225

219   Cliff Notes, Inc. v. Bantam Doubleday Dell Publ’g Grp., 886 F.2d 490, 495 (2d Cir. 1989). See 2 Thomas J. McCarthy, McCarthy on Trademark and Unfair Competition § 10:22 (4th ed. 2006). 220   Id. See also Hoffman v. Capital Cities/ABC, Inc., 255 F.3d 1180, 1183 (9th Cir. 2001). 221   ETW Corp., 332 F.3d at 937. 222   Rogers, 875 F.2d at 999. 223   E.S.S. Entm’t 2000, Inc. v. Rock Star Videos, Inc., 444 F. Supp. 2d 1012, 1038–39 (D. Cal. 2006). 224   Id at 1045. 225   See, e.g., MCA Records, 296 F.3d at 900; L.L. Bean, Inc. v. Drake Publishers, Inc., 811 F.2d 26, 29 (1st Cir. 1987). In contrast, the important distinction is how the fictional work is presented to the public. If someone wishes to produce a comic parody of commercialization by creating a television skit about the making of a TV commercial using athletes, the producer of the skit should not require permission of the athletes or manufacturers to depict real athletes or products in the parody. The skit producer is not intending to mislead the public into thinking the commercial or the endorsement is real.

BARFIELD PRINT.indd 547

13/11/2018 11:53

548  Research handbook on the law of virtual and augmented reality As such, the Rogers test provides a well-balanced test to protect expressive works while recognizing the need to protect the public and trademark owners. When used in complement with the Restatement (Third) of Unfair Competition and the full copyright fair use tests, the need to protect the property interests of publicity rights and the need to promote expressive works can be effectively balanced.

3. WHERE SPEECH ENDS IN VIRTUAL WORLDS: WHAT CAN AND SHOULD BE REGULATED 3.1  Where Publicity Meets Free Expression: From Fair Use to Avatar-Jacking NCAA Football, which ceased production in 2014, and Grand Theft Auto: San Andreas together represent the state of the modern case law regarding computer games and virtual worlds. These games are entitled to First Amendment protection, thwarting city and state attempts to ban ultraviolent or socially repugnant works. NCAA Football protected the publicity rights of the players who had their images used without consent and licensed many of the other elements within the game. Grand Theft Auto represented the other extreme. It is an expressive work, entitled to fair use and First Amendment protection, which allows unauthorized use of intellectual property—copyrights, trademarks, and publicity rights—to be incorporated into content. These are highly valuable commercial properties earning their owners tremendous revenue and setting the stage for a high stakes battle regarding control of the intellectual property contained in these games. Against this backdrop, it will be useful to assess the legal consequences of the introductory example, www.PRO.us (“PRO”), a virtual world incorporating the best and worst aspects of Second Life, Madden NFL, and Grand Theft Auto. In this hypothetical, Players, Inc., the for-profit licensing arm of the National Football League Players Association, will bring suit against the owners of PRO and will also file takedown notices against the website hosting companies where its member-players’ trademarks and publicity rights are being exploited without authorization.226 In analyzing the hypothetical case, courts following Keller, Digital Software, Kirby, E.S.S. Entertainment 2000, and CBC will treat the game as an expressive work rather than merely a product. Just as with the cases attempting to ban ultraviolent games,227 226   Such a lawsuit would be filed not just by Players, Inc. but also by all unions and leagues affected. The focus on a single plaintiff is for purposes of illustration only. The league and individual teams would bring trademark actions for the use of the registered trademarks, logos, and mascots of the teams; broadcasters may have copyright claims for copying of television footage and still photography, as well as copyright claims for intermediate copying for the use of such photographs and footage into what later became the avatars and game graphics. The players would also be expected to bring defamation claims regarding unsportsmanlike activities and steroid use allowed in the game, and false light claims for the use of the identity in gladiatorial games. The hypothetical also avoids the distinction between amateur and professional athletes at the heart of the Electronic Arts licensing practices that distinguished between NCAA “amateur” athletes and professional athletes. This article provides no opinion as to the efficacy of those additional causes of action. 227   See, e.g., Interactive Dig. Software Ass’n v. St. Louis Cty., 329 F.3d at 956; Video Software

BARFIELD PRINT.indd 548

13/11/2018 11:53

Reordering the chaos of the virtual arena  549 recent court decisions have consistently accepted video games as expressive.228 Since the various legal tests for fair use and First Amendment protection no longer rely on merely categorizing the game as expressive or commercial, however, the recognition that video games are expressive is merely the first step in the analysis. Turning first to the use of the trademarks within the game, claims that PRO is unable to use these trademarks are directly comparable to CBC. Both the teams and the players will make trademark claims, but both will face similar challenges for protection. Under trademark law, the teams and leagues would have a stronger claim than the players because the use of team names, logos, and colors for various fantasy leagues goes beyond the nominative fair use tests of the Ninth and Third Circuits. While the use of the player’s professional affiliation would be necessary for the user to create one’s team, there is less support for the claim that names, logos, or colors are necessary to build and play the team. Both the Third Circuit and Ninth Circuit nominative fair use tests focus on the necessity of the use and the amount taken suggests that it is substantial. As a result, nominative fair use may provide little protection from the teams. The sufficiency and need for taking will be influenced by the expressive use made of these elements. Publishers of PRO will assert that the game itself creates a parody of the leagues and the values of professional sports. To-the-death fighting, steroid availability, free agency, and interchangeability of players among sports all comment on the inherent nature of professional sport. Unlike the slavish realism of NCAA Football, PRO provides an experience that makes a caricature of the realistic sporting environments. Taken one step further, parody versions of these publicity rights trademarks, created in a whimsical fashion like that in Cardtoons, would take less of the original and enhance the expressive elements. The additional analysis required under the First Amendment further reflects a predisposition to protect free expression. It is quite difficult to parody the teams or players without access to the team names, colors, or logos. The public interest in free expression will outweigh any concerns regarding likelihood of confusion.229 The satirical view of professional sports built into the game will have a strong influence on the First Amendment defense. The most difficult area to predict remains the publicity rights of the athletes. After Keller and Hart, the Third and Ninth Circuits have shifted the focus to the transformation of the individual, with limited discussion of the transformative or nontransformative environment for that individual. Accurate depictions of the athletes would militate against a finding of transformative fair use. Nonetheless, in PRO, there is an overwhelming amount of additional content, manipulation of the images, use of the images in the games, and implicit social commentary on the players, leagues, and sports. Like the image of Tiger Woods being embraced by golf’s pantheon, the players’ images are merely a part of the message communicated. In the context of nominative fair use, courts following the reasoning of these cases will likely allow PRO to use the entirety of an identity. Given the extensive nature of the game Dealers Ass’n v. Webster, 968 F.2d 684, 688 (1992); Am. Amusement Mach. Ass’n v. Kendrick, 244 F.3d 572, 577 (7th Cir. 2001); Wilson v. Midway Games, Inc., 198 F. Supp. 2d 167, 181 (D.C. Conn. 2002). 228   Id. 229   See, e.g., Hoffman v. Capital Cities/ABC, Inc., 255 F.3d 1180, 1183 (9th Cir. 2001).

BARFIELD PRINT.indd 549

13/11/2018 11:53

550  Research handbook on the law of virtual and augmented reality and the comprehensive, highly transformative use to which the identity is put, celebrities and athletes should expect to have little ability to stop such use. Like an unauthorized biography, there is no amount of detail that crosses the line because of the context in which the raw material is used. If a California court were to take the analysis of Comedy III and Kirby seriously, then the additional, highly transformative elements of PRO should overcome the core concern of Keller and Hart that the players were the heart of the game. Football cannot exist without football players and the essence of NCAA Football was its highly detailed realism. At an instinctual level, the game felt like a commercial exploitation of the players. PRO has no slavish representation of particular players and teams. It presents much more like an adult-rated, hyper-realistic version of a sports programmer’s nightmare. Given the protection afforded to free expression under Kirby and CBC, courts would find it difficult to rule against PRO despite Keller, even if the judges found the content undesirable. To the extent that Doe creates any inconsistency in the court decisions, the facts of that decision render it far less persuasive. Doe emphasized that the commercial exploitation of hockey players occurred in a manner unrelated to the comic book Spawn, but directly related to marketing efforts to sell the comic books.230 A parallel between Doe and PRO would only exist if PRO utilized the players’ likenesses to target the marketing of the game. Given the factually specific nature of Doe, there is little chance that it would sway a court’s opinion. Having said that, the outcome could depend on the jurisdiction in which the case is brought. The Missouri court would be highly critical of the analysis just described. As in Doe, the creators of PRO are emphasizing the game for the sports fan. The more a gamer knows of players and statistics, the more valuable the game features become. The use of real players is unequivocally intended to create a commercial advantage for the producers of the game.231 Applying “a sort of predominant use test,”232 PRO’s producers would be hard pressed to establish that the predominant purpose of the game was “to make an expressive comment” on or about a player or players.233 This very narrow exception to publicity rights requires the object of the publicity rights claim to be the subject of the work at issue. Although taken as a whole, the players are the subject of the game; the emphasis of the work is on the gameplay and not on the players. Real identities are not required to make the game work or even to comment on the violence or drug use associated with professional sports. Were a court closely to follow Doe, the players would be much more likely to have their identity rights protected. The Doe approach highlights the economics underlying the game. Why should the players be forced to forgo income that will be made by third parties? By including the effect on the market, copyright fair use includes the ability to assess this question. But at present, only Zacchini fully reiterates this concern.234 Even under Zacchnini’s approach, however, a court can distinguish the right of the player to control his or her performance   Doe v. TCI Cablevision, 110 S.W.3d at 373.   See id at 369. 232   Id at 374. 233   Id. 234   “The rationale for [protecting the right of publicity] is the straight-forward one of preventing unjust enrichment by the theft of good will.” Zacchini v. Scripps-Howard Broad. Co., 433 U.S. 230 231

BARFIELD PRINT.indd 550

13/11/2018 11:53

Reordering the chaos of the virtual arena  551 from the right to recreate a user-generated performance based on attributes of the performer. Under a proper copyright fair use approach, courts would find fair use because a voluntary license is unlikely, given the nature of the parody in PRO. 3.2  Finding Balance Among the Balancing Tests As applied in Doe, the Missouri predominant use test goes too far in requiring a direct reference to the particular celebrity as a prerequisite to protecting free expression interests. On the other hand, California goes too far in excluding the market from publicity fair use analysis. As provided in the Restatement (Third) of Unfair Competition, the full fair use analysis of copyright provides such a synthesis as does the combination of trademark law’s fair use, nominative fair use, and First Amendment defenses. Copyright, of course, is not an entirely useful comparison. If the players were literary characters or comic book characters, then PRO would be an infringing work unless the lifting of the characters constituted fair use. The copyright in a celebrity’s authorized avatar will have greater legal protection than the publicity rights available to the person behind the avatar. Absent fair use, the extensive use of the characters in a video game would be an unauthorized derivative work.235 More importantly, the players are not themselves the work. The identity rights protected for humans are far less than the literary rights of characters protected for their authors. Looking to a comprehensive fair use copyright analysis, the unauthorized use of the names and likeness would be balanced against the commercial nature of the game, the highly transformative use of the identities, the entirety of the identities taken, and the potential effect on the market. The effect on the market for publicity rights is an important factor under a copyright style analysis. Although PRO does not substitute for watching professional sports, PRO does directly compete with Madden Sports and other toys and licensed games. The economic rights to the market will be affected by PRO. So either all of these products should no longer be protected or PRO will require a license. This is an accurate reflection of fair use, but it also has problematic consequences. A fair use approach to publicity rights ignores the transaction cost difficulties of such a doctrine. Fair use, as applied in the copyright context, is a highly fact specific regimen which may create a significant chilling effect on content that merely risks running afoul of infringement. Although this problem exists in copyright as well, centuries of practice and industry norms have allowed publishers to make many assumptions about the contours of fair use for purposes of quotations, reviews, news, commentary, and many other unauthorized reproductions of third party content. The extent of fair use parody may be hard to predict,236 but many other forms of copyright fair use are largely routine. In contrast, the two cases that have thus far attempted to create a new common law fair 562, 576 (1977) quoting Kalven, Privacy in Tort Law—Were Warren and Brandeis Wrong? 31 Law & Contemp. Prob. 326, 331 (1966). 235   Cf. Pirone v. MacMillan, Inc., 894 F.2d 579 (2d Cir. 1990) (Babe Ruth’s publicity rights and trademark rights in his name and likeness insufficient to stop use of photographs in a calendar). 236   See, e.g., Campbell v. Acuff-Rose Music, Inc. 510 U.S. 569, 579 (1994). Cf Dr. Seuss Enters., L.P. v. Penguin Books USA, Inc., 924 F. Supp. 1559, 1575 (S.D. Cal. 1996), aff’d, 109 F.3d 1394 (9th Cir. 1997).

BARFIELD PRINT.indd 551

13/11/2018 11:53

552  Research handbook on the law of virtual and augmented reality use doctrine for publicity rights are quite complex and the role of such fact-specific rules in the mercurial worlds of computer games and virtual worlds makes such rule making more harmful than helpful. A court applying a more comprehensive approach, such as the Sixth Circuit, would be more likely to find that the commercial nature of the game and the ability of the ­publishers and the players to earn revenue, the comprehensive use of all attributes of identity, and the history of video game licensing would weigh heavily against the transformative nature of the game. Under such a balancing test, it would be hard to predict if PRO would be required to license the players’ publicity rights or permit only fictional avatars to be used. The law both within and between the circuits remains unresolved. There are far too many tests, and the application of the tests to particular facts remains highly subjective. Only once the parameters of the online activities become clear can usefully predictive fair use rules begin to develop. Absent good fair use rules, the public is probably better served by a rejection of a cause of action for misappropriation of performances than it would be if such a cause of action existed. 3.3  Advertising Inside the Game and the Need for Disclosure Virtual worlds have the ability to include advertising and commercial speech within the game, an attribute that separates them from books and other literary works.237 Billboards in video games can be sold to advertisers, or change dynamically depending on the attributes of the game or the individual user.238 In PRO, for example, the producers of the game could show Gatorade being used on football sidelines since that product is commonly used in real-world football games.239 In the alternative, PRO’s publishers could sell the branding rights to a particular company, create a fictional product, or have the product which is displayed be based on the computer browsing habits of each user in relation to the advertiser’s products.240 In the not too distant future, the choice of drinks advertised to the player of the game could be based on the individual consumer preferences of that gamer.241 Needless to say, the ability to affiliate particular products in the game to each 237   But see Steve Rose, As Seen on TV: Why Product Placement Is Bigger than Ever, Guardian (June 24, 2014, 13:06 EDT), www.theguardian.com/tv-and-radio/2014/jun/24/breaking-bad-tvproduct-placement. The motion picture industry and television industries are also increasing the incorporation of paid commercial content and product placement. Id. The professional actors involved in these works have the ability to negotiate for an income stream from these revenues through collective bargaining agreements. Id. 238   See, e.g., Derrik Lang, What if Watching TV Were Like Playing a Game? Mpls Star-Trib., Dec. 23, 2007, at F3. 239   The technique of replacing advertising within movies has been both exploited and litigated. See Sherwood 48 Assocs. v. Sony Corp. of Am., 213 F. Supp. 2d 376, 377 (S.D.N.Y. 2002), aff’d, 76 Fed. Appx. 389, 2003 U.S. App. LEXIS 20106 (2003) (digital replacements of billboards in New York’s Times Square for Sony’s Spiderman were not infringing of the building’s trade dress). 240   See Kaylene Williams et al, Product Placement Effectiveness: Revisited and Renewed, J. Mgm’t and Mkt’g Research, 5–6, www.aabri.com/manuscripts/10712.pdf. 241   See FTC Staff Revises Online Behavioral Advertising Principles, Federal Trade Comm’n (Feb. 12, 2009), www.ftc.gov/news-events/press-releases/2009/02/ftc-staff-revises-online-behavioraladvertising-principles. The behavioral advertising could be based on general trends or track each

BARFIELD PRINT.indd 552

13/11/2018 11:53

Reordering the chaos of the virtual arena  553 gamer’s user behavior will eventually have far-reaching economic consequences with regard to reinforcing brands and adjusting marketing strategies. Such individuation of advertising would also change the relationship between the celebrity and the product to be advertised or endorsed. The commercial endorsements and advertising in virtual worlds should be viewed as a form of commercialization rather than expressive speech because the gamer will be able to move directly to the commercial product on the Internet. Whether the commercial is directly linked to the product’s website or whether the gamer must open a second browser to reach the product, the user is much closer to the point of purchase than a billboard or television advertisement. If advertisements or endorsements take place inside the game for real products that are readily available for purchase, those advertisements and endorsements should require the licensing of the identity. Of course, if the commercial is for a “wacky package”242 or parody along the lines of Cardtoons,243 then a lack of commercial payment and the parodic nature of the ad can be taken into account. A comedic TV sketch for a “Bass-o-Matic” remains a parody whether on television or playing on a television within a virtual world,244 so the placement of an advertisement within the virtual world should not materially affect the analysis. The determination should be based on whether the product is intended to be promoted to a gamer or viewer. If stadium billboards inside the game show celebrities endorsing commercially available products, those billboards should be treated the same as their brick and mortar counterparts. A commercial display of a product directly associated with a character or avatar should not be treated as transformative merely because the commercial appears in the game or virtual world. The need to identify commercial endorsements and advertising for products that occur within the video games and virtual worlds suggests that the various tests of fair use, transformative use, and the First Amendment must be applied to the specific use of the publicity rights within the game rather than more generally to the game as a whole. Similarly, the sale of virtual products with publicity rights or trademarks should be treated as commercial transactions if the sale results in an exchange of something of value that has economic consequences outside of the virtual environment. Unlike the complex balancing surrounding the correct fair use approaches to publicity rights protection, concerns regarding the use of publicity rights and avatars within the video games and virtual worlds have a relatively simple regulatory solution. Broadcasting user individually. The former would be accomplished through brand surveys, cross-tabulation of the gamer to spending and purchasing habits tracked with credit cards and store affinity cards, or inferred through interaction with other computer content. In the latter case, for example, the company could assess the amount of screen time (or “stickiness”) various products had for each gamer, then incorporate the advertising associated with the stickiest product on a user-by-user basis. Id. 242   Wacky Packages, Topps, www.topps.com/shop-entertainment-brands/wacky-packages.html (last visited July 19, 2017) (Wacky Packages are magnets, stickers, and postcards which parody particular brand names such as “Canaduh Dry,” “Enlisterine,” and “Cavemanwich.”) 243   Dave Anderson, Sports of the Times; Baseball’s New Satire: “Cardtoons,” N.Y. Times, June 3, 1995, at C1. 244   See Urban Dictionary, www.urbandictionary.com/define.php?term=bass-o-matic (“Coined by SNL alumnus Dan Ackroyd in the 1970s, the bass-o-matic is an ordinary blender used to turn fresh bass (fish) into a disgusting red slurry that no one in his right mind would actually drink”).

BARFIELD PRINT.indd 553

13/11/2018 11:53

554  Research handbook on the law of virtual and augmented reality already has comprehensive rules for identification of paid sponsorships,245 and email advertising requires disclosures.246 The online endorsement guidelines further reduce the ambiguity regarding what constitutes advertising online.247 A simple extension of these laws would create a regulation requiring that paid advertising and sponsorships be identified for the user. Such a rule would not be difficult to implement since the law could simply require a notice page of the endorsements, sponsorships, and advertisers included in the product. The relevance here is that the disclosure requirement would help define which uses of an avatar were in furtherance of a commercial purpose and which were expressive. If a person is depicted in a biographical movie drinking a Coke, there is no assumption that CocaCola paid to have its product associated with that famous person. Listing the companies paying for product placement might change that.248 Consider a biographical movie of Humphrey Bogart and Lauren Bacall. Authenticity might require that the famed couple’s actual cigarette brand preferences be shown, but there will be a tremendous financial temptation to create a sponsorship opportunity to sell the brand featured in the motion picture. Were such a production made for the internet, a requirement that the advertisers be listed on a web page associated with the site would have little chilling effect. At the same time, the disclosure of commercial endorsement would help satisfy the public’s compelling interest regarding the commercial promotion of products. 3.4  Who Owns the Unauthorized Content Anyway? Yet one additional wrinkle unique to the issues of virtual worlds is the ownership tension between the creators of content in the worlds and the publishers of the worlds themselves. Virtual world publishers require end-user license agreements, which set forth the contractual rights of the gamer and the publisher.249 The terms of service or end-user license agreements used in these relationships typically require that the user not violate the rights of third parties, and set forth the ownership of any intellectual property created or imported into the virtual world by the user. Publishers try to immunize themselves from liability to third parties by including provisions in the agreements prohibiting the user from uploading content that violates the rights of third parties.250 As a result, to the extent that the use of celebrity publicity rights violates the rights of the celebrities, such user conduct is a breach of the 245   47 U.S.C. § 317 (2007) (Announcement of payment for broadcast rules); 47 CFR 73.4180 (Payment disclosure: Payola, plugola, kickbacks). 246   Can-Spam Act of 2003, Pub. L. No. 108-187, 117 Stat. 2699 (2003) (codified at 15 U.S.C. §§ 7701–7713 and 18 U.S.C. § 1037). 247   See 16 C.F.R. § 255.0(a) (2016). 248   Cf. Superman II, Imdb.com, www.imdb.com/title/tt0081573/ (the movie prominently featured the Coca Cola sign which adorned Times Square.) 249  Reuveni, supra note 13 at 269–70. See also F. Gregory Lastowka & Dan Hunter, The Laws of the Virtual Worlds, 92 Calif. L. Rev. 1 (2004). The interpretation and enforceability of these agreements is beyond the scope of this article. 250   See generally Steven J. Horowitz, Competing Lockean Claims to Virtual Property, 20 Harv. J.L. & Tech. 443 (2007); Andrew E. Jankowich, Property and Democracy in Virtual Worlds, 11 B.U. J. Sci. & Tech. L. 173 (2005).

BARFIELD PRINT.indd 554

13/11/2018 11:53

Reordering the chaos of the virtual arena  555 ­ greement. The consequences are typically set out in the agreement, but generally a include the right of the publisher to terminate the user’s account and prohibit the user from resubscribing. The specific contractual language may be important with regard to publicity rights. If the terms of service agreement were to specify copyright law and defamation, then publicity rights would be omitted from the prohibition. Were the prohibition written not to violate any other person’s “intellectual property rights,” then additional ambiguity would be introduced due to the inconsistency of the jurisdictions in labeling publicity rights as a personal right, property right, or intellectual property right.251 The agreements may have a similar ambiguity with regard to the rights retained by the user and the publisher. Commercial sites often require that players waive any claims to the intellectual property created by the gamer inside the game and assign those rights to the publisher.252 In contrast, Second Life does not require the users to relinquish the intellectual property created and uploaded to the virtual world, although it does continue to protect its own content.253 Each publisher has a slightly different variation on the language used. Assuming, arguendo, that no third party interests were violated by gamers when creating characters and avatars incorporating the publicity rights of athletes and performers, then under many of the license agreements, the copyrightable character created from the user’s adaptation of the celebrity would become the property of the game publisher. The publishers, not the end users, would gain the economic value of these interests. It would be highly ironic for courts to encourage a liberal interpretation of First Amendment jurisprudence to allow the creation of these avatars, only to have the rights then transferred by contract to the major commercial publishers by operation of clickwrap contract law.

4. REORDERING CHAOS—COLLECTIVE BARGAINING AND BEST PRACTICES FOR THE PROFESSIONAL NEIGHBORHOODS OF VIRTUAL WORLDS At present, with the exception of Doe, Keller, and Hart, the courts reviewing claims for publicity rights have not been generous to their owners. Courts have been highly respectful of free expression for the publishers of virtual worlds and computer games and slow to protect the property interests of athletes and performers. Were this trend to continue, it would extend beyond virtual worlds to create a legal basis for challenging the type of exclusive deals presently being made by EA Sports and other publishers.

251   A similar example arose with unauthorized uses of musicians’ images on t-shirts produced by Kendall and Kylie Jenner. See Jason Newman & Jon Blistein, The Doors Issue Cease-and-Desist Letter to Jenners Over T-Shirts, Rolling Stone (June 29, 2017), www.rollingstone.com/music/ news/the-doors-issue-cease-and-desist-letter-to-jenners-over-t-shirts-w490497. 252   See, e.g., Terms of Service and User Agreement, Sony Interactive Ent. (Jan. 1, 2017), www. playstation.com/en-us/network/legal/terms-of-service/. See Reuveni, supra note 13 at 263–64. 253   Second Life Terms and Conditions, Linden Lab, www.lindenlab.com/legal/second-life-terms-a​ nd-conditions.

BARFIELD PRINT.indd 555

13/11/2018 11:53

556  Research handbook on the law of virtual and augmented reality Keller and Hart may represent a significant change in judicial temperament regarding these cases. If not, one response to the inconsistent and increasingly unfriendly judiciary could be an attempt to change the state laws or enact federal legislation. Such laws would inevitably face First Amendment challenges, raising even more questions regarding the scope of available protection.254 Moreover, to the extent that the judicial approach reflects the political will to protect celebrity interests, the efforts to protect these rights might result in substantial opposition in many jurisdictions.255 A preferred alternative would be for the professional athletes and actors to use their collective bargaining clout to establish an appropriate balance between the economic interests of the players and the expressive interests of the public and the publishers. The controversy that triggered Keller and Hart was the NCAA ban on amateur athlete payments,256 a policy that forced Electronic Arts to forego licensing the athletes in NCAA Football even as it paid the athletes in Madden Football and its other titles.257 Through collective bargaining and the establishment of recommended best practices, the celebrities are likely to gain an appropriate level of protection while respecting the free expression and creative interests important in the nuanced balancing between the First Amendment and the publicity rights. The collective bargaining agreements, in turn, should include provisions requiring that any other parties in business with the signatories to the collective bargaining agreements will also be required to commit to the practices required under the collective bargaining agreement. In this way the publicity rights provisions will extend beyond the bargaining unit to all common divisions under joint ownership with the employers, and eventually to many of their contracting partners. 4.1  Collective Bargaining as an Alternative to Regulation Collective bargaining provides an excellent gapfiller for those areas in which intellectual property law cannot properly balance all inequities.258 The terms of the collective bargaining agreement can establish the rights to compensation, credit, and association of the person’s identity with products or services.259 The existing collective bargaining 254   Moreover, the analysis described throughout this article is limited to United States law. Both the First Amendment and the publicity rights are highly idiosyncratic to United States law and the foreign rules applying to publicity rights, though not dealt with in this article, will turn on different property and state interests, depending on the jurisdiction. 255   But see Stats 2007 ch 439 § 1 (S.B. 771) (2008) (amending Cal. Civ. Code § 3344.1 to make retroactive publicity rights protection available to those who died prior to January 1, 1985). 256   See O’Bannon v. Nat’l Collegiate Athletic Ass’n, 802 F.3d 1049, 1055 (9th Cir. 2015) (“The gravamen of O’Bannon’s complaint was that the NCAA’s amateurism rules, insofar as they prevented student-athletes from being compensated for the use of their NILs, were an illegal restraint of trade under Section 1 of the Sherman Act, 15 U.S.C. § 1”). 257   O’Bannon, Jr. v. National Collegiate Athletic Ass’n, et. al., Order Granting Final Approval of Class Action Settlement and Final Judgment, Case Nos. 4:09-CV-1067 CW; 4:09-CV-3329 CW 1 (N.D. Cal., July 16, 2015), www.ncaa-ea-likeness-settlement.com/media/354716/2015-0819_-_final_approval_order__ea_settlement_.pdf. 258   Cf Toney v. L’Oreal, 406 F.3d 905 (7th Cir.2004); Fleet v. CBS, 50 Cal. App. 4th 1911 (1996) (contract, not copyright or right of publicity, establishes the rights of an actor who voluntarily appears in a movie). 259   See Welch v. Carson Prods. Grp., Ltd., 791 F.2d 13 (2d Cir. 1986).

BARFIELD PRINT.indd 556

13/11/2018 11:53

Reordering the chaos of the virtual arena  557 agreements are moving from traditional film and television into internet use and reuse of creative content developed by union members.260 Welch v. Carson Productions Group provides an example of how the collective bargaining agreement provides a more nuanced solution to the tensions between publicity rights and production rights than the law can provide.261 Under § 36 of the Screen Actors Guild Television Agreement then applicable to the dispute, the reuse of a commercial required payment to the actors, but also provided a mechanism to allow the producers to use the commercial if the actor could not be identified or contacted.262 The provision balanced the performers’ rights with very practical considerations regarding the transaction costs involving recordkeeping and reuse of content. The collective bargaining agreement specified the required minimum payments and set forth how the reuse would take place.263 Unlike court decisions, the agreement provides rules for the situation where the party cannot be located. In this fashion, the agreements can move beyond the legal rights of the parties to address the practical difficulties they face. Collective bargaining agreements for actors and athletes generally reach issues of endorsement and commercial use of images, protecting performers from the unauthorized use of their images to commercialize a product. At the same time, the agreements provide explicit authority to use the identity interests of the performers to market, sell, and promote the audiovisual work in which the performer appears. As a result, all rights to place the image of an actor on the cover of a DVD case or on posters have been acquired. For athletes there are additional considerations. Because groups of players may be used for some commercial endorsements, the collective bargaining agreements have the ability to specify the interests of the player, the team, and the league. The rights of a player to use his or her team jersey, team logos, colors, and so on may prove essential to the ability of that player to commercially exploit his or her publicity rights. When a group of players collectively endorse a product, the team may have an economic or proprietary interest. Legal determinations regarding the scope of publicity rights are irrelevant to the balancing of these interests and their role in the overall bargain between the performers and their employers. The collective bargaining agreement is the law of the deal. Under a collective bargaining approach to exploitation of publicity rights in video games and virtual worlds, the publishers receive the right to use the identity of a celebrity for noncommercial, expressive activities in exchange for clear prohibitions against using the identity for commercial exploitation. In keeping with this basic approach, the following points would need to be negotiated:

  Peter Sanders, Directors, Studios Reach a Deal, W.S.J., January 18, 2008 at A3.   791 F.2d 13 (2d Cir. 1986). 262   Id at 14–15. “‘Reuse of Photography or Sound Track,’ provides that a producer may not reuse television film of an actor in a manner other than that for which the actor originally was employed “without separately bargaining with the player and reaching an agreement regarding such use . . . [If] the Producer is unable to find the player, it shall notify the Guild [SAG], and if the Guild is unable to find the player within a reasonable time, the Producer may use the photography . . . without penalty . . .”Id, quoting 1977 Screen Actors Guild Television Agreement (“Green Book”) § 36. 263   Id. 260 261

BARFIELD PRINT.indd 557

13/11/2018 11:53

558  Research handbook on the law of virtual and augmented reality   (1) The right of publicity would have to be recognized by the publishers and producers of the video games, websites, and virtual worlds. The right would extend to members of the bargaining unit as well as any other person, except for nonunion members appearing in crowds or large groups.264  (2) The definition of commercial activity would be specified to focus on the sale, marketing, or endorsement of any product or service offered to the public. It would exclude all expressive, literary, or editorial content, and provide limited use of unlicensed identity interest for advertising or marketing of the work in which the celebrity appeared as a performer or athlete.  (3) The publishers and producers would be obligated to make an effort to avoid interfering with the exploitation of the publicity rights by union members with products or services which were in competition with the products and services for which the union members had existing exclusive arrangements.265 For example, because Tiger Woods has an exclusive endorsement agreement with Nike, absent a compelling editorial need, he would be depicted with a Nike shirt rather than a Reebok shirt. Similarly, Serena Williams would be shown using a Hewlett-Packard computer rather than an Apple to recognize her exclusive agreement with the manufacturer, HP.266 Such a rule would not be absolute, however, and would allow for necessary editorial use by the publishers, including the ability to parody or satirize the endorsements.   (4) For athletes, collective bargaining agreements with teams and leagues would need to extend to the use of identity interests in computer games and virtual worlds. The publishers and producers would be required to recognize those agreements either by becoming signatories or by recognizing the third party beneficiary interests and obligations set forth in the players’ collective bargaining agreements.   (5) The enforcement of the publicity rights provisions would need to extend beyond the publisher or producer of the game or site to include end users. The publishers and producers would accomplish this through inclusion of the provisions in its end-user license agreement.267   (6) Rules governing the importation of content originally created for television, radio, billboards, and motion pictures should be developed. This will allow for better movement of advertising and other content to the games and virtual worlds. For example, a depiction of a drive-in theatre in a virtual world could be running movie trailers, movies, or commercials, and the availability of this use must be obtained 264   The difference between an outline of the terms and the specific terms is highlighted by this concept. While it may be intuitively easy to understand when a person is featured in an advertisement and when a person is merely part of the background, the language necessary to make that distinction may be highly technical and beyond the scope of this article. 265   Such a clause could range from a contractual requirement to a reasonable efforts obligation. A “best efforts” clause would be a reasonable starting point. 266   See Susan Dominus, Dangerous when Interested, N.Y. Times (Aug. 19, 2007), www.nytimes. com/2007/08/19/sports/playmagazine/0819play-serena.html; Kate Maddox, HP Serves up TV Spot, Advertising Age (Aug. 27, 2007), http://adage.com/article/btob/hp-serves-tv-spot/268173/. 267   See Andrew Jankowich, EULAw: The Complex Web of Corporate Rule-Making in Virtual Worlds,  8 Tul. J. Tech. & Intell. Prop. 1 (2006); Robert L. Oakley, Fairness in Electronic Contracting: Minimum Standards for Non-Negotiated Contracts, 42 Hous. L. Rev. 1041 (2005).

BARFIELD PRINT.indd 558

13/11/2018 11:53

Reordering the chaos of the virtual arena  559 from both the copyright holder and the performers. Billboards and print advertising should similarly be clarified.   (7) Compensation would be paid to the performers. The competition might be limited to the commercial exploitation only or could be based on other uses. Compensation might even be appropriate in situations such as PRO, where substantial revenues are being generated based on the union member’s identities notwithstanding the fact that those identities are used in an expressive manner. This reflects the distinction between the editorial control over the image and the economic advantage being taken by the use of the image. Such compensation deals, however, would be appropriate only where a very substantial taking of identity interests occurred.   (8) Credit for the performers would be provided and full disclosure of the relationship between the celebrities and the characters and avatars would make explicit the extent to which permission had been granted for the depictions in each game or virtual world.   (9) The terms of the agreement will apply to all divisions under the common ownership of the producers. The parties to the agreement will not enter into agreements with any third party that does not abide by a substantially similar agreement or an enforceable statement of best practices. (10) The producers will provide a notice and takedown mechanism for complaints of publicity rights infringement similar to that utilized for complaints of copyright infringement.268 The particular rule adopted under each of these topics does not necessarily matter. Different unions will likely seek different outcomes, at least initially. Actors, musicians, and athletes may each have slightly different goals for the agreements and these differing interests will result in slightly different resolutions. Parties will learn from each other as a consensus emerges. The rules provide a much more symbiotic relationship between the parties than any that could be derived through court decisions. 4.2  Expanding the Jurisdiction through Viral Distribution of the Agreements To be effective, these provisions must ultimately be fairly ubiquitous across the internet and among game publishers. Although content on the internet is not directly regulated, many policies derive from regulation,269 or from commonly adopted practices.270 A contract provision such as the one suggested in point 9 of the foregoing list, which requires the contracting party to enter the same clause with its other contracting partners, takes

268   See 17 U.S.C. § 512(c) (2007) (Limitations on liability relating to material online). See also 47 U.S.C. § 230 (2007) (Protection for private blocking and screening of offensive material). 269   See, e.g., 42 § U.S.C. § 1320d-2 (2007) (HIPAA regulations attaching data security obligation to any entity having control over the regulated health information). 270   See, e.g., voluntary ratings for motion pictures or video games. See Classification and Rating Administration, filmratings.com (last visited July 30, 2017); Ent. Software Rating Board, www.esrb.org (last visited Jan. 31, 2008). Professor Larry Lessig would suggest that the computer code utilized to standardize Internet sites has this de facto effect. Lawrence Lessig, Code and Other Laws of Cyberspace 6 (1999).

BARFIELD PRINT.indd 559

13/11/2018 11:53

560  Research handbook on the law of virtual and augmented reality on the behavior of a virus, “infecting” every contracting party that comes into contact with the source of the provision.271 These provisions provide that the parties to the first contract are a form of third party beneficiary to the subsequent agreements. Assuming these provisions are drafted in a manner consistent with antitrust laws,272 such provisions can quickly transform the common practices of an industry. A practice codified into collective bargaining agreements for those companies which are unionized and followed as best practice for those businesses which have no union agreements will be applicable for a major portion of the commercially significant activities on the internet and by game publishers.273 The technique for spreading new contractual expectations through viral provisions in contracts will not extend everywhere, but it should track most of the economically relevant activity. Through the combination of contractual and voluntary adoptions, voluntary self-regulation spread virally through the transactional web of the Internet. The plan to regulate publicity rights in gaming and virtual worlds through collective bargaining may appear flawed because the jurisdiction seems too limited on both sides. The Wild West of the internet is not home to extensive collective bargaining agreements. A closer look, however, suggests that the application of these viral contracting techniques creates both the ability to marshal collective bargaining power clout and the economic interests to do so. This is not a case of traditional labor organizing. The avatars will not be holding rallies, collecting cards, or counting votes. Instead, the willingness to enter into collective bargaining talks will stem from the leverage held by the players and performer unions. Today, most games and virtual worlds have versions playable on mobile devices or gaming consoles, if not both. This means users will acquire their games from either the Apple App Store or Google Play.274 A much smaller market exists for Amazon Fire users and for Microsoft mobile users.275 In a similar manner, the three top gaming console companies—Sony, Microsoft, and Nintendo—account for much of the console gaming market.276 Although there are many other sources for internet gaming, a common practice

271   This approach is similar to the effect of federal health privacy law to extend beyond health care organizations to include all organizations that come into contact with health care data. See Business Associates, 45 CFR 164.502(e) (providing that a business associate must protect health care information under contractual language provided by regulation). 272   Sherman Antitrust Act, 15 U.S.C. § 1 (2017). See Lawrence A. Cunningham, Language, Deals, and Standards: The Future of XML Contracts, 84 Wash. U. L. Rev. 313 (2006); Henry H. Perritt, Towards a Hybrid Regulatory Scheme for the Internet, 2001 U. Chi. Legal F. 215, 286–8 (2001). 273   The nonstatutory labor exemption from antitrust laws avoids Sherman Act complications when the parties are involved in good faith bargaining regarding topics such as wages, hours, and working conditions. Brown v. PRO-Football, Inc., 518 U.S. 231, 235 (1996); Connell Constr. Co. v. Plumbers, 421 U.S. 616, 622 (1975). While the antitrust implications are significant, they are not likely to interfere with a broadly adopted set of policies, particularly if the policies were generally consistent with and clarifying of existing law. Such practices would be reasonable and efficient, benefiting competition and trade, and therefore upheld under either the exemption or under a rule of reason analysis. See id. 274   Apple, www.appstore.com/; Google Play, https://play.google.com/store/. 275   Amazon, www.amazon.com/Fire-TV-Apps-All-Models/; Microsoft, www.microsoft.com/ en-us/store/apps/windows. 276   John Gaudiosi, How Sony, Microsoft, and Nintendo Are Preparing for Their Gaming Future, Fortune (Dec, 23, 2016), http://fortune.com/2016/12/23/sony-microsoft-nintendo-future/.

BARFIELD PRINT.indd 560

13/11/2018 11:53

Reordering the chaos of the virtual arena  561 among these six vendors driven by collective bargaining would transform the rules for content licensing. In addition, GE, News Corp, Disney, Viacom, Time/Warner, and CBS, the six largest media conglomerates, together own 90 percent of American media.277 Twelve companies are not too many to bring to the table. Under the viral approach to enforcing the collective bargaining agreements, the performers’ unions initially negotiate for the relevant provision in collective bargaining agreements with the producers presently under contract. The ninth provision, listed above, requires that the members of the producers or league enforce the agreement by requiring the provisions to be adopted by their contracting partners. In the first instance, this would require that the entirety of a commonly owned enterprise be subject to the rules. The rules would become the “law” everywhere within the 12 gatekeeper companies. Broadcasters would feel the pressure from the actors’ union, the players’ union, and the leagues. There would naturally be some resistance to the proposal because payments are required. But each of these companies also owns large portfolios of brands. Some of these brands are promoted through endorsement, and they make significant revenue from other advertisers that would prefer a better way to protect the investment in the performers who endorse their products. Properly managed, a campaign to win over the advertisers and the leagues would help enable the celebrities to negotiate effectively. The negotiations will also be more successful if the balance sought is reasonable and consistent with the historical copyright fair use and free expression approach to balancing ownership and expression. Publishers will appropriately fight attempts to frustrate biographies or unauthorized uses of identities. Producers will defend the right to produce parodies of celebrities on television and on the internet. The rules outlined for the collective bargaining agreement must not impinge on these acts of commentary, criticism, and public discourse. Once the provisions are incorporated into existing collective bargaining agreements and applied to the enterprises in which those bargaining units exist, the obligation to recognize these rights will spread quickly throughout much of the industry. As a result of contractual relations among the major content producers and advertisers, many of the sources of advertising will be contractually obligated to follow the rules. The jurisdiction will extend slowly throughout the internet, computer gaming, and virtual world environment to provide protection for union members and nonunion members alike.278 In this way, it will reach college athletes and nonunion musicians as well as the professional celebrities who are notorious rather than talented. Although unions are often reluctant to extend union benefits beyond their members, such protection is in the union’s economic best interest because it precludes the signatory from reducing cost by approaching nonunion performers for their rights.

277   Michael Corcoran, Twenty Years of Media Consolidation Has Not Been Good for Our Democracy, Moyers & Company (Mar. 30, 2016), http://billmoyers.com/story/twenty-years-ofmedia-consolidation-has-not-been-good-for-our-democracy/; Ashley Lutz, These 6 Corporations Control 90% of the Media in America, Bus. Insider (Jun. 14, 2012), www.businessinsider.com/ these-6-corporations-control-90-of-the-media-in-america-2012-6. 278   See proposal no 1 above.

BARFIELD PRINT.indd 561

13/11/2018 11:53

562  Research handbook on the law of virtual and augmented reality 4.3  Best Practices for Everyone Else There are limits to the model of viral contracting as a method of revising the normative behavior on the internet. First, the collective bargaining and contract negotiating process intentionally spreads the obligation somewhat slowly through the media network. Second, the reach does not extend to publishers like PRO, which are user created or outside traditional media. Both of these limits are benefits of the approach. For the rules of publicity to gain widespread acceptance, the rules need to become part of the normative balance of rights expected by the public. Copyright, on one hand,279 and trademark, on the other,280 allow the scope of enforcement to be based on the behaviors and expectations of the public. While not always effective,281 law in general—intellectual property law in particular—benefits tremendously from having it reflect accepted behavior rather than trying to preclude conduct the public does not generally find inappropriate. Particularly within the context of virtual worlds, there needs to be a great deal of experimentation before the genre matures. Some of these experiments will be highly popular and successful; most will go unnoticed. Owners of intellectual property rights will waste valuable resources if they futilely try to chase every act of infringement. They will also stymie the very innovation necessary for the next stage in growth for immersive and interactive media. The use of a viral licensing regime helps align the economic interests of the parties and leaves those parties with few connections generally out of the private ordering. As an unlicensed project gains momentum, it is likely to need contractual relations with an entity under the licensing regime. At that point, the publishers can choose to join the regime or to stop growing in order to avoid those contractual entanglements. The best practices are even simpler than the provisions needed in the collective bargaining agreement because the duties are self-imposed, not requiring any provisions for enforcement outside of the virtual world under operation: (1)  The right of publicity would have to be recognized by the virtual world. (2) The enforcement of the publicity rights provisions would need to extend beyond the publisher or producer of the game or site to include the end users. (3) The definition of commercial activity would be specified to focus on the sale, marketing, or endorsement of any product or service offered to the public. It would exclude all expressive, literary, or editorial content, and provide limited use of unlicensed identity interest for advertising or marketing of the work in which the celebrity appeared as a performer or athlete. (4) The virtual world and its members would be obligated to make efforts to avoid interfering with exploitation of publicity rights by celebrities with products or services 279   See Jon M. Garon, Normative Copyright: A Conceptual Framework for Copyright Philosophy and Ethics, 88 Cornell L. Rev. 1278 (2003) (emphasizing public’s acceptance of practices establish fair use). 280   See Two Pesos, Inc. v. Taco Cabana, Inc., 505 U.S. 763, 780 (1992). 281   See, e.g., Metro-Goldwyn-Mayer Studios, Inc. v. Grokster, Ltd., 545 U.S. 913 (2005); A&M Records, Inc. v. Napster, Inc., 239 F.3d 1004 (9th Cir. 2001).

BARFIELD PRINT.indd 562

13/11/2018 11:53

Reordering the chaos of the virtual arena  563 which were in competition with the products and services for which the celebrity had existing exclusive arrangements. (5) Use of publicity rights for commercial activities can only be conducted with the permission of the celebrity. (6) Credit for the performers would be provided and full disclosure of the relationship between the celebrities and the characters and avatars would make explicit the extent to which permission had been granted for the depictions in each game or virtual world. (7) The producers will provide a notice and takedown mechanism for complaints of publicity rights infringement, similar to that utilized for complaints of copyright infringement. These provisions are very similar to the provisions applied to collective bargaining agreements, and should be relatively easy for operators of virtual worlds to add to end-user license agreements and put in place as practices governing their own conduct. Over time, many of the entities not otherwise contractually obligated to accept the provisions will choose to adopt them as best practices. If the policies are well balanced and take into account both the social desire for robust expression and the social desire for accurate, fully disclosed commercial endorsements, then publishers of websites will recognize that the burdens of the provisions are very minor and the benefits to the site’s customers are significant.

5. CONCLUSION New rules for publicity rights in computer games and virtual worlds are best developed through creation of a best practices standard and a series of collective bargaining agreements that slowly pervade the online and gaming worlds. These rules can provide robust protection for celebrities when their identity is hijacked to sell or endorse products, while promoting wide latitude for writers, artists, and publishers to use such identities when telling stories about the celebrities or merely using celebrities as the raw material for other expressive works. Private ordering will allow for this newest of intellectual property rights to mature, and would avoid the widespread confusion created by the multiple tests proliferating among the courts and the somewhat inconsistent application of the publicity rights doctrine alongside copyright and trademark law. Given the importance of virtual worlds in the future social fabric, a solution to the identity rights chaos is necessary. Already protected by constitutional law, the content of these worlds reflects an important new platform for public discourse, as ideas and mores are explored within these environments. At the same time, these worlds are also sites of commercial transactions, advertising, marketing, and endorsements that take advantage of performers, athletes, and other celebrities to sell their wares. Celebrities are likely to embrace this new opportunity only if their intellectual property interests can be protected. Through the combination of collective bargaining agreements and best practices, the problems of court interpretation can be avoided, First Amendment challenges averted, and legitimate rights protected. Let the games continue.

BARFIELD PRINT.indd 563

13/11/2018 11:53

564  Research handbook on the law of virtual and augmented reality

SOURCES Cases A&M Records, Inc. v. Napster, Inc., 239 F.3d 1004 (9th Cir. 2001). Abdul-Jabbar v. Gen. Motors Corp., 85 F.3d 407, 416 (9th Cir. 1996). ACLU v. Gonzales, 478 F. Supp. 2d 775 (D. Pa. 2007). Allison v. Vintage Sports Plaques, 136 F.3d 1443 (11th Cir. 1998). Am. Amusement Mach. Ass’n v. Kendrick, 244 F.3d 572 (7th Cir. 2001). Apple Comput., Inc. v. Franklin Comput. Corp., 714 F.2d 1240 (3rd Cir. 1983). Ashcroft v. ACLU, 542 U.S. 656 (U.S. 2004). Bragg v. Linden Research, Inc., 487 F. Supp. 2d 593, 595 (E.D. Pa. 2007). Brother Records, Inc. v. Jardine, 318 F.3d 900, 905–6 (9th Cir. 2003). Brown v. Entm’t Merch. Ass’n, 131 S. Ct. 2729 (2011). Brown v. PRO-Football, Inc., 518 U.S. 231, 235 (1996). C.B.C. Distrib. & Mktg., Inc. v. Major League Baseball .Advanced Media, L.P., 443 F. Supp. 2d 1077, 1084 (D. Mo. 2006). Cairns v. Franklin Mint Co., 292 F.3d 1139, 1152 (9th Cir. 2002). Campbell v. Acuff-Rose Music, Inc., 510 U.S. 569 (1994). Cardtoons, L.C. v. Major League Baseball Players Ass’n, 95 F.3d 959, 966 (10th Cir. 1996). Carson v. Here’s Johnny Portable Toilets, Inc., 698 F.2d 831, 835–6 (6th Cir. 1983). Cent. Hudson Gas & Elec. Corp. v. Pub. Serv. Comm’n, 447 U.S. 557, 563 (1980). Century 21 Real Estate Corp. v. Lendingtree, Inc., 425 F.3d 211, 229 (3rd Cir. 2005). City of Los Angeles v. Alameda Books, Inc., 535 U.S. 425 (2002). City of Renton v. Playtime Theatres, Inc., 475 U.S. 41 (1986). Cliff Notes, Inc. v. Bantam Doubleday Dell Publ’g Grp., 886 F.2d 490, 495 (2nd Cir. 1989). Cohen v. California, 403 U.S. 15, 26 (1971). Comedy III Prods., Inc. v. Gary Saderup, Inc., 21 P.3d 797 (Cal. 2001). Comput. Assocs. Int’l Inc. v. Altai, Inc., 982 F.2d 693, 703 (2nd Cir. 1992). Connell Constr. Co. v. Plumbers, 421 U.S. 616, 622 (1975). Dall. Cowboys Cheerleaders, Inc. v. Pussycat Cinema, Ltd., 604 F.2d 200, 206 (2d Cir.1979). Doe v. TCI Cablevision, 110 S.W.3d 363, 369 (Mo. 2003). Downing v. Abercrombie & Fitch, 265 F.3d 994, 1009 (9th Cir. 2001). Dr. Seuss Enters., L.P. v. Penguin Books 109 F.3d 1394 (9th Cir. 1997). E.S.S. Entm’t 2000, Inc. v. Rock Star Videos, Inc., 444 F. Supp. 2d 1012, 1027 (D. Cal. 2006). Eldred v. Ashcroft, 537 U.S. 186, 219 (U.S. 2003). Entm’t Software Ass’n v. Blagojevich, 404 F. Supp. 2d 1051 (E.D. Ill. 2005). Entm’t Software Ass’n v. Foti, 451 F. Supp. 2d 823 (M.D. La. 2006). Entm’t Software Ass’n v. Granholm, 426 F. Supp. 2d 646 (E.D. Mich. 2006). Entm’t Software Ass’n v. Hatch, 443 F. Supp. 2d 1065 (D. Minn. 2006). ETW Corp. v. Jireh Publ’g, Inc., 332 F.3d 915 (6th Cir. 2003). ETW Corp. v. Jireh Publ’g, Inc., 332 F.3d 915, 928 (6th Cir. 2003). FCC v. Pacifica Found., 438 U.S. 726, 749–50 (1978). Feist Publ’ns, Inc. v. Rural Tel. Serv. Co., 499 U.S. 340, 349–350 (1991). Fleer Corp. v. Topps Chewing Gum, 658 F.2d 139, 142 (3d Cir. 1981). Fleet v. CBS, 50 Cal. App. 4th 1911 (Cal. Ct. App. 1996). Gionfriddo v. Major League Baseball, 114 Cal. Rptr. 2d 307 (Ct. App. 2001). Gitlow v. New York, 268 U.S. 652 (1925). Golan v. Holder, 565 U.S. 302 (2012). Grosjean v. Am. Press Co., 297 U.S. 233 (1936). Guglielmi v. Spelling-Goldberg Prods., 25 Cal. 3d 860 (1979). Haelan Labs., Inc., v. Topps Chewing Gum, Inc., 202 F.2d 866, 868 (2nd Cir. 1953). Harper & Row, Publishers v. Nation Enters., 723 F.2d 195, 203 (1983). Harper & Row, Publishers, Inc. v. Nation Enters., 471 U.S. 539 (1985). Hart v. Elec. Arts, Inc., 717 F.3d 141 (3rd Cir. 2013). Hilton v. Hallmark Cards, 599 F.3d 894, 912 (9th Cir. 2010). Hoffman v. Capital Cities/ABC, Inc., 255 F.3d 1180 (9th Cir. 2001). Horphag Research Ltd. v. Pellegrini, 337 F.3d 1036, 1040–1 (9th Cir. 2003). Hustler Magazine, Inc. v. Falwell, 485 U.S. 46, 52 (1988). In re Dual-Deck Video Cassette Recorder Antitrust Litig., 11 F.3d 1460, 1467 (9th Cir. 1993).

BARFIELD PRINT.indd 564

13/11/2018 11:53

Reordering the chaos of the virtual arena  565 In re NCAA Student-Athlete Name & Likeness Licensing Litig., 724 F.3d 1268 (9th Cir. 2013). In re R.M.J., 455 U.S. 191, 203 (1982). Interactive Dig. Software Ass’n v. St. Louis Cty., 329 F.3d 954 (8th Cir. 2003). Interstate Circuit, Inc. v. City of Dallas, 390 U.S. 676, 690 (1968). Joseph Burstyn, Inc. v. Wilson, Comm’r of Educ. of N.Y., 343 U.S. 495, 502 (1952). Keller v. Elec. Arts, Inc., 2010 WL 445190 (Feb. 8, 2010). Kirby v. Sega of Am., Inc., 144 Cal. App. 4th 47, 50 (2006). KP Permanent Make-Up, Inc. v. Lasting Impression I, Inc., 543 U.S. 111, 117 (2004). L.A. News Serv. v. CBS Broad., Inc., 305 F.3d 924 (9th Cir. 2002). L.A. News Serv. v. KCAL-TV Channel 9, 108 F.3d 1119, 1122 (9th Cir. 1997). L.L. Bean, Inc. v. Drake Publishers, Inc., 811 F.2d 26, 29 (1st Cir. 1987). Lloyd Corp. v. Tanner, 407 U.S. 551 (1972). Lugosi v. Universal Pictures, 603 P.2d 425 (Cal. 1979). Madsen v. Women’s Health Ctr., Inc., 512 U.S. 753 (1994). Mattel, Inc. v. MCA Records, 296 F.3d 894 (9th Cir. Cal. 2002). MCA, Inc. v. Wilson, 677 F.2d 180, 200 (2nd Cir. 1981). Metro-Goldwyn-Mayer Studios, Inc. v. Grokster, Ltd., 545 U.S. 913 (2005). Midler v. Ford Motor Co., 849 F.2d 460, (9th Cir. 1988). Midway Mfg. Co. v. Dirkschneider, 543 F. Supp. 466, 479 (D. Neb.1981). Montana v. San Jose Mercury News, Inc., 40 Cal. Rptr. 2d 639, 640 (Ct. App. 1995). Mut. Film Corp. v. Indus. Comm’n of Ohio, 236 U.S. 230 (1915). Mut. of Omaha Ins. Co. v. Novak, 836 F.2d 397, 402–3 (8th Cir. 1987). N.Y. Times Co. v. Sullivan, 376 U.S. 254, 280 (1964). N.Y. Times Co. v. United States, 403 U.S. 713 (1971). Nemani v. St. Louis Univ., 33 S.W.3d 184, 185 (Mo. 2000) (en banc). New Era Publ’ns Int’l v. Henry Holt & Co., 873 F.2d 576 (2nd Cir. 1989). New Era Publ’ns Int’l, ApS v. Henry Holt and Co., 695 F. Supp. 1493, 1528 (S.D.N.Y. 1988). New Kids on the Block v. News Am. Publ’g, Inc., 971 F.2d 302 (9th Cir. 1992). O’Bannon v. Nat’l Collegiate Athletic Ass’n, 802 F.3d 1049, 1055 (9th Cir. 2015). O’Bannon, Jr. v. National Collegiate Athletic Ass’n, et. al., Order Granting Final Approval of Class Action Settlement and Final Judgment, Case Nos. 4:09-CV-1067 CW; 4:09-CV-3329 CW 1 (N.D. Cal., July 16, 2015). Pavesich v. New England Life Ins. Co., 50 S.E. 68 (Ga. 1905). Peel v. Attorney Disciplinary Comm’n, 496 U.S. 91 (1990). Pirone v. MacMillan, Inc., 894 F.2d 579 (2nd Cir. 1990). Roberson v. Rochester Folding Box Co., 64 N.E. 442 (N.Y. 1902). Rogers v. Grimaldi, 875 F.2d 994 (2nd Cir. 1989). Rosemont Enters., Inc. v. Random House, Inc., 366 F.2d 303 (2nd Cir. 1966). Salinger v. Random House, Inc., 811 F.2d 90 (2nd Cir. 1987). Sherwood 48 Assocs. v. Sony Corp. of Am., 213 F. Supp. 2d 376, 377 (S.D.N.Y. 2002), aff’d, 76 Fed. Appx. 389, 2003 U.S. App. LEXIS 20106 (2003). Snyder v. Phelps, 562 U.S. 443 (2011). Spahn v. Julian Messner, Inc., 221 N.E.2d 543 (N.Y. 1966). Stephano v. News Grp. Publ’ns, Inc., 474 N.E.2d 580, 584–5 (N.Y. 1984). Stewart v. Rolling Stone LLC, 105 Cal. Rptr. 3d 98, 104 (Ct. App. 2010). Suntrust Bank v. Houghton Mifflin Co., 268 F.3d 1257, 1264 (11th Cir. 2001). Thomas v. Collins, 323 U.S. 516, 531 (1945). Time, Inc. v. Bernard Geis Assocs., 293 F. Supp. 130 (S.D.N.Y. 1968). Time, Inc. v. Hill, 385 U.S. 374 (1967). Toney v. L’Oreal, 406 F.3d 905 (7th Cir. 2004). Two Pesos, Inc. v. Taco Cabana, Inc., 505 U.S. 763, 780 (1992). Tyne v. Time Warner Entm’t Co., 901 So. 2d 802 (Fla. 2005). Uhlaender v. Henricksen, 316 F. Supp. 1277, 1278 (D. Minn. 1970). United States v. Stevens, 559 U.S. 460, ___, 130 S. Ct. 1577, 1582–3, 176 L.Ed.2d 435 (2010). Va. State Bd. of Pharmacy v. Va. Citizens Consumer Council, Inc., 425 U.S. 748, 771 (1976). Video Software Dealers Ass’n v. Maleng, 325 F. Supp. 2d 1180 (W.D. Wash. 2004). Video Software Dealers Ass’n v. Schwarzenegger, No. C-05-04188 (RMW), 2007 U.S. Dist. LEXIS 57472 at 25 (N.D. Cal. Aug. 6, 2007). Video Software Dealers Ass’n v. Webster, 968 F.2d 684, 688 (1992). Volkswagenwerk Aktiengesellschaft v. Church, 411 F.2d 350 (9th Cir. 1969). Waits v. Frito-Lay, Inc., 978 F.2d 1093, 1108 (9th Cir. 1992). Welch v. Carson Prods. Grp., Ltd., 791 F.2d 13 (2nd Cir. 1986).

BARFIELD PRINT.indd 565

13/11/2018 11:53

566  Research handbook on the law of virtual and augmented reality Wendt v. Host Int’l, Inc., 125 F.3d 806 (9th Cir. 1997). Wendt v. Host Int’l, Inc., 125 F.3d 806, 811 (9th Cir. 1997). White v. Samsung Elecs. Am., Inc., 971 F.2d 1395, 1399 (9th Cir. 1992). White v. Samsung Elecs. Am., Inc., 989 F.2d 1512 (9th Cir. 1993). Williams Elecs., Inc. v. Artic Int’l, Inc., 685 F.2d 870 (3d Cir. 1982). Wilson v. Midway Games, Inc., 198 F. Supp. 2d 167, 181 (D.C. Conn. 2002). Winter v. DC Comics, 30 Cal. 4th 881 (Cal. 2003). Winters v. New York, 333 U.S. 507, 510 (1948). Wojnarowicz v. Am. Family Ass’n, 745 F. Supp. 130, 142 (D.N.Y. 1990). Yankee Publ’g, Inc. v. News Am. Publ’g, Inc., 809 F. Supp. 267, 276 (S.D.N.Y. 1992). Young v. Am. Mini Theatres, Inc., 427 U.S. 50 (1976). Zacchini v. Scripps-Howard Broad. Co., 433 U.S. 562, 576 (1977).

Statutes Cal. Civ. Code § 3344 (2007). Cal. Civ. Code § 3344 (2007). Can-Spam Act of 2003, Pub. L. No 108–87, 117 Stat. 2699 (2003) (codified at 15 U.S.C. §§ 7701–13 and 18 U.S.C. § 1037). Children’s Online Privacy Protection Act of 1998, 15 U.S.C. §§ 6501-03 (2007); Stats 2007 ch. 439 § 1 (S.B. 771) (2008). H.R. Rep. No 94-1476, at 57 (1976). La. Stat. Ann. § 14:91.14 (2015). N.Y. Civ. Rights Law §§ 50–51 (McKinney 2002). Ohio Const. art. I, § 11. Sherman Antitrust Act, 15 U.S.C. § 1 (2017). Telecommunications Act of 1996, Pub. L. No. 104-104, 110 Stat. 56 (1996); 47 U.S.C. 303(w) (2007). 15 U.S.C. § 1051. 15 U.S.C. § 1115(b)(4). 15 U.S.C. §1125(a). 15 U.S.C. § 1125(d). 15 U.S.C. §§ 7701–13. 16 C.F.R. § 255.0(a) (2016). 17 U.S.C. § 102. 17 U.S.C. § 102(b). 17 U.S.C. § 107 (2000). 17 U.S.C. § 107(4). 17 U.S.C. § 512(c) (2007). 18 U.S.C. § 1037. 42 U.S.C. § 1320d-2 (2007). 47 U.S.C. § 303(a) (2007). 47 U.S.C. § 317 (2007). 47 U.S.C. § 317 (2007).

CFRs FTC Guides Concerning the Use of Endorsements and Testimonials in Advertising, 74 Fed. Reg. 53,124, 53,125 (Oct. 15, 2009) (codified at 16 C.F.R. pt. 255). 16 C.F.R. § 255.0(a) (2016). 16 C.F.R. § 255.1(a) (2016).

County Ordinances Indianapolis and Marion County, Ind., Ordinance 72 (July 10, 2000). St. Louis County, Mo., Ordinance 20,193 (Oct. 26, 2000).

BARFIELD PRINT.indd 566

13/11/2018 11:53

PART V CONTRACT, PROPERTY LAW, AND JURISDICTION

BARFIELD PRINT.indd 567

13/11/2018 11:53

17.  Property rights in virtual and augmented reality: Second Life versus Pokémon Go Hannah YeeFen Lim

1. INTRODUCTION Video gaming has undoubtedly been revolutionised by the internet. Single player games no longer seem to be as attractive as dynamic, networked games, such as virtual worlds. A number of nonlevel-based, social interaction-based, three-dimensional virtual worlds have become hugely popular in recent decades. Two of the better known in the Western world have been Second Life, operated by Linden Research (‘Linden’), and There.com, operated by Makena Technologies. Aside from the virtual-world games, in 2016 the augmented reality game Pokémon Go was launched worldwide and reached unprecedented levels of popularity. Being an augmented reality game, which necessarily intrudes upon the real physical world, the game has raised a number of serious legal issues – certainly more than virtual-world games. This chapter will focus on the rights related to land arising from both the virtual world Second Life and the augmented reality game Pokémon Go. For Pokémon Go, it will analyse the various rights related to land from the perspective of a number of different parties, namely the game developer, the players and the real world land or property owners/occupiers. For the virtual world Second Life, it will in particular expound the legal concept of land in Second Life, which found itself in the international legal limelight because of the Bragg litigation against Linden in the United States.1 The Bragg litigation raised squarely, for the first time in a Western court of law, the issue of the legal status of items in virtual worlds and whether these virtual items can indeed be recognised as property under the Western legal tradition. This is an issue separate from and independent of the question of intellectual property protection, which has already been discussed elsewhere.2 There has been much dispute concerning whether virtual items can be considered property. If they are not considered property, then players have very little control over their creations vis-à-vis the virtual world provider and against other players; in effect, they have a mere contractual licence. If virtual items are property, then the legal landscape changes dramatically, as will be discussed in section 7 of this chapter. It can be said that at the heart of the issue is the End-User Licence Agreement (‘EULA’). Before a player commences using the virtual world, the player must agree to the terms of the EULA, which dictates the conditions for playing the game.3 A game provider ‘can

  Bragg v. Linden Research, Inc., 487 F. Supp. 2d 593 (E.D. Penn. 2007) [Bragg].   Yee Fen Lim, ‘Is It Really Just a Game? Copyright and Online Role-Playing Games’ (2006) 1 J Intellectual Property Law & Practice 481. 3   Mia Garlick, ‘Player, Pirate or Conducer? A Consideration of the Rights of Online Gamers’ (2004–5) 7 Yale Journal of Law and Technology 422, online: www.yjolt.org/7/spring/garlick-422. 1 2

568

BARFIELD PRINT.indd 568

13/11/2018 11:53

Property rights in virtual and augmented reality  569 discipline players who violate the EULA, take away their privileges and powers, or even kick them out of the game space and eliminate their avatars.’4 While EULAs may provide consumers nearly everything they need to know by setting out the terms of the relationship between the player and the game provider,5 certain terms of these standard contracts, found among the billing arrangements and software requirements, often rob players of a number of rights – most often, the intellectual property rights in any characters or items they create while playing the game. The legal problems essentially begin here. Online gamers feel an almost passionate sense of ‘entitlement’ to ownership of the unique characters and items they create while playing the game.6 Most virtual world providers feel the opposite: almost always, the EULA states that the intellectual property rights of gamers vest with the game ­providers, depriving the player of any rights over the potential use of a character or item.7 A more crucial layer on top of the intellectual property issue is the status of the virtual items, that is, whether they can be classified as property, an issue which has been further fuelled by a lucrative online market for the sale of game characters and items. The value of this market was estimated in 2005 at approximately US$880 million.8 Whether sold on eBay or sold by private transactions, game items can fetch a fortune. In December 2004, University of Sydney graduate David Storey purchased an online island, complete with beaches and an abandoned castle, for AUD$35,000.9 In March 2005, a Chinese gamer sold a ‘dragon sabre’ sword in the game Legends of Mir 3 for AUD$1,129. Other items have sold for as much as US$3,000.10 On eBay, a purchaser can buy a magic sword from the game Dark Age of Camelot for $100.11 After sending 100 real-world dollars to the seller, the purchaser will then meet the seller inside the game so the seller can hand over the magic sword.12 Inside the game, the purchaser can now defeat a dragon; outside the game the seller is $100 richer. This is essentially the sale of something independent from the intellectual property. The closest analogy would be the purchase of a book; when one buys a book, one buys the physical thing and not any intellectual property rights. Similarly, when one buys a virtual castle, one buys the castle in its virtual form and not necessarily any intellectual property rights.

 4   Jack M. Balkin, ‘Institute for Information Law and Policy Symposium State of Play: I. Essay & Reflection: Law and Liberty in Virtual Worlds’ (2004/2005) 49 N.Y.L. Sch. L. Rev. 63 at 65.  5   Daniel C. Miller, ‘Note: Determining Ownership in Virtual Worlds: Copyright and Licence Agreements’ (2003) 22 Rev. Litig. 435 at 461.  6   Supra note 3.  7   Andrew Jankowich, ‘EULAw: The Complex Web of Corporate Rule-Making in Virtual Worlds’ (2006) 8 Tulane Journal of Technology & Intellectual Property 1.  8   Edward Castronova, ‘Secondary Markets: $880 Million’ Terra Nova (30 October 2004), online: http://terranova.blogs.com/terra_nova/2004/10/secondary_marke.html.  9   Amalie Finlayson and Reuters, ‘Online Gamer Killed for Selling Virtual Weapon’ The Sydney Morning Herald (30 March 2005), online: www.smh.com.au/news/World/Online-gamerkilled-for-selling-virtual-weapon/2005/03/30/1111862440188.html. 10   Molly Stephens, ‘Note: Sales of In-Game Assets: An Illustration of the Continuing Failure of Intellectual Property Law to Protect Digital-Content Creators’ (2002) 80 Tex. L. Rev. 1513 at 1515. 11   Listings of internet gaming characters, items, and other paraphernalia may be viewed on eBay. 12   Supra note 3.

BARFIELD PRINT.indd 569

13/11/2018 11:53

570  Research handbook on the law of virtual and augmented reality Game companies have been quick to ban the sale of virtual items. While most EULAs state that players do not have the right to resell any items they create in the game, in most instances game companies have purported to rely on intellectual property law, in particular copyright, to stop such sales. As will be discussed, several companies have threatened sites, including Yahoo! and eBay, with legal action, claiming that this practice infringes the intellectual property rights of the game provider. Players have complained, however, that it is not intellectual property being sold, but the fruits of the time the players have invested in these games, and in essence the ‘property’ in their creations. It will have become apparent to the reader that there are numerous legal issues arising in this area. This chapter will focus on the property issues. The best analogy is that of a book. As already mentioned, in a physical book, multiple layers and forms of ownership of different types of property can coexist – the physical property in the book and the intangible copyright or copyrights in the book, for instance.13 The same can be argued for virtual items, in that there is property in the virtual item separate from the intellectual property of the virtual item. From a legal point of view, it is unfortunate that the dispute between Bragg and Linden was settled in early October 2007. The hearing was originally set for December 2007, and the decision was destined to produce important and interesting precedents in the area of virtual property. Section 2 of this chapter elaborates on the nature of virtual worlds and section 3 examines Second Life and its thriving synthetic economy. Section 4 examines the policy on property in Second Life and its interaction with property law and intellectual property law. Section 5 expounds on the meaning of land ownership in Second Life under the current conception of property law and concludes that proprietary interests do subsist in land ownership in Second Life. Section 6 argues that the bundle of property rights that arises from owning land in Second Life is very much like the property rights associated with holding a modified form of leasehold property. The analysis will draw upon the leasehold in public housing in Singapore as well as grazing licences in Australia. Property is a legal construct and in the real world the law has recognised property in many guises, from the tangible to the intangible, such as shares and leases. Land ownership in Second Life can and should be regarded as another form of property that gives rise to proprietary rights. Section 7 will discuss the practical significance of this analysis. Sections 8 to 10 will focus on the augmented reality game Pokémon Go and the proprietary rights implicated by the augmented reality game.

2.  VIRTUAL WORLDS Massively multiplayer online role-playing games (‘MMORPGs’), or, simply, online games, are a craze that has swept the globe. These online games provide virtual worlds that are as rich as the imagination allows. Being powered on the internet and with players from countries around the world, the games never stop. A player can go to bed at night and find

13   There can be multiple copyrights in a book such as the copyright in the text and the copyright in the layout of the published editions of books; see, e.g., the Copyright Act 1968 (Cth.), s. 92 (Australia).

BARFIELD PRINT.indd 570

13/11/2018 11:53

Property rights in virtual and augmented reality  571 an online kingdom’s ruler has changed by morning. In 2005, it was found that on average, online gamers spent approximately 22 hours per week online playing their favourite games.14 There have been reports of players spending up to 55 hours at a time playing,15 and in some Asian countries, where these games are enormously popular, laws have been introduced designed to limit the number of hours spent playing them on the internet.16 Online games allow users to create items and visual characters while playing the game. A player who subscribes to a game must first create an ‘avatar’ – a character that can have whatever features one chooses. Characters and other in-game items can take two forms: either text-based, literary characters or visual, animated, two or three-dimensional figures.17 The character does not even need to resemble a human being. While playing the game, users can battle dragons, escape dungeons and rescue princesses (or handsome princes), acquiring abilities, swords, shields and coins (which can also be created). While some online games emphasise problem-solving and adventuring akin to the fantasy tabletop roleplaying game Dungeons & Dragons (first published in 1974), others are less task-oriented and focus more on interpersonal relationships and goals less competitive than slaying beasts and gaining power. These types of online social worlds are not competitive by design in that players do not need to ‘level up’, that is, there is no need to defeat obstacles to increase the avatar’s power from a weak level 1 avatar to a more powerful level 2 avatar and so on. Instead, the online game allows for social interaction, letting users communicate or otherwise interact with each other. There are of course also hybrids such as EVE Online, where players aim to acquire power and treasure but player interaction is paramount, through combinations of trade, resource collection and/or theft and fraud. The heightened interactivity between the player and the game has no doubt been one of the major attractions of this phenomenon – and has arguably created the legal controversies surrounding online games. In order to play an online game, a user must first either purchase a CD or download the game software from the website of the game provider for free or for a flat fee.18 A monthly subscription fee is usually payable to access the game. The downloaded software allows the player to connect to the game server and play the game.

3.  SECOND LIFE AND ITS ECONOMY Second Life is a noncompetitive, social, three-dimensional virtual world for which Linden has deliberately recreated real-world experiences such as wind, oceans, mountains, islands, grass and buildings so that players would feel as if they were actually living in some sort

14   Tom Loftus, ‘Virtual Worlds Wind Up in Real World’s Courts’ MSNBC.com (7 February 2005), online: www.msnbc.msn.com/id/6870901/. 15   ‘S Korean Dies after Games Session’ BBC News (10 August 2005), online: http://news.bbc. co.uk/2/hi/technology/4137782.stm. 16   ‘Time Limits for Online Games Meant to Protect Children from Addiction’ Shanghai Daily (21 October 2005), online: www.shanghaidaily.com/. 17   Supra note 5 at 449. 18   Mathias Klang, ‘Avatars: From Deity to Corporate Property – A Philosophical Inquiry into Digital Property in Online Games’ (2004) 7(3) Information, Communication & Society 389 at 390.

BARFIELD PRINT.indd 571

13/11/2018 11:53

572  Research handbook on the law of virtual and augmented reality of virtual reality. But unlike the real world, players may also encounter vampires, talking animals, flying humans and anything else the players and Linden fancy. To participate in Second Life, the user first registers at no cost on Second Life’s homepage and creates an avatar. After downloading and installing the software, one is ready to begin living in Second Life. A player can simply live in Second Life and interact with other players in the online community without touching the Second Life economy. However, Second Life was designed to be more than an interactive chat facility for conversations, and indeed Linden explicitly rejects calling Second Life a ‘game’.19 According to the Second Life website, Second Life has a ‘fully integrated economy architected to reward risk, innovation and craftsmanship’.20 On the same website, it boasts to the world, ‘Make real money in a virtual world. That’s right, real money.’21 It also advises that businesses succeed in Second Life by the ‘ingenuity, artistic ability, entrepreneurial acumen, and good reputation of their owners’.22 The virtual currency in Second Life is ‘Linden Dollars’ (or L$), which can be purchased with real-world US dollars. The exchange rate in 2016 was around US$1 for L$250 and this rate has been relatively stable since 2007. One can also earn Linden Dollars by making and selling goods and services, holding events, and similar activities in Second Life. In the period prior to the Bragg litigation, if one upgraded to a Premium account, which cost the player US$9.95 or more per month, then Linden provided a small weekly stipend.23 Back them, only Premium account players could own land in Second Life. Many items in Second Life, such as clothes, cars and castles, have been created by its players using scripting tools and other design programs. Linden allows players to create, buy, sell and otherwise trade any product or service both within Second Life and in the real world. Some residents of Second Life have made significant profits from the sale and exploitation of virtual items and virtual land.24 These residents often spend as much time online in Second Life as they would working a full-time job, to the point that many view their investment in Second Life as their livelihood.25 A case in point is Anshe Chung (the creator of which is Ailin Graef), who in late 2006 became the first online avatar to achieve a net worth exceeding one million US dollars from profits entirely earned inside a virtual world.26 Indeed, it was through virtual real estate that Anshe Chung made her fortune. Within Second Life, Linden sells parcels of land through online auctions where registered users bid against other users for parcels of land. Once the parcels of land   Kenneth James, ‘Real Benefits in Virtual Worlds’ Business Times (11 December 2006).   ‘The Marketplace’, online: Second Life https://secure-web6.secondlife.com/whatis/­market​ place.php (website available until early 2008). Thereafter, the webpage was changed. See also Bragg supra note 1. 21   Ibid. (emphasis in the original). 22   Ibid. 23   See generally Second Life’s FAQ website, online: http://secondlife.com/whatis/?lang=en-US. 24   See Mark Wallace, ‘The Game Is Virtual. The Profit Is Real’ The New York Times (29 May 2005) at 37. 25   See F. Gregory Lastowka and Dan Hunter, ‘The Laws of the Virtual Worlds’ (2004) 92 Cal. L. Rev. 1 at 8–9. 26   Rob Hof, ‘Second Life’s First Millionaire’ Business Week (26 November 2006), online: www. businessweek.com/the_thread/techbeat/archives/2006/11/second_lifes_fi.html. 19 20

BARFIELD PRINT.indd 572

13/11/2018 11:53

Property rights in virtual and augmented reality  573 have been purchased, they can be subdivided and resold.27 Over a period of around two and a half years, Anshe Chung bought and developed virtual real estate from an initial investment of US$9.95 for a Second Life account. She began with smallscale purchases of virtual real estate, which she then subdivided and developed with landscaping and themed architectural builds for rental and resale. Since then, her profits have also come from the development and sale of properties for largescale real-world corporations. Anshe Chung’s holdings in Second Life have been reported to include virtual real estate that is equivalent to 36 square kilometres of land, several virtual shopping malls, virtual store chains, and several virtual brands, as well as ‘cash’ holdings of several million Linden Dollars. She also has significant virtual stock market investments in Second Life companies.28 Importantly, all of this grew from her real estate investments in Second Life, which indicates that real estate in Second Life is a significant commodity in the virtual world. Given the successes seen by individuals, it is not surprising that traditional real-world businesses have also set up shop in Second Life, with international brands selling products such as clothing, virtual and real.29

4.  SECOND LIFE AND VIRTUAL ITEMS The ability of players such as Anshe Chung to profit from their time spent in Second Life stems from the policy on ownership of virtual items in it. On 14 November 2003, Linden released a statement announcing that it had revised the terms of service (‘TOS’) for Second Life and that it now allowed subscribers to retain full intellectual property protection for the digital content they create while playing the game, including characters, clothing, scripts, textures, objects, and designs.30 This was seen by most as a bold move on the part of the company. Specifically, clause 3.2 of the Second Life TOS was changed to state: ‘Linden acknowledges and agrees that, subject to the terms and conditions of this Agreement, you will retain any and all applicable copyright and other intellectual property rights with respect to any Content you create using the Service.’31 The rationale given for the change in policy was presented by Linden in its press release as follows:32 Until now, any content created by users for persistent state worlds, such as EverQuest or Star Wars Galaxies, has essentially become the property of the company developing and housing the world . . . We believe our new policy recognizes the fact that persistent world users are making

27   See Julian Dibbell, ‘The Unreal Estate Boom’ Wired (January 2003), online: www.wired.com/ wired/archive/11.01/gaming.html?pg=1&topic=&topic_set=. 28   Press Release, ‘Anshe Chung Becomes First Virtual World Millionaire’ Anshechung.com (26 November 2006), online: www.anshechung.com/include/press/press_release251106.html; Sean F. Kane, ‘Virtual Wealth Management: Asset Creation, Seclusion, and Money Laundering in the Online World’ (2006) 185 N.J.L.J. 988. 29   Wagner James Au, ‘Adidas, Toyota, Come to Second Life’ GigaOM (20 August 2006), online: http://gigaom.com/2006/08/20/adidas-toyota-come-to-second-life. 30   Press Release, ‘Second Life Residents to Own Digital Creations’ Linden Lab (14 November 2003), online: http://lindenlab.com/press/releases/03_11_14. 31   ‘Terms of Service’, online: http://secondlife.com/corporate/tos.php. 32   Supra note 30.

BARFIELD PRINT.indd 573

13/11/2018 11:53

574  Research handbook on the law of virtual and augmented reality significant contributions to building these worlds and should be able to both own the content they create and share in the value that is created. The preservation of users’ property rights is a necessary step toward the emergence of genuinely real online worlds.

It would appear from this press release alone that Linden was truly adhering to the position that it was relinquishing all property rights in the virtual-world items created by players. It stated that ‘the preservation of users’ property rights’ was essential and that players should ‘own the content they create’ as well as ‘share in the value that is created’. In addition, it would appear that Linden recognised that general property rights, which are inclusive of intellectual property rights, exist in Second Life. However, the actual wording of the clause in the TOS referred only to intellectual property rights and not general property rights, and therefore questions arose as to what exactly Linden’s intention was. As was referred to in the Linden press release, many if not all virtual world providers state in their EULAs or TOS that the virtual-world provider owns all property rights, not just the intellectual property rights, in all players’ creations. It could be argued that at the time of the Linden press release, well-publicised disputes had only been seen in the intellectual property area, and this is the reason for the change in the TOS to only refer to intellectual property rights, whereas Linden’s intention was that all property rights of the players would be respected. Indeed, as will be discussed presently, Linden’s representations over the course of the period between that press release and the Bragg litigation in 2006 tend to support the contention that it intended to refer to all property rights, not just those in intellectual property. 4.1  The Tension The EULAs of other virtual worlds and the ensuing intellectual property disputes – which were well publicised before the Linden press release – resulted in the unhappy position that if one were to exercise one’s imagination and create a highly original online character complete with weaponry, pursuant to most EULAs or TOS, the creator would nonetheless not be legally able to, for example, write a comic book or sell t-shirts with pictures of the character.33 As Miller has noted, ‘A subscriber creates. He is encouraged to create. However, he owns nothing but a good time.’34 Another useful analogy here would be the software and platforms for word processing. Whilst software developers provide and own the facilities for creating documents, they do not own the documents and artworks that are created with the software. Most complaints in this area arise from the virtual world provider asserting ownership over the intellectual property created by gamers, with the use of EULAs or TOS being blamed for this injustice. These clauses are fairly sweeping and are extremely onesided.35 They do not allow 33   At least in an American context, an EverQuest subscriber ‘is denied from creating any derivative fiction based on his character, his character’s adventures, or items he creates in space’. See supra note 5 at 466. 34   Supra note 5 at 466. 35   See clause 7 of ‘Terms of Use for Star Wars Galaxies: An Empire Divided’, online: http:// help.station.sony.com/cgi-bin/soe.cfg/php/enduser/std_adp.php?p_faq​id=15629&p_created​=​1​114​ 4850​03&p_sid=KJByKL*j&p_accessibility=0&p_redirect=1&p_srch=1&p_lva=16​20​6&​p_sp=cF 9zcmNoPTEmcF9zb3J0X2J5PSZwX2dyaWRzb3J0PSZwX3Jvd19jbnQ9MjI5LDIyOSZ​wX3Byb​

BARFIELD PRINT.indd 574

13/11/2018 11:53

Property rights in virtual and augmented reality  575 any leeway for the player to retain any rights – whether general property or intellectual property rights – over their creations. One of the earliest reasons put forward for the denial of players’ copyrights was that if players/creators were permitted to retain any intellectual property or other property rights, the law would be allowed into the game and it would no longer be a game.36 This argument is unconvincing, as it has not stopped game companies from claiming copyright infringement when they experience a problem, whether legal, commercial or economic. It is submitted that the use of such broad intellectual property law claims was largely to curb the out-of-game sales of virtual-world items discussed previously, which would in turn help increase the revenue of the virtual world providers. After all, if players can purchase characters and virtual-world items for real world dollars, this would decrease the amount of time players, especially inexperienced ones, would need to spend in the virtual world, which would in turn decrease the revenue for virtual-world operators as they would be collecting monthly subscription fees from fewer people. The situation was exacerbated by the many ‘gold farming’ companies that have been operating in developing countries. These ‘gold farmers’ are basically sweatshops that hire users to play these games solely to acquire or create valuable virtual world items and then sell the items outside of the virtual world.37 An illustrious example is Black Snow Interactive, which was based in Tijuana, Mexico and which hired players to play the Mythic Entertainment-owned game Dark Age of Camelot.38 Like many virtualworld providers, Mythic Entertainment claimed this ‘farming’ infringed the company’s copyrights and asked online auction sites such as eBay to shut down any auctions of these items. However, reliance on copyright law as a means of curbing this practice is somewhat artificial, and indeed ineffective if virtual items are characterised as property. In fact, it could be argued that copyright has been used as a legal tool by some virtualworld providers to indirectly maintain their economic position. 4.2  The Policy Decision Made by Second Life In light of the foregoing, when Linden entered the virtual-world market in 2003, it understood the tensions that already existed between existing virtual-world operators and their players. Being a new entrant in the market, in order to succeed, Linden also needed to distinguish itself from the other established virtual worlds. Hence, it made a business decision to depart from the industry standard of denying that players had any rights to virtual items. As has already been discussed, this decision could be justified from the view of players, and indeed was probably intended to draw players to Second Life. From the point of view of a virtual-world provider, especially the provider of a nonlevel-based virtual world, there were no convincing reasons why virtual-world providers needed 2RzPTAmcF9jYXRzPTAmcF​9wdj0mcF9jdj0mcF9zZWFyY2hfdHlwZT1hbnN­3ZXJzLnNlYXJ​ jaF9ubZwX3BhZ2U9MSZwX3NlYXJjaF90ZXh0PWludGVsbGVjdHVhbCBwcm9wZXJ0eQ!!​ &p_li=&p_topv​iew=1. 36   Supra note 4 at 73. 37   Mark Ward, ‘Fantasy Fuels Games with Finances’ BBC News (30 December 2005), online: http://news.bbc.co.uk/2/hi/technology/4543212.stm. See also supra note 3. 38   Supra note 14.

BARFIELD PRINT.indd 575

13/11/2018 11:53

576  Research handbook on the law of virtual and augmented reality to assert intellectual property ownership over player creations. So it appeared to be a win–win business decision for Linden. Following the announcement on 14 November 2003 in which Linden expressly recognised players’ property rights in their online creations, Linden continued to make statements and representations that preservation of players’ property rights is a fundamental part of Second Life.39 Much of the language used by Linden seems to indicate that Linden was referring to property rights in general and not just intellectual property rights. In December 2003, Linden began allowing players to ‘own’ land in Second Life. Players were able to buy land at land auctions and payment would be made to Linden in US dollars for parcels of land. In June 2004, Philip Rosedale, the Chief Executive Officer of Linden, was quoted as saying: ‘The idea of land ownership and the ease with which you can own land and do something with it . . . is intoxicating . . . land ownership feels important and tangible. It’s a real piece of the future.’40 One year later, in June 2005, Rosedale was again quoted in an interview published by Guardian Unlimited: Gamesblog. During the course of the interview, Rosedale said: ‘We like to think of Second Life as ostensibly as real as a developing nation . . . The fundamental basis of a successful developing nation is property ownership . . . We started selling land free and clear, and we sold the title, and we made it extremely clear that we were not the owner of the virtual property.’41 In July 2006, in an interview with After TV, Rosedale was asked: ‘So your economic model is selling virtual land; do you have an advertising model?’ Rosedale stated in response: ‘[E]veryone owns their own stuff, their own property – there’s no way we could just advertise on that property without asking because it isn’t ours you know. It belongs to the land owners.’42 These statements and representations are just a selection of those made by Linden over the years. They clearly indicate that as far as Linden is concerned, owners of land in Second Life do indeed own the land, including the title. It would also be fair to conclude that Second Life’s month over month record growth in subscriber acquisition during those years is attributable to the economic and investment opportunities offered by Second Life,43 which entice players to spend and make substantial amounts of money.44

39   See Complaint, Bragg v. Linden Res., Inc., No. 06-08711 (Pa. Ct. Com. Pl. 2006) [Bragg (2006)] and Bragg, supra note 1. 40   Michael Learmonth, ‘Virtual Real Estate Boom Draws Real Dollars’ USA Today (3 June 2004). 41  ‘Second Life and the Virtual Property Boom’ Guardian Unlimited: Gamesblog (14 June 2005), online: http://blogs.guardian.co.uk/games/archives/2005/06/14/second_life_and_the_virtual_­ property_boom.html (emphasis added). 42   See interview with Philip Rosedale (20 July 2006) online: http://andrewkeen.typepad.com/ aftertv/2006/07/index.html (emphasis added). 43   Linden and Rosedale’s Answer & Counterclaim against Bragg in Bragg, supra note 1 at para. 51. 44   See Robert D. Hof, ‘My Virtual Life’ Business Week (1 May 2006), online: www.businessweek.com/magazine/content/06_18/b3982001.htm, where it is stated that in January 2006, ‘inside Second Life alone, people spent nearly $5 million in some 4.2 million transactions buying or selling clothes, buildings, and the like’; ‘Virtual Online Worlds’ The Economist (28 September 2006), at 62, online: www.economist.com/business/displaystory.cfm?story_id=7963538.

BARFIELD PRINT.indd 576

13/11/2018 11:53

Property rights in virtual and augmented reality  577 In the following section, the concept of land ownership in Second Life under property law will be examined.

5.  LAND OWNERSHIP IN SECOND LIFE The legal meaning of land ownership in Second Life found itself in the legal spotlight because of the dispute between Marc Bragg and Linden. The case resulted in a decision at the United States District Court level concerning one aspect of contract law.45 The remaining issues, including the property issue, were originally set for hearing in December 2007,46 but the parties settled in early October 2007. While many legal questions remain unanswered, it is heartening to note that Linden has amended some of its procedures and websites to overcome some of the weaknesses highlighted by the case. This section will first examine the Bragg dispute, and then analyse the features of land ownership in Second Life within the context of existing legal scholarship on property law. 5.1 The Bragg Dispute – Overview Marc Woebegone was the online avatar of Mark Bragg, a lawyer based in Pennsylvania. Woebegone had accumulated a substantial portfolio of real estate and currency in Second Life and also ran nightclubs and other businesses.47 Bragg was keen to increase his profits from his land investments in Second Life and began investigating means of taking advantage of the online auction interface. Bragg contacted another player through the Second Life online chat facility, where he learnt of a way of accessing unlinked URLs or web addresses in order to prematurely start land auctions that would not be visible to players who did not know how to access them.48 This meant that the minimum bid of US$1,000 set by Linden for its parcels of land at its land auctions would not be applicable and there would be very few, if any, competing bidders. According to court records, another player self-initiated a land auction. Bragg was one of the bidders and became the owner of a parcel of land named ‘Taessot’ for which he paid Linden US$300.01. It is unclear from Second Life’s TOS and EULA if the practice was prohibited. Certainly, it is arguable that Bragg knew Linden had set a minimum bid for the land parcels which it auctions as he had previously purchased land from Linden in the normal way. However, there was nothing in the TOS or EULA which stated that all land sold by Linden had a minimum price of US$1,000, nor was there anything that stated that land could only be purchased through the online auction in a certain way. Indeed, the ‘exploit’ in the system of which Bragg and others took advantage was a result of a loophole left open by Linden’s computer programmers. While Bragg’s actions may be considered morally reprehensible, Linden also had a responsibility, if not a duty, to its players to ensure that certain important transactions and key features of the virtual world are properly   Bragg, supra note 1.   The case was originally placed in the trial pool for 17 December 2007. 47   Kathleen Craig, ‘Second Life Land Deal Goes Sour’ Wired (18 May 2006), online: www. wired.com/news/culture/0,70909-0.html. 48   Supra note 43. 45 46

BARFIELD PRINT.indd 577

13/11/2018 11:53

578  Research handbook on the law of virtual and augmented reality secured and functioning properly. The auction and sale of land by Linden, which usually involves transactions involving US$1,000 and upwards, would certainly come under such a special category. If it could be said that perhaps Linden was acting negligently for not ensuring that its computing systems were functioning properly, then perhaps Linden should have relied on a legal, not a technological, solution. As Moore has said of the EULA:49 ‘It is a legal device designed to limit the gamers’ capacity for legitimate ownership when playing the game, but because installation of the game on the player’s computer requires acknowledgement of the contract’s terms, players are powerless to object.’ This being so, it would have been open to Linden to dictate in the TOS or EULA that land auctions could only occur validly in a certain way and that all other methods of acquisitions would be null and void. This would have incorporated these conditions so they form part of the contract of subscription to the virtual world or the contract for sale of virtual land. Alternatively, the statements could have appeared on Second Life’s website in prominent and relevant places. This would have alerted the players and could arguably have been an implied term of the subscription contract or the contract for the sale of land in Second Life. Linden, however, did not do any of these things. Furthermore, it accepted Bragg’s money and Bragg was given the parcel of land known as Taesot, thereby concluding the contract for the sale of land. Soon after the land was transferred to Bragg, Linden discovered the anomaly and froze Bragg’s account on Second Life. Linden subsequently resold all of Bragg’s real estate in Second Life (not just Taesot), deleted his avatar and denied him access to all of his virtual property. Bragg did not receive any compensation or reimbursement. Linden had initially told Bragg that it intended to rescind the sale of Taesot and refund the price paid, this being US$300.01, but this never occurred, even though it would have been the equitable course of action. Bragg retained counsel and sued Linden, seeking financial damages and specific performance and requesting the return of the land and access to it, in part for a breach of a virtual land auction contract, fraud and violations of the Pennsylvania Unfair Trade Practices and Consumer Protection Law50 and California Unfair and Deceptive Practices Act,51 among other claims.52 Bragg alleged that at the time his avatar was deleted from Second Life he had provided to Linden US$13,900 for items such as land use fees and land purchases and that his ongoing rentals, firework shows and projects under construction were generating approximately US$50 per month. He also alleged that his purchase and sale of land in Second Life was generating approximately US$1,200 per month. Bragg alleged that he relied upon statements made by Linden which indicated that property rights in Second Life were inviolable, and that he was thereby induced to invest in Second Life.53 49   Christopher Moore, ‘Commonising the Enclosure: Online Games and Reforming Intellectual Property Regimes’ (2005) 3(2) Australian Journal of Emerging Technologies and Society 100. 50   Unfair Trade Practices and Consumer Protection Law, 73 P.S. (1968). 51   Unfair and Deceptive Practices Act, Cal. Bus. & Prof. Code § 17200. 52   Bragg (2006), supra note 39. 53   Ibid.

BARFIELD PRINT.indd 578

13/11/2018 11:53

Property rights in virtual and augmented reality  579 5.2  The Bragg Dispute – TOS The first round of the dispute was fought over the TOS when Linden filed motions to dismiss the suit and compel Bragg to arbitrate his claim out of court, consistent with Linden’s TOS.54 Judge Eduardo Robreno held that Linden’s TOS constituted a ‘contract of adhesion’, which was also a substantively unconscionable contract, and that the arbitration clause was invalid. The court therefore refused to enforce the compulsory arbitration clause contained in the TOS and ordered the legal suit to proceed.55 The purpose of this chapter, however, is to focus not on the contract law issues, but rather on the property law issues. 5.3 The Bragg Dispute – The Concept of Land in Second Life The concept of land in Second Life became an issue in the dispute when Linden denied the existence of any ‘property’ in the land it sold in Second Life. In the court documents, Linden and Rosedale, the defendants, stated as follows:56 Defendants admit that Second Life contains an integrated economy that enables users to purchase and sell rights in Objects or other user-created content for various forms of consideration. Defendants further admit that, subject to the Second Life Terms of Service and other applicable rules and policies, Second Life users may also purchase and sell representations of parcels of ‘land’ in Second Life, hereinafter referred to as ‘virtual land.’ Defendants deny that Second Life allows for the actual ‘conveyance of title’ in ‘virtual land,’ as ‘virtual land’ is not property to which one may take ‘title,’ but instead a license of access to Linden’s proprietary servers, storage space, bandwidth, memory allocation and computational resources of the server, which enables the experience of ‘land’ and the things that one can do with ‘land’ on the Second Life platform. Defendants further admit that the access rights represented by ‘virtual land’ in Second Life can be purchased using Linden Dollars or U.S. Dollars.

It would appear that what the defendants were arguing was that they were simply leasing out (very expensive) computer processors when they ‘sold land’ in Second Life. Indeed, the defendants’ description was probably an accurate scientific explanation of what users are engaged in when they interact with land in Second Life. The law, however, has not always accepted the scientific view as the legal view.57 Second, the defendants’ explanation completely ignored the booming and energetic economy that had developed around land in Second Life – players did not pay thousands of US dollars just to give themselves ‘the experience of “land” and the things that one could do with “land” on the Second Life platform’, as Linden claimed. Lastly, the defendants, in denying the existence of any property, also closed their eyes to the expectations of the players – expectations which had been formed as a result of the representations made by Linden and its officers, as already discussed.

    56   57   54 55

The court also had to consider jurisdiction arguments over Rosedale, the CEO of Linden. Bragg, supra note 1. Supra note 43 at para. 8. See, e.g., Gutnick v. Dow Jones Co Inc, [2001] Vic. S.C. 305.

BARFIELD PRINT.indd 579

13/11/2018 11:53

580  Research handbook on the law of virtual and augmented reality 5.4  The Legal Concept of Property and Land Ownership in Second Life In order to properly examine the legal status of land and land ownership in Second Life, the first question that needs to be asked is, simply, ‘What is property?’58 It should be remembered that property is a legal construct. The High Court of Australia has said: ‘The concept of property may be elusive . . . and it may be . . . that ‘the ultimate fact about property is that it does not really exist: it is mere illusion.’’59 The Court went on to cite Professor Kevin Gray: ‘[M]uch of our false thinking about property stems from the residual perception that “property” is itself a thing or resource rather than a legally endorsed concentration of power over things and resources.’60 In Western Australia v. Ward, another decision by the High Court of Australia, the majority held that ‘the common law’s concern [is] to identify property relationships between people and places or things as rights of control over access to, and exploitation of, the place or thing’.61 These notions of property reflect the thinking in a number of Commonwealth jurisdictions, such as Australia, Singapore and the United Kingdom, on the concept of property. In essence, property is not a thing but a socially approved power relationship in respect of socially valued assets.62 The two key features of property, then, are a presumptive right to exclude others (or control over access) and discretion in the manner of exploitation.63 We will now turn to these two features of land ownership in Second Life to examine whether land ownership does take on the features of ‘property’ and to examine what similarities exist between ownership of land in the real space and the ownership of land in Second Life. As previously mentioned, once one has successfully bid for a parcel of land in Second Life, one can subdivide the land and resell the land. In addition to this, the Second Life website for a time provided details of what else one can do with the land one owns: ‘Owning land allows you to control land. You can prevent others from visiting or building there, change the shape of the land, subdivide and sell it, and much more.’64 The statement is clear. When one owns land, one can control the land. If the control amounts to control over access to, and exploitation of, the land, as was held in Ward, then clearly, property rights would arise over the land in Second Life. On the same webpage as the abovementioned statement, much detail was given as to

58   For an excellent discussion of virtual property in general from the perspective of the American legal tradition, see Joshua A.T. Fairfield, ‘Virtual Property’ (2005) 85 B.U.L. Rev. 1047. 59   Yanner v. Eaton [1999] 201 C.L.R. 351 (H.C.A.) [Yanner] at para. 17, citing Kevin Gray, ‘Property in Thin Air’ (1991) 50 Cambridge L.J. 252 at 252. 60   Yanner, ibid at para. 18. 61   Western Australia v. Ward, [2002] 213 C.L.R. 1 (H.C.A.) [Ward] at para. 88, per Gleeson C.J. and Gaudron, Gummow, and Hayne JJ. 62   Kevin Gray, ‘Equitable Property’ (1994) 47(2) Curr. Legal Probs. 157 at 160. 63   There is also of course the need that the so-called property is definable or identifiable, but this is not an issue here as each parcel of land in Second Life has not only its unique name and location in Second Life but also a unique location on the computer processor. 64   ‘Land: How To’ online: Second Life http://secondlife.com/community/land-howto.php (website available until early 2008). Thereafter, the information was separated into many webpages in the webpages on Land, online: http://secondlife.com/land/?lang=en-US.

BARFIELD PRINT.indd 580

13/11/2018 11:53

Property rights in virtual and augmented reality  581 the types of control one has.65 These will now be considered to demonstrate the breadth and depth of control a user has over the virtual land. As it will become apparent, much of these controls are also controls which landowners and tenants have over physical land.66 First, the owner can allow others to, or restrict others from, creating, editing or placing objects on the land. Note that this only refers to others as the landowner always has the ability to build or place things on his or her own land. In the real world, an example of this would be whether one allows others to build a house or plant trees on one’s land or graffiti one’s walls and so on. Second, the owner can dictate whether the land is ‘safe’ – that is, no damage can be done to those visiting the land – or not safe. For example, in the real world, one could (with notice, of course) keep dogs within one’s land to attack any person or thing that comes onto the land. This would render the land not safe in the Second Life sense. Third, the owner of the land can determine whether or not sound can enter or leave the parcel of land. The equivalent of this in the real world would of course be the use of soundproofing to keep sound in or keep it out. The fourth feature of land ownership in Second Life is the ability, at a cost, to have the land publicly listed so that other avatars can easily locate the land. One would need to specify the category under which the land is to be listed, for example, homestead or store. The real-world analogy of this would be something along the lines of a telephone book or a street directory, except that the listing would be of the property and not of the person or business as in the telephone book, or the street name as in the street directory. Fifth, in Second Life, the owner can dictate whether other avatars can fly on the land. It is possible for others to fly over one’s land at a prescribed height – similar to real life, where one cannot generally prevent airplanes from flying above one’s land – but landowners in Second Life who do allow others access to their land also have the option to prevent others from starting to fly while they are on the land. Given that in real life human beings cannot fly, it can however be argued that in some respects this capability to control goes beyond that possessed by a real-world landowner. A real-world landowner generally cannot, for example, prevent birds from starting to fly from within their land (without the ­construction of horizontal nets or similar devices). Sixth, landowners in Second Life can allow others to run, or prevent others from running, scripted objects on their land. What this means is essentially that landowners can prevent others from doing things such as firing scripted weapons on one’s land. Avatars may be allowed to visit but while there they are prevented from doing things, good or bad, that are not in the default. Examples of harmless activities may include animations or chimeras for customised dancing or the launch of particle effects to liven up the atmosphere. Because a large variety of effects, some harmless and some harmful, can be achieved with scripted objects, a lot of places, such as nightclubs, in Second Life do not permit scripted objects to be run on their land. This type of control in some sense goes beyond the control a real-world landowner possesses. In practice, it is not possible (short of imposing security checks) to out and out prevent people from bringing things such as 65   Ibid. The information on Second Life in the remainder of this section is taken from the same webpage unless stated otherwise. 66   See also Lawrence Lessig, Code and Other Laws of Cyberspace (New York: Basic Books, 1999), which discusses the power of code to regulate in the virtual, powers that are markedly ­different from the power of architecture in real space.

BARFIELD PRINT.indd 581

13/11/2018 11:53

582  Research handbook on the law of virtual and augmented reality mini firecrackers and fireworks into clubs and shopping malls and setting them off. In the real world, however, the deterrent for this kind of behaviour is simply civil or sometimes criminal sanctions. Seventh, not only may avatars purchase land and make improvements to the land, they can also exclude other avatars from entering onto the land. Access may be denied to the general public or limited to a certain group and/or up to 50 particular avatars. A landowner in Second Life may also sell passes to access the land and select the price and time limit for the pass. As noted above, those who are banned from accessing a parcel of land in Second Life may still fly over the land, but the flight must be above a certain height. This is clearly analogous to the real world, where prying helicopters and planes can fly over private properties at a certain height. The ability to restrict access to the land in Second Life is clearly a feature found in the real-world ownership of land. Similarly, Second Life landowners can freeze and eject people from their land. Ejection of persons is an action available to landowners in the real space, but the freezing of avatars certainly goes beyond what is physically possible in the real world. On the whole, the classes of persons allowed or not allowed to enter land in the real world are replicated in Second Life. Whilst occupiers cannot physically freeze invited guests on the land, occupiers do have the legal power to restrict access to certain parts of the land or premises by, for example, locking doors so that invited guests cannot move freely about on the premises. Eighth, landowners in Second Life can work the land by ‘terraforming’, which means flattening, raising, lowering, smoothing or roughening the land. One can terraform the whole parcel of land or parts of the parcel. All this is clearly possible with real property in the real world. Ninth, as already mentioned, landowners can subdivide the land and, conversely, can also join together two or more parcels of land if they own the parcels. Tenth, if an avatar is feeling generous or simply does not have the time to sell the parcel of land, it can give it away, just as in real life. Finally, Linden charges landowners in Second Life a monthly ‘Land Use Fee’ if the avatar owns more than 512 square metres of land.67 The first 512 square metres is free but thereafter the fee increases according to the size of the land holding. The most obvious equivalent of the Land Use Fee in the real world is land tax. In conclusion, most if not all of these controls are also controls that real property holders, such as occupiers and landowners, possess in the real world. Granted, occupiers cannot subdivide land, but they can sublease part of the land. From the foregoing, it is clear that ownership of land in Second Life gives the owner a presumptive right to exclude others. Further, the second element of discretion in the manner of exploitation is also satisfied, as there is a wide discretion available to the landowner in the manner of exploitation of the land. Hence, under the principles of property law as it is known in many Commonwealth jurisdictions, it would appear that land ownership in Second Life gives rise to some kind of property rights. It could also be said that there is ‘property’ in the land in Second Life. But the question is: what form of property rights? The foregoing analysis has shown that a substantial number of similarities exist between ownership of land in the real space and ownership of land in Second Life. The law has recognised all sorts of property rights in relation to land – for example, 67

  See ‘Mainland Pricing and Fees’, online: http://secondlife.com/whatis/landpricing.php.

BARFIELD PRINT.indd 582

13/11/2018 11:53

Property rights in virtual and augmented reality  583 mortgages and leases. Could it be that the property rights that subsist in land ownership in Second Life are akin to some forms of property rights related to land as we know it in the real world? It is submitted that owning land in Second Life resembles holding some modified form of leasehold property in the real world. In particular, section 6 argue that parallels can be drawn with the modified form of leasehold in public housing in Singapore.

6. THE LEASEHOLD IN PUBLIC HOUSING IN SINGAPORE, GRAZING LICENCES AND LAND OWNERSHIP IN SECOND LIFE We saw in the previous section that excludability and the discretion in the manner of exploitation form the foundation of the concept of property. It may be suggested that while the landowner in Second Life has the requisite control over the land, ultimately he or she may not have meaningful real control because that control can be lost easily if, for example, the user fails to pay the monthly subscription fee or the monthly land use fee. In most cases, the nonpayment of such fees will result in one losing ownership of the land. The second objection that may be raised as to the control held by a landowner in Second Life is in respect to one particular clause in the TOS that overshadows this control. Clause 2.6 reads:68 Linden Lab has the right at any time for any reason or no reason to suspend or terminate your Account, terminate this Agreement, and/or refuse any and all current or future use of the Service without notice or liability to you. In the event that Linden Lab suspends or terminates your Account or this Agreement, you understand and agree that you shall receive no refund or exchange for any unused time on a subscription, any license or subscription fees, any content or data associated with your Account, or for anything else.

Leaving aside issues of whether the above clause is unconscionable or unfair and whether it is enforceable, it would certainly seem that a landowner’s control in Second Life is at the absolute discretion of Linden. After all, Linden can, for no reason, suspend or terminate an account and give no refund. We will now turn to these two issues. In the next section, the argument will be made that just because land ownership in Second Life can come to a premature and abrupt end, this should not deprive land ownership in Second Life of the character of property. The foundation of the argument will be the leasehold in public housing in Singapore to which property rights do attach, coupled with analysis of legal and equitable interests in land. 6.1  Leasehold in Public Housing in Singapore – Premature End to Ownership The leasehold is the most prevalent method of land holding in Singapore. Importantly, the public housing system administered by the Housing and Development Board (‘HDB’) provides the key to understanding land ownership in Second Life. Under the Housing and

68   Supra note 31. Note that this was originally clause 7.1 when Bragg launched the legal proceedings but was subsequently renumbered. The text remains unchanged.

BARFIELD PRINT.indd 583

13/11/2018 11:53

584  Research handbook on the law of virtual and augmented reality Development Act,69 the HDB has broad powers to manage all lands, houses and buildings or other property vested in or belonging to HDB, including preparing and executing proposals, plans and projects for the erection, conversion, improvement and extension of any building for sale, lease, rental or other purposes.70 Part IV of the Housing and Development Act deals with the HDB’s power to sell flats, houses and other buildings. As noted by William Ricquier,71 the HDB in fact rarely sells property, but rather leases it. The definition of ‘owner’ in s. 2 of the Housing and Development Act is interesting as it includes a person who has purchased a leasehold interest in the property and also includes a purchaser under an agreement for a lease. Indeed, most people who buy from the HDB generally regard their transactions as sales.72 The scheme is in effect a modified form of the leasehold as it is known at common law. The public housing scheme meets and exceeds social policy and desires by ensuring sufficient, fair and equitable housing. Apartment prices for flats sold by HDB are pegged in relation to specific income levels to ensure the flats are affordable. However, the HDB encourages an active resale market after the ‘minimum occupation period’73 and because of this, the resale value of flats in sought-after locations can be several times higher than that of a new flat sold by the HDB. The public housing scheme is an equitable one and for it to be so there must be strict rules in place ensuring fair play, social justice and equal access to public housing. To this end, ss. 55 and 56 of the Housing and Development Act list circumstances that will trigger a premature end to the leasehold ownership. The most common duration for a leasehold flat that the HDB sells is 99 years, but if the owner (or, in some cases, other authorised occupiers) has breached any of the provisions in s. 55 or s. 56, then the HDB may reenter the premises and determine the lease absolutely under s. 55 or compulsorily acquire the premises under s. 56. For example, under s. 55(1)(a), the nonpayment of rent for three months after the HDB has sent a letter of demand will trigger a right of reentry by the HDB. Similarly, under s. 56(1)(b), the HDB can compulsorily acquire any flat if the owner, the owner’s spouse or any authorised occupier has acquired any title to an estate or interest in any other flat, house or building or land, with the exception of some forms of commercial property.74 Significantly, unlike forfeiture, the owner of a HDB flat has recourse to an appeals structure as well as compensation as set out under s. 56. In short, under the scheme, there is a framework to instill confidence in home ownership. The relevance of these two sections to land ownership in Second Life is that parallels can be drawn in terms of the rights of control of the property owner coming to a premature end as a result of some triggering event. In Second Life, the nonpayment of the subscription fee or land use fee would be a triggering event. Under the public housing

69   Housing and Development Act (Cap. 129, 2004 Rev. Ed. Sing.) [Housing and Development Act]. 70   Ibid at s. 13. 71   William J.M. Ricquier, Land Law, 3rd ed. (Singapore: LexisNexis, 2007) at 22. 72   See also Tan Sook Yee, Private Ownership of Public Housing in Singapore (Singapore: Times Academic Press, 1998) at 129–38, where Professor Tan concludes that owners of HDB flats can be said to own their flats and have property rights over them. 73   Supra note 69, s. 49A. 74   Ibid at s. 56(2).

BARFIELD PRINT.indd 584

13/11/2018 11:53

Property rights in virtual and augmented reality  585 scheme in Singapore, one example of an event triggering a premature end to ownership would be nonpayment of rent for three months. The only difference between buying land in Second Life and buying an apartment from the HDB is that the HDB cannot and does not arbitrarily, at its absolute discretion, terminate the ownership and to do so without compensation. This difference, however, should not impact upon whether or not land or land ownership in Second Life has proprietary rights attached. Rather, it reflects the lack of justice as well as substantive and procedural unfairness in the Second Life regime. Land ownership in Second Life and ownership in public housing in Singapore can both come to a premature and abrupt end: this is the crucial element. The point being made here is that the fact that the methods by which the ownership can be determined are different should not be relevant to questions of whether property rights subsist. Owning land in Second Life is in fact very much like owning a modified form of leasehold property. To put it another way, the holding of freehold real property can also be uncertain and defeasible as the land can always be compulsorily acquired, sometimes for seemingly arbitrary reasons.75 6.2  Discretion to Terminate Given the argument above that the focus should be on the fact that property interests can come to a premature and abrupt end, Second Life’s ability to terminate at will needs further examination under traditional real property (land) law. In particular, there may be lingering doubts as to whether this discretion may render the property right to be only a personal right. The discussion will now turn to grazing licences in Australia and the law on profit-à-prendre. The leading authority on Australian grazing licences is the High Court of Australia’s decision in R v. Toohey; Ex parte Meneling Station Pty Ltd.76 The case concerned grazing licences granted pursuant to statute. A number of legal points were at issue. We will only deal here with the property issue, on which the five justices of the High Court all concurred. The grazing licences were not assignable and they were terminable on three months’ notice by the relevant minister. The question was whether the holder of a grazing licence has an ‘estate or interest’ in the land, the subject of the licence, within the meaning of another piece of legislation. The applicants’ contention was that this is a question to be answered by determining whether the grazing licence created under statute is a profità-prendre – if so, that this would mean the grazing licence constitutes an interest in land, as a profit-à-prendre is a proprietary interest in land. Under the legislation, a grazing licence is one granted by the relevant minister to the holder to graze stock, or any particular kind of stock, on specified lands for such period not exceeding one year.77 The relevant Regulations set out the amount payable for a grazing licence based on the area of the land. A grazing licence must include a condition prescribing the maximum number and type of stock that may be depastured on the land

75   See, e.g., the facts of Sole v. Secretary of State for Trade and Industry [2007] EWHC 1527 (Admin). 76   (1982), 158 C.L.R. 327 (H.C.A.) [Toohey]. 77   Ibid at 341.

BARFIELD PRINT.indd 585

13/11/2018 11:53

586  Research handbook on the law of virtual and augmented reality and may include any other conditions that the minister may think necessary or desirable in a particular case.78 Applicants for a grazing licence who have not commenced grazing stock on the land may withdraw the application if they are not satisfied with the conditions imposed by the minister.79 Under the regulation, a grazing licence ceases to be in force on 30 June each year but may be renewed, at the minister’s discretion, for a further period not exceeding 12 months.80 A licence can come to a premature end if there is a failure to comply with a condition of it.81 Further, a licence may be cancelled if the minister gives three months’ notice in writing of the intention to cancel it.82 There is also provision for the surrender of a grazing licence.83 Justice Mason (as he then was), with whom Gibbs CJ and Brennan J (as he then was) agreed, took as a starting point Lord Wilberforce’s statement in National Provincial Bank Ltd. v. Ainsworth:84 ‘Before a right or an interest can be admitted into the category of property, or of a right affecting property, it must be definable, identifiable by third parties, capable in its nature of assumption by third parties, and have some degree of permanence or stability.’85 In Justice Mason’s view, the rights of the holder of a grazing licence created under the statute fall short in two respects of the concept of property or proprietary rights expressed by Lord Wilberforce. First, the existence of reg. 71A, which enables the minister to cancel a licence with three months’ notice in writing of his intention to do so without any necessary default on the part of the licensee, suggests that the licensee has no interest in the land at all, because a right terminable in such a manner lacks the requisite degree of permanence envisaged in Lord Wilberforce’s passage.86 Second, Mason J said that nothing in the statute indicated that the grazing licence is assignable. In fact, he said, all indications are to the contrary. Further, His Honour also placed much emphasis on the personal nature of the right conferred by the grant of a grazing licence. For example, he pointed to the fact that under the statute, a licensee must apply for permission if he wishes to make or erect improvements on the land. This, according to Mason J, is a very strong indication that property in the land remains in the Crown and does not pass to the licensee. Justice Wilson, with whom Gibbs CJ and Murphy J agreed, found the ‘resolution of the issue more than ordinarily difficult’.87 Wilson J was of the view that the statutory grazing licence is ‘more than a mere licence’,88 but the answer to the problem, according to His Honour, is to be ‘found in the view taken of the totality of the legal rights conferred by the statute set against the question, not whether a grazing licence confers a right cor    80   81   82   83   84   85   86   87   88   78 79

Ibid. Ibid. Ibid. Ibid. Ibid. Ibid. [1965], A.C. 1175 (H.L.) [Ainsworth] at 1247–8. Supra note 76 at 342. Ibid. Ibid at 353. Ibid at 352.

BARFIELD PRINT.indd 586

13/11/2018 11:53

Property rights in virtual and augmented reality  587 responding to the common law category of a profit a prendre, but whether it is “an estate or interest in land” within the meaning of that term in the Act’.89 The decision of His Honour turned on a close examination of the statute, which he found to have made a clear distinction between estates in fee simple, leases and easements – all of which are estates or interests in land – and licences. First, Wilson J found that the discretion in the relevant minister to terminate the licence unilaterally and more or less summarily, without compensation (except to improvements erected with the permission of the minister), tends strongly to deny to the grazing licence the character of an interest in land.90 Second, the statute does not contemplate any assignment of the licence, and this can be contrasted with leases under the statute that are transferrable with the permission of the minister. According to Wilson J, it would be extraordinary if the legislation placed such a control on the assignment of leases while allowing grazing licences to be freely transferrable without permission or even notice to the minister. Hence, Wilson J concluded that the grazing licence remains personal to the grantee of the licence.91 The judgments of both justices hung on three main factors: the discretion resting with the relevant minister to terminate the licence unilaterally; the inability of the licensee to assign; and the inability to make or erect improvements on the land without the permission of the relevant minister. The only one of these factors present in the situation in Second Life is the discretion to terminate the grazing licence unilaterally. However, it is important that the judgments are interpreted within the context of the particular statutory regime. Indeed, as with Wilson J, whose views are already noted, Mason J. was also careful to point out that ‘[t]he grazing licence is the creature of statute forming part of a special statutory regime governing Crown land. It has to be characterized in the light of the relevant statutory provisions without attaching too much significance to similarities which it may have with the creation of particular interests by the common law owner of land.’92 In the more recent case of Commissioner of Taxation v. Orica Ltd,93 Gummow J indicated that one should not attach too much significance to the approval by Mason J, in Toohey, of the statement by Lord Wilberforce in Ainsworth. Gummow J said: ‘Mason J was concerned to analyse particular statutory rights, and Lord Wilberforce was dealing with the novel development of the “deserted wife’s equity”. Neither was dealing with rights created under the general law of contract.’ 94 The insight of Gummow J is particularly relevant to the situation of Second Life, where we are dealing with rights created under the general law of contract. While it is true that the Ainsworth formulation requires some permanence or stability, in both Ainsworth and Toohey the courts were concerned with very particular sets of facts and circumstances which may not be universally relevant.

    91   92   93   94   89 90

Ibid. Ibid at 353. Ibid. Ibid at 344. [1998] 194 C.L.R. 500 (H.C.A.). Ibid at para. 110.

BARFIELD PRINT.indd 587

13/11/2018 11:53

588  Research handbook on the law of virtual and augmented reality Furthermore, as Professor Gray has already pointed out, the requirement of ‘permanence’ or ‘stability’ is circular. He states: 95 Quite often – as for instance in National Provincial Bank Ltd v Ainsworth – the reason for asking whether a particular right is proprietary is precisely in order to determine whether the right is capable of binding third parties and thereby attaining just such a quality of ‘permanence’ and ‘stability’. It is radical and obscurantist nonsense to formulate a test of proprietary character in this way.

Returning to the case of Toohey, it was the applicants’ contention that the question of whether the holder of a grazing licence has an ‘estate or interest’ in the land should be answered by determining whether the grazing licence created under statute is a profità-prendre. As we have just seen, the court refused to follow such an approach. If we examine a profit-à-prendre under common law, we will find that the minister’s discretion to terminate the grazing licence unilaterally would not deprive the grazing licence of the character of property. The relevant case on profit-à-prendre is the Australian case of Unimin Pty Ltd v. The Commonwealth.96 In this case, the previous landowner had conferred on the plaintiff the right to remove sand from its land, with royalties payable on the sand removed. Clause 12 of the agreement stated that the arrangement was terminable by either party upon one month’s notice to the other. No default of any kind was required. The court held that the plaintiff was entitled to an interest in land in the nature of an equitable profit-à-prendre and in doing so, it confirmed that a valid profit-à-prendre could come to an end by the giving of an agreed notice. Connor J said: ‘I do not think that an interest which would amount to a profit-à-prendre if it were granted in perpetuity or for a term of years, loses its essential character because it is determinable on a month’s notice, any more than a tenancy does so.’97 The principle to be drawn from Unimin is that a contract clause that gives either party the discretion to terminate the agreement at will does not affect the proprietary nature of the right. Thus, proprietary rights in land ownership in Second Life will survive the clause in the TOS that gives Linden the discretion to terminate the user’s account at will.

7. RAMIFICATIONS Having reached the conclusion that some form of proprietary rights should be recognised in virtual land in Second Life, this section will examine the significance or practical importance of this. The most obvious impact of a right over property is that on the relationships between a landowner in Second Life and the virtual-world provider, Linden, as well as those between virtual landowners and other users. Taking the facts of the Bragg dispute as a starting point, what would this mean? Would the only possible result be that even though Bragg had a proprietary right over the virtual

  Kevin Gray, ‘Property in Thin Air’ (1991) 50 Cambridge L.J. 252.   (1974), 22 F.L.R. 299 [Unimin]. 97   Ibid at 78. 95 96

BARFIELD PRINT.indd 588

13/11/2018 11:53

Property rights in virtual and augmented reality  589 land, that right would come to a premature end as a result of his account and avatar being validly terminated by Linden? And would Linden be required to pay some sort of compensation for the loss in property ownership? Although this may seem to be a simple solution, it may not be the fairest or only solution. What follows will be a discussion of some other possible outcomes. Assuming that the clause giving Linden the discretion to terminate Bragg’s account is not struck down as unconscionable, one possibility would be that the land ownership could be detached from the access to the virtual world, meaning that while Bragg may own the land, he would have no access to Second Life as Linden would have validly terminated his access. In this scenario, Bragg ought to be allowed to dispose of his land to anyone he so wishes and Linden would have to honour access to the new landowner. Linden certainly should not be able to sell off all of Bragg’s land, like it did, without accounting to Bragg. Alternatively, could the proprietorial interest be read to impose a limit on the provider’s contractual discretion to terminate Bragg’s account? This may be slightly more difficult, though not impossible, to argue. From the first principles of property ownership, one ought to be able to exclude others from one’s property, which means that one ought to be able to control and access the virtual land. Would it be technologically possible for a landowner like Bragg to be able to access his virtual land but not the rest of the virtual community? The answer is dependent on the software code of Second Life. If it were possible to do this, then perhaps this would be the simplest solution, as Bragg’s right to his property in the virtual land is preserved, while at the same time Linden would be able to terminate Bragg’s access to and interaction with the rest of Second Life. The land would still be attached to the rest of Second Life, and while others would be able to come to visit Bragg and his ‘secluded’ land, Bragg would have no access beyond the boundaries of his own land, similar to a prisoner who cannot leave the prison but who can be visited by outsiders. In this way, if, for example, Bragg wished to continue running his nightclub on the land to generate revenue, others could come to it; Bragg might however need to find alternative ways of advertising his products and services, as he would not be able to access the other areas of Second Life to advertise. It would also be questionable how much value a ‘secluded’ piece of land would have in Second Life, and it may be even more difficult for Bragg to dispose of such a piece of property. However, if the software is written in such a way that it is not technologically possible to separate access to the virtual land from access to Second Life as a whole, then the question would fall back to whether there are any compelling reasons why technologically and contractually denying Bragg access to Second Life should not be permitted even though he may own land in Second Life. Given that the property owned by Bragg sits within the larger framework of the virtual world of Second Life, it is only fair to take the framework into consideration when balancing the rights and interests of landowners like Bragg and virtual-world providers like Linden. As a virtual-world provider, Linden has overall responsibility for the reputation and integrity of the virtual community as a whole; it would be fettering Linden’s ability to effectively discharge this responsibility if Linden is repeatedly prevented from terminating avatars and users who may be in flagrant breach of the rules. This would also negatively impact upon the Second Life experience and likely devalue the product of Second Life. Hence, the answer would appear to be in the negative – but to be fair to Bragg, he should

BARFIELD PRINT.indd 589

13/11/2018 11:53

590  Research handbook on the law of virtual and augmented reality be allowed to dispose of his land within a reasonable period of time and Linden should recognise the new landowner as the legitimate successor in title. It is difficult to find a perfect solution to the problem, but another analogy from the physical world may assist in clarifying matters. A commercial tenant may own a leasehold on a small island accessible only by a bridge. There is no access by boat as the island is surrounded by swampy marshland infested with crocodiles. It is also not accessible by air as there is no place to land a helicopter, let alone a plane. If the bridge is blown up, then of course the tenant still owns the leasehold, but he is unable to access or control it in any meaningful physical way – although he is free to dispose of the leasehold if the contract does not prevent this. If the bridge being blown up was the fault of an identifiable party, it seems arguable that the tenant can claim some compensation under tort law from the bridge blower. And if the bridge blower was the landlord, the tenant may be able to claim under either tort law or the contract. If, however, the bridge being blown up was the result of the actions of the tenant himself, then he would surely have no recourse against any party (except perhaps claiming on his own insurance policy). Similarly, if a user such as Bragg breaches the rules or fails to pay the subscription fees, resulting in Linden terminating his access to Second Life, he should not be compensated by anyone for the denial of access. And perhaps Linden should be regarded as temporarily holding the virtual land on trust for a set period of time for the landowner until it is acquired by a new landowner. If the virtual land is not sold within a reasonable period of time, it could perhaps revert back to Linden. The issue becomes tricky, however, if there is no breach or no clear identifiable breach by the player, and Linden, at its whim, terminates a user. Although it is unlikely that Linden would embark on this course of action as it is earning good money from its user, in the event that it does happen, should the player obtain any extra compensation from Linden in addition to whatever amount he would receive from selling the virtual land? This is a difficult question, as there is no tortious act (as there is in the blowing up of a bridge) and, importantly, the user has already agreed in the TOS to be subjected to a discretionary power to be terminated by Linden at any time. Further, no doubt the basis for termination would be hotly contested, that is, Linden would be arguing that there was a breach or default while the deprived landowner would be arguing the opposite. Perhaps in such a scenario the user’s best course of action would be to argue the unconscionability of the original contractual clause that gives Linden the full discretion to terminate the user. These outcomes are, however, not without their problems. While they may settle, to some extent, the problem thrown up by the Bragg dispute, treating virtual land as property may have difficult consequences in other settings. For example, what would happen if a virtual-world provider like Linden goes into liquidation? The server would be shut down and the virtual property in the virtual land would be ‘switched off’ until the server is switched back on again. What happens to the virtual land if it is never ‘switched on’ again? Perhaps this is a risk that any user of Second Life has to accept. Characterising virtual land as property may be helpful for tax collectors in jurisdictions where gains derived from dealings in property are taxable. It would provide clarity for both the taxpayer and the tax authorities.

BARFIELD PRINT.indd 590

13/11/2018 11:53

Property rights in virtual and augmented reality  591

8.  AUGMENTED REALITY – POKÉMON GO The augmented reality game Pokémon Go by Niantic was released in July 2016 and was one of the first augmented reality games to achieve widespread popularity, not just in the US, but also in other parts of the world. Pokémon Go is an app and can be downloaded for free by players onto their mobile devices. The key feature of augmented reality is the overlay of in-game locations on real-world landmarks. Pokémon Go makes use of geolocation data to associate in-game locations with real-world places and landmarks. As players move about with their mobile devices, one of the aims of the game is to catch Pokémon, which are virtual in-game creatures. In addition, there are Pokémon Go gyms where Pokémon trainers go to train their Pokémon and battle for control of the gym, and as the game has evolved, it has become essential to join a team to take part in the battle.98 In addition to gyms, there are also PokéStops where players can go to collect items such as eggs, potions and Pokéballs to catch more Pokémon.99 Players initially were able to create new gyms and PokéStops but the game developer has ceased to allow this function, most likely due to the various legal challenges it has faced.100 The gyms and PokéStops have raised significant problems for society at large, that is, nonplayers, as people, institutions, churches and others have suddenly found their homes, churches, land or property listed as gyms and PokéStop locations, meaning that players congregate on or outside their property for the purpose of doing battle for the gym, for example, at all hours of the day and night. The key feature of this augmented reality game is that a player or user must be on the physical location site to play the game or use the application. Apart from gatherings of large numbers of strangers on private land, the appropriateness of the gyms’ and PokéStops’ locations has also been questioned. The United States Holocaust Memorial Museum was, at the inception of the game, a PokéStop. The Museum is a memorial to the victims of Nazism, so playing the game is not appropriate here. There was also an alleged sighting of the Pokémon Koffing, which emits poisonous gas, near the museum’s  Helena Rubinstein Auditorium – an auditorium which shows testimonials of Jews who survived the gas chambers. It is unclear whether the sighting was a hoax or a genuine part of the game.101 At the simplistic level, augmented reality is the projection of virtual images onto realworld surroundings. Simple as this may sound, it has produced a host of legal problems, such as intrusions upon the privacy of players – implicated by Niantic’s ability to track players’ movements – and injuries to users who wander into dangerous situations while

 98   See the Pokémon Go support website https://support.Pokémongo.nianticlabs.com/hc/en-us/ articles/222049347-Battling-at-Gyms.  99   See the Pokémon Go support website https://support.Pokémongo.nianticlabs.com/hc/en-us/ articles/221957688-Gather-items-at-PokéStops-and-Gyms. 100   See the Pokémon Go support website https://support.Pokémongo.nianticlabs.com/hc/en-us/ articles/225375247-How-do-I-create-a-new-PokéStop-or-Gym. 101   Andrea Peterson, ‘Holocaust Museum to Visitors: Please Stop Catching Pokémon Here’ The Washington Post (12 July 2016), www.washingtonpost.com/news/the-switch/wp/2016/07/12/ holocaust-museum-to-visitors-please-stop-catching-Pokémon-here/.

BARFIELD PRINT.indd 591

13/11/2018 11:53

592  Research handbook on the law of virtual and augmented reality playing the game. These and other nonland rights-related legal issues are beyond the scope of this chapter, which instead focuses on the land or property-related rights that may arise. Section 9 will consider the rights, if any, of Pokémon Go players and the developer of the game. Section 10 will expound on the rights of the real world property owners and occupiers, focusing in particular on trespass laws and nuisance laws.

9.  POKÉMON GO DEVELOPER AND PLAYERS The rights of Pokémon Go players over land – as with any such rights, and like those in virtual-world game Second Life – depend on the contract between the players and the developers: the terms of service. Unlike Linden Lab, Niantic made no grand statements about players owning any land-related rights, whether as part of the game, inside the game, or in the real world. The Pokémon Go terms of service (PG ToS) expressly state that users and players have a limited, personal, noncommercial, nonexclusive, nontransferable, nonsublicensable revocable license to all in-game content with no ownership rights. Interestingly, the PG ToS define content to include, without limitation, user content, which is in turn defined to mean any content that players provide to be made available through the game.102 Regarding User Content, Niantic does not claim any ownership rights in any user content, nor does it restrict any rights that players may have to use and exploit their user content. Content itself is broadly defined to mean ‘the text, software, scripts, graphics, photos, sounds, music, videos, audiovisual combinations, interactive features, works of authorship of any kind, and information or other materials that are posted, generated, provided, or otherwise made available through the Services’.103 From the foregoing, it would appear that Niantic is limiting content to the kind of content protected by intellectual property laws and does not intend to make any pronouncements on any kind of land-related rights. This interpretation is confirmed by another clause which states: You acknowledge that you do not acquire any ownership rights in or to the Virtual Money, Virtual Goods, or other Content; any balance of Virtual Goods or Virtual Money does not reflect any stored value. You agree that Virtual Money and Virtual Goods have no monetary value and do not constitute actual currency or property of any type.104

‘Virtual goods’ are defined as virtual items or services that the developer expressly make available for use in the game. It is quite clear from this clause that the developer does not intend to create any kind of property or property rights in the game of Pokémon Go. Hence, with the PG ToS not awarding any property or property rights, there are certainly no land-related rights created or awarded to the players, even if players may battle and obtain control of gyms and PokéStops located at real-world places such as churches and museums.   See the Pokémon Go Terms of Service www.nianticlabs.com/terms/Pokémongo/en/.   Ibid. 104   Ibid. 102 103

BARFIELD PRINT.indd 592

13/11/2018 11:53

Property rights in virtual and augmented reality  593 To round off the discussion, Niantic also evinces the position in the PG ToS that it does not envisage that there are any property or property rights in the virtual items or services, and hence does not claim to hold any land-related rights. Furthermore, another section of the PG ToS expressly states: ‘You agree not to use the App to violate any applicable law, rule, or regulation (including but not limited to the laws of trespass) . . . Without limiting the foregoing, you agree that in conjunction with your use of the App you . . . will not enter onto private property without permission.’105 In this clause, the developer expressly recognises the rights of real-world property owners and occupiers and implores players not to trespass or to enter private property without permission. The developer is in effect upholding the traditional understanding of property ownership and occupancy in the real world. This leaves the final group of people to be considered: those who own land or occupy land in the real physical world.

10.  REAL-WORLD PROPERTY OWNERS AND OCCUPIERS The PG ToS expressly recognises the rights of real world property owners and occupiers and exhorts players not to trespass on private property without permission. It is worth noting too that in the PG ToS, Niantic prohibits players from encouraging or enabling any other individual to violate any applicable rule or regulation. It seems incongruous then for Niantic to encourage players to violate laws and commit torts such as trespass through the placement of gyms and PokéStops on private property or in places where access is restricted. Under the PG ToS, the developer also disclaims all liability related to any property damage, personal injury or death that may occur during the use of their game; it has also categorically stated that they are not liable for the actions of the game’s players.106 This disclaimer may or may not have effect in the game developer’s contractual relationship with its players but it would be ineffective against other parties. This section of the chapter will consider the rights of real-world property owners and occupiers and how these rights may be applicable to augmented reality games such as Pokémon Go. As discussed above, the two key features of property and of land in a number of common law jurisdictions are a presumptive right to exclude others (or control over access) and discretion in the manner of exploitation. Would these rights extend to the augmented reality realm, directly or indirectly? 10.1  Trespass Laws Traditional trespass requires the physical presence on or entry into property without permission. If the developer places a gym or PokéStop on private property, it cannot be said that there is any physical presence as it takes place in the augmented reality realm. Thus, no action of trespass can arise against the game developer unless the right to

  Ibid.   Allison Grande, ‘“Pokémon Go” Maker Says Players’ Trespass Not Its Problem’ (31 January 2017) www.law360.com/articles/886257/Pokémon-go-maker-says-players-trespass-not-its-problem. 105 106

BARFIELD PRINT.indd 593

13/11/2018 11:53

594  Research handbook on the law of virtual and augmented reality exclude others can be extended to the augmented realm; legislators are free to make such pronouncements if there are cogent reasons for doing so. If a player enters onto private property in the course of playing the game, that would certainly be trespass and it is actionable. Under US law, the Restatement (Second) of Torts provides that a person is liable regardless of whether harm was caused.107 Further, in the US, punitive damages can be awarded for trespass.108 The traditional view in Commonwealth jurisdictions is that unless harm has been caused, only nominal damages are recoverable.109 Where harm can be shown, the courts will embark upon an exercise of calculating the harm caused and the compensatory damage.110 Such damages should be based on what sum the parties would, hypothetically, have agreed for the right to use the land. This however is not an easy task, as the landowner would argue the highest amount and the trespasser the lowest. In a recent UK Court of Appeal case, the court held that the assessment of damages for trespass should be based on the amount of a hypothetical licence for the period during which the trespass occurred, and not a permanent hypothetical licence.111 In the case of Plenty v Dillon,112 the High Court of Australia explained the notion of general damages for trespass to land as fulfilling vindicatory purposes:113 True it is that the entry itself caused no damage to the appellant’s land. But the purpose of an action for trespass to land is not merely to compensate the plaintiff for damage to the land. That action also serves the purpose of vindicating the plaintiff’s right to the exclusive use and occupation of his or her land. The appellant is entitled to have his right of property vindicated by a substantial award of damages . . . If the occupier of property has a right not to be unlawfully invaded, then . . . the ‘right must be supported by an effective sanction otherwise the term will be just meaningless rhetoric’.

It appears that the courts have moved to the position that nominal damages may no longer be the norm and trespassers would need to account to the land owners or occupiers, especially if the occupier or land owner is swamped with players trespassing on their property.114 Trespass laws, however, may not have the sort of deterrent effect that would be practically effective or useful, as players may think they cannot be caught, or that they can disappear off the property as soon as the owner or occupier appears to detain them or to identify them.

  Restatement (Second) of Torts § 158 (Am. Law Inst. 1965).   See for example, Jacque v Steenberg Homes, Inc. 563 N.W.2d 154 (Wis. 1997). 109   See however recent developments in the UK with the introduction of the Countryside and Rights of Way Act 2000, 2000 c 37, which is applicable only to certain categories of land. 110   See for example Corlis v Blue Grass Sod Farms Ltd 2016 ABPC 55 (Canada). 111   Eaton Mansions (Westminster) Ltd v Stinger Compania De Inversion SA [2013] EWCA Civ 1308 (30 October 2013). 112   Plenty v Dillon (1991) 171 CLR 635. 113   Plenty v Dillon (1991) 171 CLR 635 at 655. 114   Madeline Bilis, ‘This Church-turned-House Is also Unwillingly a Pokémon Gym’ Boston Magazine (11 July 2016) www.bostonmagazine.com/property/blog/2016/07/11/church-house-Poké​ mon-go-gym/. 107 108

BARFIELD PRINT.indd 594

13/11/2018 11:53

Property rights in virtual and augmented reality  595 10.2  Nuisance Law – US As noted, in the early days of the game there were many complaints about the locations of the gyms and PokéStops, with players converging on and disrupting funeral services in churches and cemeteries;115 incidents of robberies also surfaced.116 In response to the complaints, the developer has since allowed individuals to request online that their property be added and/or removed from the game as a gym or PokéStop.117 While this is certainly a step in the right direction, an individual may make a request but the developer may not necessarily grant the request, or even provide any substantive reply, as epitomised by the class action suit to be discussed presently. Further, the question remains of whether the system should be an opt-out one, as there are many individuals who do not play the game and who would have no idea that their properties have been automatically added as a gym or PokéStop until the real-world disturbances begin. However troublesome it may or may not be, the express consent of property owners or occupiers should be obtained before those properties are designated as gyms or PokéStops. This is so because Pokémon Go, unlike virtual-world games, is an augmented reality game that transcends the virtual world and makes a significant, real-world physical impact. The location of gyms and PokéStops at or near private residences has led to the filing of a number of lawsuits in the United States District Court for the Northern District of California in 2016, consolidated into a single suit.118 The plaintiffs alleged nuisance, trespass, and unjust enrichment, seeking an injunction and other relief against Niantic’s alleged wrongful conduct. Two sets of plaintiffs had made repeated requests that their properties be removed as gyms and PokéStops, but without success – no action regarding this had been taken by the defendants. In fact, the standard experience was that the plaintiffs received an automatic reply stating that it would review the request and appropriate action would be taken but the plaintiffs then heard nothing further from the game developer. All of the plaintiffs alleged that shortly after the release of the game in mid-2016, visitors to their homes and the vicinity increased significantly, from around 15–20 visitors at any given time to at least several hundred, some at odd hours in the morning and evening. The visitors would, inter alia, knock on their doors requesting permission to enter private backyards, park their cars illegally, block driveways, interact at very high volumes that would prevent residents from getting to sleep at night, leave garbage on the property, trespass on the lawns, trample the landscape, peer into the private windows of residences, and use properties as the visitors’ own private toilet late at night, since no public bathroom

115   Paddy Naughtin, ‘Box Hill Cemetery Swamped by People Playing Pokémon Go, as Chief Executive Philip Bachelor Urges People to Be Respectful’, Herald Sun (15 July 2016) www.heraldsun. com.au/leader/news/box-hill-cemetery-swamped-by-people-playing-Pokémon-go-as-chief-executivephilip-bachelor-urges-people-to-be-respectful/news-story/aec7297c9f8afe59faed4​c78ef94bd91/. 116  ‘Pokémon Go: Armed Robbers Use Mobile Game to Lure Players into Trap’ The Guardian (11 July 2016) www.theguardian.com/technology/2016/jul/10/Pokémon-go-armed-robbers-dead-body. 117   Jeff Grub, ‘Pokémon Go: How to Request or Remove a PokeStop or Gym’ https://­venture​ beat.com/2016/07/14/Pokémon-go-how-to-request-or-remove-a-pokestop-or-gym/. 118   Pokémon Go Nuisance Litigation, Case No. 3:16-cv-04300-JD (N.D. Cal. Sept 23 2016) (No 34).

BARFIELD PRINT.indd 595

13/11/2018 11:53

596  Research handbook on the law of virtual and augmented reality facilities were available. One group of plaintiffs even had to hire off-duty police officers to patrol the property in the early hours of the morning.119 While the plaintiffs took legal action on a number of legal grounds, the strongest landrelated rights argument would lie in nuisance. In the US, the Restatement (Second) of Torts defines private nuisance as ‘a non-trespassory invasion of another’s interest in the private use and enjoyment of land’.120 The phrase ‘interest in the private use and enjoyment of land’ is broadly defined to include any disturbance of the enjoyment of property. The comment in the Restatement describes the landowner’s interest protected by private nuisance law as follows:121 The phrase ‘interest in the use and enjoyment of land’ is used in this Restatement in a broad sense. It comprehends not only the interests that a person may have in the actual present use of land for residential, agricultural, commercial, industrial and other purposes, but also his interests in having the present use value of the land unimpaired by changes in its physical condition. Thus the destruction of trees on vacant land is as much an invasion of the owner’s interest in its use and enjoyment as is the destruction of crops or flowers that he is growing on the land for his present use. ‘Interest in use and enjoyment’ also comprehends the pleasure, comfort and enjoyment that a person normally derives from the occupancy of land. Freedom from discomfort and annoyance while using land is often as important to a person as freedom from physical interruption with his use or freedom from detrimental change in the physical condition of the land itself.

From the facts, it seems quite clear that the plaintiffs’ interest in the use and enjoyment of their property has been severely interfered with and significant and unreasonable harm has been caused,122 from numerous intrusions, property damage and threats of physical harm, to costs expended in cleaning up, maintenance and securing their properties. Section 822 further provides as follows: One is subject to liability for a private nuisance if, but only if, his conduct is a legal cause of an invasion of another’s interest in the private use and enjoyment of land, and the invasion is either (a) intentional and unreasonable, or (b) unintentional and otherwise actionable under the rules controlling liability for negligent or reckless conduct, or for abnormally dangerous conditions or activities.

Interestingly, the wording of section 822 specifies that one is liable if one’s conduct is ‘a legal cause’ of an invasion of another’s interest in the private use and enjoyment of land. There is no requirement that one’s conduct must be the only cause, or a direct cause; it only requires that the conduct be one of the legal causes. On this point, it would seem quite evident that the game developer’s placement of gyms or PokéStops at the properties belonging to the plaintiff is one of the legal causes of the harms and interruptions 119   Andrew L Rossow, ‘Gotta Catch . . . a Lawsuit? A Legal Insight Into the Intellectual, Civil and Criminal Battlefield Pokémon Go Has Downloaded onto Smartphones and Properties Around the World’ (2017) 16 J. Marshall Rev. Intell. Prop. L. 329 at 346–7. 120   Restatement (Second) of Torts § 821D (Am. Law Inst. 1965). 121   Restatement (Second) of Torts § 821D, Comment b (Am. Law Inst. 1965). 122   See the recent Texas Supreme Court case of Crosstex North Texas Pipeline, L.P. n/k/a Enlink North Texas Pipeline LP v Gardiner No 15-0049, 2016 WL 3483165 (Tex. June 24, 2016) which emphasised that it is the effects of the substantial interference on the plaintiff that must be unreasonable, not the conduct or land use of the defendant.

BARFIELD PRINT.indd 596

13/11/2018 11:53

Property rights in virtual and augmented reality  597 experienced by the plaintiff. Further, the game developer’s refusal to remove the plaintiffs’ properties as gyms or PokéStops is another legal cause of the damage suffered by the plaintiffs. In other words, if the plaintiffs’ properties were not designated as gyms or PokéStops in the game, none of the strangers would have turned up to cause the harms experienced by the plaintiffs – in this sense, the developer’s conduct could be seen as a direct cause of the damages. Even if it could be a legal argument that, as the game developer, Niantic has no control over the players’ movements, the developer’s conduct would be at minimum an indirect cause, which would still satisfy section 822. Being the developer of the game, Niantic has every power, control and ability to prevent the nuisance from occurring by simply removing those sites as gyms or PokéStops. It is the defendants’ conduct of the placements of gyms or PokéStops that has caused the harms to the plaintiffs. In a sense, Niantic’s placements of Gyms or PokéStops have enticed players to trespass on and/or to cause nuisance to private property, and this enticement is also a cause of the harms. Indeed, it would appear that any contention by Niantic on the nuisance action that it required players to agree that they would not trespass on other people’s property, thereby shielding Niantic from liability, would not succeed.123 Nuisance is actionable irrespective of whether there is trespass – the fact that Niantic’s augmented reality game has brought hundreds of strangers daily to an originally quiet neighbourhood area is irrelevant to the question of trespass. None of the strangers need to be trespassing but since Niantic’s game is the cause of the strangers’ unwanted and disruptive presence, this is the key to the nuisance allegations. The next criteria that needs to be satisfied is that the invasion must be either intentional and unreasonable or unintentional and otherwise actionable under the rules controlling liability for negligent or reckless conduct. The developer of Pokémon Go could fall within either limb, but it would certainly fall under the intentional invasion category as it has been notified many times of the interferences to the enjoyment of land. The unreasonableness element would not be difficult to satisfy given the stated facts outlining harms such as prevention from quiet sleep at night and expenses incurred in hiring security patrols. While Niantic may attempt to argue that it required its users to agree to only engage in lawabiding activities, this might be a step that is simply insufficient, and in reality it is an ineffective clause. The reasonable course of action would be simply to relocate gyms or PokéStops – after all, there are plenty of public spaces available where the developer could locate these. If the defendants attempt to claim that they have done all that is within their power to abate the enjoyment of property interferences, it may be an unconvincing argument given the defendants’ unresponsive customer service. While Niantic created a system in late July 2016 whereby people could request a PokéStop be removed, most requests have reportedly been unanswered.124 Lastly, section 821F needs to be considered. It provides as follows: ‘There is liability for a nuisance only to those to whom it causes significant harm, of a kind that would be

123   Jeffrey N. Rosenthal, ‘Property Owners Continue to say “No” to Pokémon Go’ The Legal Intelligencer (25 April 2017) www.thelegalintelligencer.com/id=1202784505873/PropertyOwners-C​ontinue-to-Say-No-to-Pokeacutemon-GO. 124   Ibid.

BARFIELD PRINT.indd 597

13/11/2018 11:53

598  Research handbook on the law of virtual and augmented reality suffered by a normal person in the community or by property in normal condition and used for a normal purpose.’ This requirement is also satisfied given the set of facts. The normal person in the community living in a once quiet residential neighbourhood would not under normal conditions have to suffer the significant harms and intrusions caused by hundreds of strangers congregating outside their homes at all hours of the day and night on a daily basis. Worthy of note are the developments in the state of Illinois; there, property owners have turned to a more cost-effective route of asking the legislature to address the problem. Kelly Cassidy introduced a bill, the Location-Based Video Game Protection Act, HB 6601, which, if passed, would allow a real property owner to bring a civil action; Niantic would be liable to a fine of up to $100 per day for not removing a PokéStop if the request was not met within two business days of receiving the request.125 10.3  Nuisance Law – Australia and the UK In other common law jurisdictions, the tort of private nuisance is conceived with similar concepts and requirements. The law of private nuisance in Australia is by and large similar to that of the UK, with much of Australia’s law in the area having followed the UK caselaw. In Australia, private nuisance arises when a party substantially and unreasonably interferes with, or disturbs someone else’s ordinary and reasonable use of the land they own or occupy.126 As with the law in the US, the interference occurs without entry onto the plaintiff’s land. For a successful action in private nuisance, the party taking legal action must have a legal interest in the land, such as being an owner of the land, or have a right to occupy or exclusively possess the land, such as being a tenant. The plaintiff must show that the defendant has interfered with the use and enjoyment of that land and that the inference was both substantial and unreasonable. The question of whether the interference was substantial and unreasonable will depend on the circumstances of each case.127 Interference is generally only considered to be substantial where it is more than that which the average person in that neighbourhood can reasonably expect in the circumstances. The courts will examine a variety of factors such as the frequency, extent and nature of the interference.128 Inconvenience to a sensitive or fussy person will not constitute a nuisance.129 The duration and timing of the nuisance will also be relevant; for example, noise in the middle of the night or at other quiet times will be more likely to be found to be a nuisance.130 The location of the land will play a role in determining the ordinary use

125   Location-Based Video Game Protection Act, HB 6601 www.ilga.gov/legislation/BillStatus. asp?DocTypeID=HB&DocNum=6601&GAID=13&SessionID=88&LegID=98270. 126   Grand Central Car Park Pty Ltd v Tivoli Freeholders [1969] VR 62 at 72 per McInerney J (public nuisance); Sedleigh-Denfield v O’Callaghan [1940] AC 880 at 896–7 per Lord Atkin; Elston v Dore [1982] HCA 71; (1982) 149 CLR 480 (14 December 1982). 127   Halsey v Esso Petroleum Co Ltd [1961] 2 All ER 145 at 151, 155 per Veale. 128   Oldham v Lawson (No 1) [1976] VR 654 at 655 per Harris J. 129   Halsey v Esso Petroleum Co Ltd [1961] 2 All ER 145. 130   Harrison v Southwark and Vauxhall Water Co [1891] 2 Ch 409 at 414 per Vaughan Williams J.

BARFIELD PRINT.indd 598

13/11/2018 11:53

Property rights in virtual and augmented reality  599 of land in the area; thus, an emission of fumes in a residential area would be more likely to be considered a nuisance than if fumes were emitted in an industrial area.131 The fault standard is, first, that the defendant knew or ought to have known of the nuisance; hence, a person may be liable where a nuisance was created without their knowledge. Second, the interference or damage from the nuisance must be reasonably foreseeable; last, the defendant did not take reasonable action or steps to end the nuisance.132 The traditional thinking in Australia and in the UK has been that the private nuisance action has been conceived as actions from the enjoyment and use of one piece of land against another piece of land. Most of the cases have been between neighbours in the popular sense of the word, whether the defendant owned or occupied the land. There is however, precedent that the creator of the nuisance can be held liable even if the creator does not have any interest in the land from which the nuisance emanates, nor does nuisance need to arise on private land. In the House of Lords decision of Esso Petroleum Co Ltd v Southport Corporation,133 Lord Devlin aptly summed up the common law position as follows:134 It is clear that to give a cause of action for private nuisance the matter complained of must affect the property of the plaintiffs. But I know of no principle that it must emanate from land belonging to the defendant. Mr. Nelson cited Cunard v. Antifyre Ltd. and I think that the statement of the principle is put there as clearly and concisely as it can be. Talbot J. said: ‘Private nuisances, at least in the vast majority of cases, are interferences for a substantial length of time by owners or occupiers of property with the use or enjoyment of neighbouring property; and it would manifestly be inconvenient and unreasonable if the right to complain of such interference extended beyond the occupier, or (in case of injury to the reversion) the owner, of such neighbouring property.’ It is clear from that statement of principle that the nuisance must affect the property of the plaintiff; and it is true that in the vast majority of cases it is likely to emanate from the neighbouring property of the defendant. But no statement of principle has been cited to me to show that the latter is a prerequisite to a cause of action; and I can see no reason why, if land or water belonging to the public, or waste land, is misused by the defendant, or if the defendant as a licensee or trespasser misuses someone else’s land, he should not be liable for the creation of a nuisance in the same way as an adjoining occupier would be.

There has been no direct pronouncement on this issue from any Australian court, but given that Australia shares a common law heritage with the UK, it would appear that Lord Devlin’s insights are equally applicable in Australia. From the foregoing, if the same facts were played out in Australia, applying the law on private nuisance in Australia would likely see the game developer being held liable for private nuisance. The location of the gyms and PokéStops has in fact caused the plaintiffs’ ordinary and reasonable use of the land which they own or occupy to be substantially and unreasonably interfered with. The fact that the frequency of the visitors increased from less than 20 to several hundreds and that it occurred at night in a residential area that was quiet before the launch of the augmented reality game adds weight to the unreasonableness and substantiality of the interferences. The game developer has clearly enticed     133   134   131 132

Halsey v Esso Petroleum Co Ltd [1961] 2 All ER 145. Robson v Leischke [2008] NSWLEC 152 per Preston CJ. [1956] AC 218. Ibid at pp. 224–5, footnotes in the original omitted.

BARFIELD PRINT.indd 599

13/11/2018 11:53

600  Research handbook on the law of virtual and augmented reality players to the vicinity of plaintiffs’ properties through the game. The defendant game developer knew of the nuisance; further, the interference or damage from the nuisance was reasonably foreseeable as the game reached unprecedented high levels of popularity within its first week of the launch; hence, it was reasonably foreseeable that large crowds of strangers would congregate at gyms and PokéStops, causing all the harms they did. Lastly, the defendant did not take reasonable action or steps to end the nuisance, even after repeated requests by the plaintiffs to remove their properties as locations of gyms and PokéStops. In the UK, the Supreme Court decision in Coventry v. Lawrence reaffirmed the basic principles constituting the tort of private nuisance.135 In the words of Lord Neuberger:136 A [private] nuisance can be defined, albeit in general terms, as an action (or sometimes a failure to act) on the part of a defendant, which is not otherwise authorised, and which causes an interference with the claimant’s reasonable enjoyment of his land, or to use a slightly different formulation, which unduly interferes with the claimant’s enjoyment of his land. As Lord Wright said in Sedleigh-Denfield v O’Callaghan . . . ‘a useful test is perhaps what is reasonable according to the ordinary usages of mankind living in society, or more correctly in a particular society’.

Lord Neuberger went on to reaffirm that that reasonableness in this context is to be assessed objectively.137 The decision in Coventry v. Lawrence did not change the fundamental understanding of what constitutes private nuisance.138 It did, however, clarify the law on five points, as follows. These five points were particular to the set of facts before the court and none of them are relevant to the circumstances that arise from Pokémon Go. First, the court considered the question of the extent to which it is open to a defendant to contend that he has established a prescriptive right to commit what would otherwise be a nuisance.139 This is not relevant to Niantic as Pokémon Go was launched less than a year before the legal suits commenced. Second, the court dealt with the merits of the contention that the claimant ‘came to the nuisance’.140 This again is irrelevant to Niantic as all of the claimants in the US class action already owned or occupied the properties before Pokémon Go was launched. Third, the court clarified the extent to which the use to which the defendant actually puts his property can be relied on when assessing the character of the locality for the purpose of assessing whether the claimant has made out the case that those activities constitute a nuisance.141 In Coventry v. Lawrence, the defendant property had been used, on and off, for similar activities for around 30 years.142 This issue is, again, by and large irrelevant to the case of Pokémon Go. In any event, for completeness, the court was of the opinion that the defendant’s activities are to be taken into account when assessing the character of the locality.143 However, such activities should only be considered to the extent that they would     137   138   139   140   141   142   143   135 136

[2014] UKSC 13. Ibid at para 3. Ibid at para 5. [2014] UKSC 13. Ibid at para 6. Ibid at para 6. Ibid at para 6. [2014] UKSC 13. Ibid at para 63.

BARFIELD PRINT.indd 600

13/11/2018 11:53

Property rights in virtual and augmented reality  601 not cause a nuisance to the claimant;144 thus, if the activities cannot be carried out without creating a nuisance, such activities will have to be entirely disregarded when assessing the character of the locality as it would be ‘illogical’ and ‘unfair’ to the plaintiff to take them into account.145 Fourth, the court examined the role and impact of planning permission for a particular use in the locality.146 Again, this is irrelevant to Niantic. Last, the court clarified the approach to be adopted by a court when deciding whether to grant an injunction to restrain a nuisance from being committed, or whether to award damages instead, and the relevance of planning permission in the approach.147 This question is relevant to the available remedies but as the discussion at hand is mainly concerned with the property-related rights that arise, it suffices to note that all the justices felt this to be a discretion that lies with the court. Lord Neuberger, along with the majority, would accept that the prima facie position is that an injunction should be granted.148 In summary, it would appear that under the law of private nuisance in Australia and that in the UK, Niantic would need to answer for interferences in the use and enjoyment of land amounting to private nuisance.

11. CONCLUSION Virtual worlds such as Second Life have become increasingly significant to their users in terms of both time and money. As such, it is important to analyse how the law may apply to and resolve disputes that originate in these virtual worlds. Linden and the users of Second Life have already imported real-world concepts, such as currency and economy, into Second Life and significant player expectations have already arisen. These have also, in part, been a result of Linden’s public representations regarding player ownership. This chapter has argued that land ownership in Second Life is very much like owning a modified form of leasehold property. Just as in the real world more than one type of property right can subsist in a given item, so this should also be the case in Second Life. So, for example, a paperback book will have property rights attached to the physical thing of the book, but there will also be intellectual property rights over the content. Similarly, an owner of land in Second Life may, for example, have intellectual property rights over the contents of the land as an artistic work, but in addition the owner of the land will have proprietary rights and interests over the land and use of the land itself. Some Asian countries such as China and Taiwan have already recognised property rights in virtual items such as virtual swords,149 and it would appear to be the next logical step to recognise

  Ibid at para 65.   Ibid at para 65–6. 146   Ibid at para 6. 147   Ibid at para 6. 148   Ibid at para 121. 149   For China, see the cases of Li Hong Chen v Beijing Arctic Ice Technology Development (2004) 二中民终字第02877号, online: www.chinacourt.org/public/detail.php?id=143455 and a case in the Chaoyang District of Beijing in December 2003 reported in ‘Video Game: Real Murder’ China Economic Review (26 October 2006). For Taiwan, see Art. 323 of the Taiwan Criminal Code (1997) and Taiwan Ministry of Justice Official Notation No. 039030 (90). 144 145

BARFIELD PRINT.indd 601

13/11/2018 11:53

602  Research handbook on the law of virtual and augmented reality virtual land as having proprietary rights attached. Under the system of law in Singapore, the United Kingdom and Australia, there is no reason why this should not be the case. The only augmented reality game that has thus far gained meteoric popularity internationally is Pokémon Go. It has also, unfortunately, caused much distress to land owners and occupiers all over the world due to the placement of the gyms and PokéStops. These intrusions and interferences to the use and enjoyment of property arguably amount to actionable torts against the developer of Pokémon Go, Niantic. This is unsurprising given that Pokémon Go, unlike virtual-world games, is an augmented reality game that transcends the virtual world and makes a significant, real-world physical impact. For these unreasonable and substantial real-world physical impacts, the developer will need to be held accountable.

SOURCES Australia Copyright Act 1968 (Cth.), s. 92 (Australia) Commissioner of Taxation v Orica Ltd [1998] 194 CLR 500 (H.C.A.) Elston v Dore [1982] HCA 71; (1982) 149 CLR 480 (14 December 1982) Grand Central Car Park Pty Ltd v Tivoli Freeholders [1969] VR 62 Gutnick v Dow Jones Co, Inc, [2001] Vic. S.C. 305 Oldham v Lawson (No 1) [1976] VR 654 Plenty v Dillon (1991) 171 CLR 635 R v Toohey; Ex parte Meneling Station Pty Ltd (1982), 158 C.L.R. 327 (H.C.A.) Robson v Leischke [2008] NSWLEC 152 Unimin Pty Ltd v Commonwealth (1974), 22 F.L.R. 299 Western Australia v Ward, [2002] 213 C.L.R. 1 (H.C.A.) Yanner v Eaton [1999] 201 C.L.R. 351 (H.C.A.)

Canada Corlis v Blue Grass Sod Farms Ltd 2016 ABPC 55 (Canada)

China Li Hong Chen v Beijing Arctic Ice Technology Development (2004) 二中民终字第02877号

Singapore Housing and Development Act (Cap. 129, 2004 Rev. Ed. Sing.)

Taiwan Art. 323 of the Taiwan Criminal Code (1997) Taiwan Ministry of Justice Official Notation No 039030 (90)

United Kingdom Countryside and Rights of Way Act 2000, 2000 c 37

BARFIELD PRINT.indd 602

13/11/2018 11:53

Property rights in virtual and augmented reality  603 Coventry v. Lawrence [2014] UKSC 13. Eaton Mansions (Westminster) Ltd v Stinger Compania De Inversion SA [2013] EWCA Civ 1308 (30 October 2013). Esso Petroleum Co Ltd v Southport Corporation [1956] AC 218. Halsey v Esso Petroleum Co Ltd [1961] 2 All ER 145 Harrison v Southwark and Vauxhall Water Co [1891] 2 Ch 409 National Provincial Bank v Ainsworth [1965], A.C. 1175 (H.L.) Sedleigh-Denfield v O’Callaghan [1940] AC 880 Sole v Secretary of State for Trade and Industry [2007] EWHC 1527 (Admin)

United States Bragg v Linden Research, Inc., 487 F. Supp. 2d 593 (E.D. Penn. 2007) Crosstex North Texas Pipeline, L.P. n/k/a Enlink North Texas Pipeline LP v Gardiner No. 15-0049, 2016 WL 3483165 (Tex. June 24, 2016) Jacque v Steenberg Homes, Inc. 563 N.W.2d 154 (Wis. 1997) Restatement (Second) of Torts §§ 158, 821D, 821F, 822 (Am. Law Inst. 1965) Restatement (Second) of Torts § 821D, Comment b (Am. Law Inst. 1965) Unfair and Deceptive Practices Act, Cal. Bus. & Prof. Code § 17200 Unfair Trade Practices and Consumer Protection Law, 73 P.S. (1968)

BARFIELD PRINT.indd 603

13/11/2018 11:53

18.  Freedom of contract in augmented reality Scott R. Peppet*

1. INTRODUCTION In the past decade, digital information has invaded our experience of physical space. Whether on a smartphone, tablet, connected watch, smartspeaker, or other device, this surge in ubiquitous internet access has opened the door to a range of mobile applications designed to saturate our daily experiences with previously unavailable information. In this increasingly “augmented reality,”1 a consumer shopping in a brick and mortar retailer can learn instantly about a product’s capabilities, safety record, environmental f­ riendliness— or contract terms. Use your phone’s camera to scan the bar code on a potential purchase, and Amazon or Consumer Reports will instantly return price comparisons and other consumers’ reviews. Enter a merchant’s name or enable your smartphone’s location capabilities, and the Better Business Bureau’s mobile app will return reviews and ratings of that seller or other local businesses, grade each merchant on a scale of A+ to F, and indicate which merchants have received a Better Business Bureau accreditation. Enter the Vehicle Identification Number (VIN) of the used car you are test driving, and CarFax will tell you whether you are buying a lemon. Our augmented reality can provide information about more than just goods, firms, and contract terms: We also have access to information about each other. Consider a startling example. In July 2011, the privacy economist Allessandro Acquisti demonstrated a prototype of an iPhone application that could take a photograph of a person, compare that photograph to the millions of public profile pictures available through Facebook, and, if a match were found, guess the person’s name and the first five digits of that person’s social security number.2 The application succeeds a staggering 30 percent of the time. Acquisti notes that “[t]he application is an example of augmented reality, in which offline and online data blend together . . . In addition to its privacy implications . . . the age of augmented reality . . . may carry even deeper-reaching . . . implications.” Although privacy *  This chapter is based on Scott R. Peppet, Freedom of Contract in an Augmented Reality: The Case of Consumer Contracts, 59 UCLA L Rev 676 (2012). 1   I will generally use the term augmented reality (AR) to describe the mobile, pervasive computing technologies that are rapidly bringing together physical and digital space. The term was coined by Tom Caudell and David Mizell. See generally Thomas P. Caudell & David W. Mizell, Augmented Reality: An Application of Heads Up Display Technology to Manual Manufacturing Processes, Proc. of the IEEE Twenty-Fifth Hawaii Int’l Conf. on Sys. Sci. 659 (1992). 2   See Allessandro Acquisti, Ralph Gross, & Fred Stutzman, Privacy in the Age of Augmented Reality 5 (working draft) (on file with author) (“To illustrate the possibility of real-time identification, we developed a smart phone demo application that captures the image of a person and then overlays on the screen her predicted name and SSN”); see also Julia Angwin, Face-ID Tools Pose New Risk, Wall St. J., Aug. 1, 2011, http://online.wsj.com/article/SB1000142405311190334140457 6480371062384798.html (describing Acquisti et al’s study).

604

BARFIELD PRINT.indd 604

13/11/2018 11:53

Freedom of contract in augmented reality  605 scholars have considered the implications of facial recognition technologies, what might these technologies imply for contract? What will it mean when merchants can identify each consumer, search for information about that consumer, and tailor the shopping experience or their standard form contract based on that consumer’s characteristics? This chapter considers the implications of these developments for contract law, and, particularly, for freedom of contract. Consider the classic case Williams v. WalkerThomas Furniture Co.,3 in which the DC Circuit famously held that a contract could be unconscionable if there was “an absence of meaningful choice on the part of one of the parties together with contract terms which are unreasonably favorable to the other party.”4 From 1957 to 1962, the Walker-Thomas furniture store had extended credit to Ms Williams for the purchase of various pieces of furniture. In each instance, the contract contained a crosscollateralization provision making each purchased item collateral for every other item; no item could be paid off completely until all had been paid off. In 1962, Ms Williams purchased a stereo for $514; when she defaulted on her payments, the store sought to repossess many of her earlier purchases as well. The court found the clause unenforceable, in part because Ms Williams was not sufficiently informed about the collateralization clause. For a moment, however, imagine that Walker-Thomas had occurred today. How might the facts and outcome have been different? What if Ms. Williams had been standing in the furniture store with her smartphone, and could easily have searched for information on the store’s reputation or the fairness of its contracts? What if she had used the Pissed Consumer app on her iPhone to learn that the furniture store had received five red stars, indicating a high volume of consumer complaints about its contracts;5 or if her GPSequipped phone had automatically notified her when she entered the store that the furniture seller had been branded “consumer unfriendly” and its location electronically tagged by a local consumer group; or if her social network had left location-based electronic notes warning her about the store’s practices? Moreover, what if the written contract itself had contained a quick response (QR) code in the margins next to the collateralization clause that, when scanned by her phone’s camera, played a short video explaining the clause to Ms Williams in layman’s terms? Would any of these quite plausible modern scenarios have changed the outcome of this seminal contracts case, the unconscionability doctrine, or, more generally, the contours of freedom of contract? These are no longer the speculations of science fiction. These are today’s technologies, and they create pressing questions for contract law. This chapter focuses on the impact of augmented reality on the interpretation and enforcement of contracts, and particularly on courts’ willingness to police consumer contracts for unfair or onesided terms. It explores this topic using the case of consumer standard form contracts—like the contract in Williams v. Walker-Thomas—because in the past decades contract scholars have argued   350 F.2d 445 (D.C. Cir. 1965).   Id at 449; see also Arthur Leff, Unconscionability and the Code—The Emperor’s New Clause, 115 U. Pa. L. Rev. 485, 552–4 (1967) (drawing the distinction between procedural and substantive unconscionability). 5   Pissedconsumer.com is a popular website that allows consumers to review products and merchants; it also provides a mobile application for use on smartphones. See Pissed Consumer, www.pissedconsumer.com. 3 4

BARFIELD PRINT.indd 605

13/11/2018 11:53

606  Research handbook on the law of virtual and augmented reality extensively about freedom of contract in this context.6 It is part of a nascent body of legal scholarship grappling with the implications of information technology developments for markets, law, and society. These implications are profound. Welfare economics generally assumes that informed and rational decisionmakers will reach efficient, welfare-maximizing contracts in competitive markets.7 Contract scholars have long argued about whether these two conditions hold in the consumer context—whether, in short, consumers are sufficiently informed and rational to read and understand standard form contracts. Classical law and economics argues that they are, and that markets will therefore discipline firms to only offer reasonable standard form consumer contracts. Some disagree: Asymmetric information arguments for distrusting consumer contracts claim that consumers have insufficient information for markets to correct standard form contracts; bounded rationality arguments for policing consumer contracts claim that consumers are insufficiently rational to do so.8 Together, these two arguments have justified judicial scrutiny of consumer standard form contracts in the past decades using unconscionability and other similar contract doctrines. This chapter makes three claims relevant to these debates. First, augmented reality profoundly changes the transaction cost economics of consumer contracting. In particular, changes in information technology have radically lowered the costs of both sorting and signaling—the two economic means by which consumers and firms choose and learn about their contracting partners9—and have made new forms of sorting and signaling possible in traditional brick and mortar transactions. Section 2 makes this argument by briefly introducing augmented reality technologies. It then reviews a typology of the types of information these technologies can deliver to today’s consumers: information about places, goods, people, firms, and contract terms. This typology provides many examples demonstrating that in an augmented reality, consumers and sellers can sort and signal each other more effectively and at lower cost than they could in the traditional analog economy. Second, these decreased sorting and signaling costs may at least partially undermine asymmetric information arguments for distrusting consumer standard form contracts. In an augmented reality, consumers can more easily sort firms and firms’ contracts using proxies of contract quality such as consumer reviews, thereby bypassing firms with

6   See Melvin Aron Eisenberg, The Limits of Cognition and the Limits of Contract, 47 Stan. L. Rev. 211, 240 (1995) (“The problems raised by the use of form contracts have been a major preoccupation of contract law scholars for the past forty years”). 7   See, e.g., Richard A. Posner, Economic Analysis of Law 3 (8th ed. 2011); Benjamin E. Hermalin, Avery W. Katz, & Richard Craswell, Contract Law § 2.2.1, in The Handbook of Law & Economics (A. Mitchell Polinsky & Steven Shavell eds, 2007) (discussing core principles of the welfare economics of contract). 8   See Robert Cooter & Thomas Ulen, Law & Economics 231 (5th ed. 2008) (“The farther the facts depart from the ideal of perfect rationality and zero transaction costs, the stronger the case for judges’ regulating the terms of the contract by law”). 9   Sorting and signaling are the two ways in which we overcome information asymmetries. For an overview of sorting and signaling, see John G. Riley, Silver Signals: Twenty-Five Years of Screening and Signaling, 39 J. Econ. Literature 432 (2001); see also Michael Spence, Informational Aspects of Market Structure: An Introduction, 90 Q.J. Econ. 591, 592 (1976) (“[Signaling and sorting] are opposite sides of the same coin”).

BARFIELD PRINT.indd 606

13/11/2018 11:53

Freedom of contract in augmented reality  607 oppressive or onesided contract terms (and creating market pressure on such firms to reform their contracts). When consumers have mobile internet applications that search more than 40 million consumer reviews on more than 4 million products from the aggregated websites of more than one thousand retailers and consumer review services—as they do today using the SearchReviews mobile app10—firms are less likely to use onesided or oppressive contract terms. This makes it more likely that standard form consumer contracts will be fair, and thus courts in an augmented reality are more likely to enforce such contracts as written. Section 3 presents this argument, but also considers possible complications to it. Third, the information available in an augmented reality may help to “debias”11 consumers and make their decisions more rational. Section 4 argues that decreased sorting and signaling costs challenge the bounded rationality arguments that behavioral law and economics scholars have made in the consumer contracting context. In particular, augmented reality can provide consumers with highly salient available information about contractual risks at the point of sale. This makes it more likely that consumers will evaluate contracting decisions rationally, which in turn suggests that, in an augmented reality, courts may more often enforce consumer contracts as written. Taken together, these three claims support the broader conclusion that technological change that increases the quantity and quality of information available to contracting parties is likely to strengthen traditional approaches to freedom of contract in the context of consumer contracts. This chapter thus concludes by making a general assertion about contract theory: (1) the contours of what we consider freedom of contract are fundamentally shaped by the availability of information; and (2) the availability and structure of information in the economy is not a constant but is instead dependent on the state of evolution of the economy’s information architecture. Debates that assume some intrinsic nature of freedom of contract, autonomy, consent, subjugation, and bargaining power are fundamentally contingent. These concepts change as information economics change. Our historical and technological moment begets one conception of the contours of freedom of contract; other moments will beget other conceptions. To me this seems obvious, and yet there is remarkably little in contract theory that acknowledges the reality of technological and historical change.12 Instead, certain technological and economic facts are generally taken as given, and debate ensues based on the faulty premise that those givens will remain constant into the future. This exploration of how augmented reality may change contract law demonstrates that contract scholars must acknowledge the technological and historical contingency of their arguments, particularly in this era of rapid technological change.

10   See SearchReviews, www.searchreviews.com (providing a mobile barcode-based application that searches a massive aggregated data set of consumer reviews). 11   On debiasing generally, see Christine Jolls & Cass R. Sunstein, Debiasing through Law, 35 J. Legal Stud. 199, 210 (2006). 12   For a notable exception, see Avery Weiner Katz, Is Electronic Contracting Different?: Contract Law in the Information Age 1 (unpublished paper) (“[T]he growth of electronic commerce reflects changes in the relative importance of various institutional transaction costs. . . . Accordingly, arrangements that were optimal  . . .  under previous configurations of transaction costs may no longer be so under configurations that will develop in the future.”).

BARFIELD PRINT.indd 607

13/11/2018 11:53

608  Research handbook on the law of virtual and augmented reality

2.  SORTING AND SIGNALING IN AN AUGMENTED REALITY Each time you watch professional sports on television, you experience augmented reality technologies: the yellow line drawn across a football field showing yardage to a first down; the strike box showing a baseball pitcher’s virtual target; the flags of Olympic swimmers seeming to shimmer in their pool lanes. These technologies are everywhere and are increasingly powerful. This section introduces the augmented reality and ubiquitous computing technologies that are changing the economy’s information architecture. Although I hesitate to begin with a technological primer, legal audiences are sometimes unaware of the rapid advances in and deployment of these technologies. It is impossible to consider the ramifications of augmented reality for contract law without understanding the state and pace of technological innovation. This section also presents a typology of the information that augmented reality technologies make available to consumers. This information is ultimately more important than which specific technological delivery method becomes popular. Put differently, whether we a­ ugment our experience of physical space via an iPhone or digital eyeglasses, for contract what matters most is what information those technologies can transmit to transacting consumers and firms, and how that information allows consumers and firms to sort and signal at lower cost. 2.1  An Introduction to Augmented Reality Technologies The retinal displays of the Terminator movies and the augmented reality glasses in Minority Report made for good science fiction, but these technologies are no longer as futuristic as they seemed even a decade ago. This section explores the history of augmented reality, the state of today’s technologies, and what the near future promises in terms of state-of-the-art delivery methods. It is important to stress that today’s smartphones are sufficiently advanced to raise the doctrinal questions at issue in this chapter. Put differently, further technological advance is not necessary for my argument to matter—the technologies in today’s pockets suffice. Nevertheless—and at the risk of sounding too much like science fiction—it is worth briefly exploring the information delivery methods currently available or under development. 2.2  A Brief History Augmented reality has been developing for decades. In 1968, Ivan Sutherland created a working prototype of the first augmented reality system.13 The system consisted of a helmet with a digital display that the user could wear to look around the room and see digital information overlaid on the physical world. In the first prototypes, the headmounted display was so heavy that it had to be suspended from the ceiling—leading to the system’s nickname of “the Sword of Damocles.”14

13   See Ivan E. Sutherland, A Head-Mounted Three-Dimensional Display, Proc. Fall Joint Comp. Conf. 757 (1968). 14   See Stephen Cawood & Mark Fiala, Augmented Reality: A Practical Guide 2 (2007) (explaining the origins of the “Sword of Damocles” nickname).

BARFIELD PRINT.indd 608

13/11/2018 11:53

Freedom of contract in augmented reality  609 The goal of early experimental systems was to bring digital information to the user’s experience of the physical world. Experimental augmented reality systems included outdoor navigation systems for the visually impaired;15 various backpack-based systems combining head-worn displays, location awareness, and computational ability;16 guided campus tours using mobile augmented reality;17 and battlefield information augmentation systems.18 These systems, developed primarily in the 1990s, seem crude today, but they demonstrated the potential for using digital information to augment physical experience. In addition to this early work on augmented reality display systems, an intertwined development was occurring simultaneously: the conception and beginnings of ubiquitous computing. Mark Weiser, a researcher at the Xerox Palo Alto Research Center, coined the phrase in a brief essay titled “Ubiquitous Computing #1.” In that and other papers, Weiser developed the concept of “invisible computing” distributed throughout the environment. Rather than massive mainframe computers or even personal computers on desktops, Weiser imagined small, lower-power, but plentiful computers embedded in everyday objects and capable of networking to provide information to individuals in a decentralized manner. As Weiser put it, “The most profound technologies are those that disappear. They weave themselves into the fabric of everyday life until they are indistinguishable from it.”19 Finally, mobile communications technology simultaneously developed quite independently. The first smartphone was introduced by IBM and Bellsouth in 1992—it served as a phone, calculator, address book, fax machine, and email device; had one megabyte of memory; and cost $900.20 The Global Positioning System (GPS) came online in 1993; the first mobile phones with GPS and internet connectivity arrived in 1999; and the first commercial camera phone was sold in 2000. Mobile technology advanced and spread throughout the 2000s, and by 2008, mobile developers were beginning to present phonebased augmented reality applications.21 2.3  Today’s Delivery Methods These developments have come together to augment physical reality with digital information. In the decade just past, we entered a new phase of the computer revolution. Now,

15   See Jack M. Loomis et al, Personal Guidance System for the Visually Impaired using GPS, GIS, and VR Technologies, Proc. Virtual Reality & Persons With Disabilities Conf. (1993). 16   The Tinmith backpacks were early pioneers, as was the Touring Machine. See Steve Feiner et al, A Touring Machine: Prototyping 3D Mobile Augmented Reality Systems for Exploring the Urban Environment, Proc. of the First IEEE Int’l Symp. on Wearable Computers 74 (1997). 17   See, e.g., Tobia Hollerer et al, Situated Documentaries: Embedding Multimedia Presentations in the Real World, Proc. of the Third IEEE Int’l Symp. on Wearable Computers 79 (1999) (demonstrating an augmented reality system contained in a backpack). 18   See Simon Julier et al., BARS: Battlefield Augmented Reality System, NATO Info. Sys. Techniques Symp. on “New Info. Processing Tech. for Military Systems” (2000). 19   Mark Weiser, The Computer for the 21st Century, 3 Mobile Computing & Commun. Rev. 3, 3 (1999). 20   See IBM Simon, Wikipedia, http://en.wikipedia.org/wiki/Simon_(phone). 21   See, e.g., Daniel Wagner et al, Pose Tracking from Natural Features on Mobile Phones, Proc. of the Seventh IEEE Int’l Symp. on Mixed & Augmented Reality 125 (2008).

BARFIELD PRINT.indd 609

13/11/2018 11:53

610  Research handbook on the law of virtual and augmented reality most of us carry at least one device—phone, tablet, or computer—capable of constant connection to the internet, impressive computational power, and location awareness. Our devices know where we are, which direction we are moving, and what is around us. This not only allows these devices to instantly search for information, but also allows for applications that “push” context-relevant information to us based on our location, our proximity to other people, our social network, or other criteria. An iPhone, iPad, or Android user can use Yelp or Google Maps, for example, to virtually look down a street and see which restaurants have good reviews or what digital information is available about local landmarks. As these technologies continue to develop, consumers will increasingly have instant access to information about the products and services around them, without having to search for it. Instead, such data will be made available as the consumer’s augmented reality device—smartphone, tablet, or otherwise—recognizes products, retrieves relevant information about those products, and shares that information with the consumer. Search is no longer the paradigm. Google’s Eric Schmidt has emphatically asserted that in the near future, search will not be Google’s business; rather, pushing relevant information to users in real time will be more important.22 Although many first generation applications have been crude or gamelike, technologists understand that the potential of augmented reality is not in whizz-bang visuals but in providing consumers with relevant information in real time—information that in many cases has not previously been available without incurring significant costs. 2.4  Tomorrow’s Delivery Methods In the past few years, these components have come together in millions of smartphones, tablets, and other devices. As Mike Kuniavsky has stated, “we are at the beginning of the era of computation and data communication embedded in, and distributed through, our entire environment.”23 Suddenly, augmented reality has moved out of the laboratory and into the market. As discussed previously, additional developments in delivery methods are unnecessary to the doctrinal discussion that follows in sections 3 and 4; today’s smartphones suffice. It is worth pausing, however, to briefly consider several technologies that exist but are not yet available to consumers. These delivery methods are truly futuristic, but they illustrate the likely directions of augmented reality’s ongoing evolution. Moreover, they illustrate the important point that although today’s smartphones already raise questions for contract law, they are not the terminus of augmented reality’s development. Far from it—today’s smartphones are relatively crude augmented reality devices just beginning to bring digital information to physical space. Although we cannot predict with accuracy which information technologies will succeed in the future, we can predict with confidence 22   See Holman W. Jenkins, Jr., Google and the Search for the Future, Wall St. J. (Aug. 14, 2010), at A9 (quoting Schmidt as saying, “[O]ne idea is that more and more searches are done on your behalf without you needing to type. I actually think most people don’t want Google to answer their questions . . . They want Google to tell them what they should be doing next”). 23   Mike Kuniavsky, Smart Things: Ubiquitous Computing User Experience Design 3 (2010); see also id at 10 (“By 2005 . . . the era of ubiquitous computing had begun”).

BARFIELD PRINT.indd 610

13/11/2018 11:53

Freedom of contract in augmented reality  611 that these technologies will continue to develop, to become more sophisticated, and to provide consumers with information in more useful, salient, and seamless ways. Various display technologies under development promise to integrate augmented reality into daily life. For example, several companies are working on projecting information directly onto car windshields and windows. Recent innovations include heads-up vehicle navigation displays that project maps and information directly onto the windshield, creating the sensation that driving directions are on the road in front of the driver.24 Similarly, Toyota is developing car windows that give backseat passengers the ability to interact with the passing outside world by drawing on and querying the window with their fingers.25 A child could touch a passing tree, for example, and the window might display the word “tree” to teach reading, or they might touch a passing building to query the window for the building’s distance from the car. Wearable displays are also in development. You can currently purchase ski goggles that will project GPS and weather information directly to you as you ski,26 and somewhat odd-looking sunglasses with embedded augmented reality displays capable of overlaying digital information on your natural field of view. Although Google Glass has faded from public consciousness, various firms are working to develop regular eyeglasses that can project digital information to their wearer, and researchers are even trying to embed display technologies in contact lenses.27 The ability to access digital information—email, instant messages, walking or driving directions, lecture notes, product information, and so on—directly through your eyeglasses would obviously bring augmented reality to a dramatically different point than being forced to use your smartphone. With Facebook’s purchase of the Oculus Rift technology, these trends only continue to accelerate. 2.5  Sorting and Signaling across Information Types Having succumbed to the temptation to peer into the (near) future, let us return to more humble topics. Before we turn, in sections 3 and 4, to the impact these technologies may have on contract doctrine, this subsection categorizes the types of information they can deliver to consumers. In other words, rather than reviewing the technologies themselves—the glasses or contact lenses or heads-up displays—we briefly consider a typology of the information those technologies can now transmit to each of us, and the ways those technologies change consumers’ and firms’ ability to sort and signal each other as they contract.

24   See Vehicle Displays: Head Up Displays, MicroVision, www.microvision.com/vehicle_­ displays/head_up_displays.html (demonstrating MicroVision’s proposed consumer heads up display system). 25   See Dave Banks, Toyota’s “Window to the World” Offers Backseat Passengers Augmented Reality, Wired (July 29, 2011), www.wired.com/geekdad/2011/07/toyotas-window-to-the-worldoffers​-back​seat-passengers-augmented-reality/. 26   See Paul Ridden, World First GPS Goggles with Head Mounted Display, Gizmag (Oct. 8, 2010), www.gizmag.com/zeal-recon-transcend-gps-head-mounted-display-goggles/16605/. 27   See Babak A. Parviz, For Your Eye Only: A New Generation of Contact Lenses Built with Very Small Circuits and LEDs Promises Bionic Eyesight, IEEE Spectrum, Sept. 2009, at 36 (­ describing contact lens research).

BARFIELD PRINT.indd 611

13/11/2018 11:53

612  Research handbook on the law of virtual and augmented reality 2.5.1 Places Smartphones and other GPS-enabled devices increasingly provide location-relevant information about the places we visit. Early and simple examples include museum and city walking tours, information about nearby restaurants or other retail establishments, and directions to movies or the nearest ATM machine. But location-sensitive information can also have more profound implications for contracting. The popular website Zillow, for example, aggregates publicly available information to display price information about almost every house in the United States; this information is now available on smartphones as well. You can point your iPhone down a street and see, on a map, prices overlaid on each house. Various companies are layering more and more information onto such maps. This includes displays of criminal activity in the area,28 or quality ratings of neighborhood schools.29 This sort of information could be of obvious import in real estate transactions. In addition, as discussed above in the context of Williams v. Walker-Thomas Furniture Co.,30 consumers can now augment physical locations with digital information. Yelp, for example, allows a user to review a retailer or restaurant and to leave that review for other users that enter that particular physical location. These technologies allow consumers to communicate with and learn from each other in entirely new ways. In the analog world, you might have learned about a good restaurant or a bad merchant if you happened to talk with a friend who had personal experience with that retailer. In an augmented reality, however, you can now learn from your social network or from the universe of strangers who have visited that retailer, just by visiting the retailer and seeing information tagged to that location. This makes a merchant’s entire reputation available to each consumer that enters the store, thereby dramatically lowering the cost for consumers to sort merchants and to determine with which merchants to transact. It is hard to overstate the radical nature of this shift to ubiquitous location-based information. In the traditional or analog economy, consumers had few ways to search for information about a given merchant. The owners of your favorite local breakfast spot, for example, have always known about the restaurant’s food quality and record of health code violations, but a potential customer standing on the sidewalk outside could not access this information. As with all sorting problems, consumers were thus forced to sort merchants by relying on imperfect proxies. Merchants would try to signal their quality to consumers by investing in the architecture or furnishings in their stores, through expensive advertising, or through branding. In an augmented reality, however, consumers can access consumer reviews indexed by location, giving them entirely new ways to sort their contracting partners.

28   See Crime Mapping, www.crimemapping.com (providing map overlays of police reports about criminal activity for a variety of U.S. cities). 29   See, e.g., Fixing D.C.’s Schools, Washington Post, www.washingtonpost.com/wp-srv/ metro/interactives/dcschools/scorecard.html (providing interactive map showing school ratings, percentage of classes taught by highly qualified teachers, crime incidents, health inspection violations, and delay in repairs for city schools). 30   350 F.2d 445 (D.C. Cir. 1965).

BARFIELD PRINT.indd 612

13/11/2018 11:53

Freedom of contract in augmented reality  613 2.5.2 Goods In addition to information about places, we can now instantly access a great deal of information about goods. Two points are relevant here. First, information about goods is increasingly available through mobile computing. Second, both the quantity and the quality of information about goods are exploding, as goods become connected to the internet and capable of recording and transmitting vast stores of information about themselves. First, accessibility. Goods possess what some in design call an “information shadow.”31 A good’s information shadow is the composite of all information available about it—its origins, makeup, and manufacturer; market information about its price; consumer and professional reviews of its quality and performance; government safety information; and any other relevant information. Goods have always had such a shadow, but in an analog world it was cost prohibitive to search for and to use such information. You might talk to a friend about which car to buy, but it was impossible to collect reviews from hundreds of acquaintances. As Mike Kuniavsky has put it, “[t]he complexity of finding, organizing, and accessing this information divided the world of objects and the world of information shadows . . . Even if accessible through a computer, information shadows were unavailable when they could provide the most value: in choosing between different products to buy, or in figuring out how to use a new tool.”32 Mobile computing increasingly makes a good’s information shadow instantly accessible anywhere. As mentioned, smartphones can now scan an object’s bar code to pull up reviews of that product. Almost half of Android and iPhone users have scanned a barcode with their phones; nearly a quarter have scanned an image of a product to search for information about that product.33 Moreover, preliminary research seems to suggest that consumers prefer—and may even be willing to pay for—the ability to access product information and reviews while shopping in physical space.34 As databases of information about goods become more organized, complete, and available, consumers will increasingly have access to the information shadows of the objects they interact with or wish to purchase in the physical world. Second, and just as important, in addition to consumers being able to sort high-quality from low-quality goods by accessing digital information, goods can now signal to consumers in entirely new ways. Shoes, refrigerators, wine bottles, car tires, and even cattle can now record relevant data about themselves and report such data electronically to 31   See Mike Kuniavsky, Smart Things: Ubiquitous Computing User Experience Design 72 (2010) (“The digitally accessible information about an object can be called its information shadow”). 32   Id. 33   See Dan Butcher, Mobile Bar Code Scanning Has Become Mainstream: Survey, Mobile Commerce Daily (Jan. 20, 2011), www.mobilecommercedaily.com/2011/01/20/mobile-bar-codescanning-has-become-mainstream-survey (providing statistics on scanning via various smartphone platforms). 34   See Tobias Kowatsch et al, The Role of Product Reviews on Mobile Devices for In-Store Purchases: Consumers’ Usage Intentions, Costs and Store Preferences, 6 Int’l J. Internet Marketing & Advertising 226, 235–6, 239–40 (2011) (finding that consumers may be willing to pay between 2 and 5 percent of a product’s cost to have access to reviews of that product).

BARFIELD PRINT.indd 613

13/11/2018 11:53

614  Research handbook on the law of virtual and augmented reality consumers. This is the “Internet of Things”—a connected web of “smart” objects capable of generating and transmitting data on themselves.35 The ability of goods to share information with consumers is, again, a radical change. As one recent ubiquitous computing study put it: Smart products—products that share information with consumers—are designed to combine the online and offline world . . . Whereas today’s in-store products provide only static information, future smart products may provide information about their journey to their current location (e.g. multiple countries of origin), information about their ingredients (e.g. news articles on problems with a particular supplier) and possibly some embedded intelligence . . . (e.g. the types of difficulties others have had [with] the product) . . . Benefits for consumers could be: First, in terms of search, the user should be able to quickly discover available online information, e.g., comparable products plus rich descriptions of its use and limits; [i]n terms of experience, the user should be able to learn from others how well-liked the product was, how long it lasted, etc.; and in terms of credence, the user should be able to find out how the company treated returns, failures, etc. These benefits could make purchase decisions easier.36

Essentially, these technologies allow goods to signal quality to consumers in new ways. Consider automobiles. Cars have long had vehicle identification numbers (VINs)—an identifier unique to each vehicle. Until recently, however, searching for information based on a VIN was relatively difficult. Today, Carfax.com allows you to search based on a VIN and learn the vehicle’s owner history, mileage, recall information, and warranty information.37 It will also report whether the vehicle has ever been in a flood or seriously damaged in an accident, whether its airbags have deployed, or whether it has ever been branded a lemon under state lemon laws. The same is true of many other goods for which such information might prove valuable. Among consumer goods, computers, appliances, cameras, and other high-value electronics come to mind. Beyond the consumer realm, all forms of commercial, agricultural, and industrial machinery are obvious candidates. Goods are increasingly smart, and information about smart goods is increasingly available to augment our interaction with the physical world. 2.5.3  People and firms People and firms also have information shadows—a digital self that augments the physical self. Consumers and firms can increasingly learn about each other prior to transacting by accessing these information shadows. Consider reputation. A person’s reputation is just a Google search away. Moreover, the internet has generated new user-generated reputation ranking systems. For example, eBay posts reputational measures of honesty for its buyers and sellers.38 Firms, too, have discovered that their virtual reputations are at least as important as traditional word of mouth. Public report cards for doctors and hospitals 35   See Scott R. Peppet, Regulating the Internet of Things: First Steps Towards Managing Discrimination, Privacy, Security and Consent, 93 Texas L. Rev. 85 (2014). 36   Florian Resatsch et al, Do Point of Sale RFID Based Information Services Make a Difference? Analyzing Consumer Perceptions for Designing Smart Product Information Services in Retail Business, 18 Electronic Markets 216, 218 (2008). 37   See CARFAX, www.carfax.com. 38   See eBay, www.ebay.com.

BARFIELD PRINT.indd 614

13/11/2018 11:53

Freedom of contract in augmented reality  615 are being touted as a means to correct an ailing healthcare system.39 And rating sites, in which home owners rate contractors, electricians, or plumbers; clients rate lawyers; students rate professors; tenants rate landlords; and consumers rate brands and manufacturers now provide the public with heretofore delayed, disaggregated, or nonexistent data about reputation.40 We have already seen examples of consumers learning about firms through online consumer and intermediary reviews of locations and goods, and subsection 2.5.4—on contract terms—provides more evidence of the ways in which such information changes the sorting process for consumers. Augmented reality changes sorting and signaling for both consumers and firms, however, not merely for consumers. Consumers can learn more about firms, but firms can also learn more about consumers. As already discussed, facial recognition technologies are now available, rather than just being the stuff of science fiction. We can also increasingly recognize others by the electronic devices they carry or by biometric markers such as iris scanning. As these technologies evolve and integrate into the economy, firms will increasingly be able to identify consumers individually. As Jerry Kang has put it: [E]nvision a future where we can “right click” on any object . . . and receive contextually relevant information. People already “Google” each other before going on dates or to interviews. Think about having the option of one-click “Googling” anyone you walk past, as you walk past . . . This process could become automated; no specific “request” to pull information will be required. Rather, software will manage our datasense and constantly seek out and filter information about nearby people.41

Beyond reputation, scholars have focused on the ways in which firms sort consumers by mining aggregated data such as credit histories. The privacy field is dominated by Daniel Solove’s concept of the “digital dossier,” which is a metaphor for the aggregate of information available in public and private databases about a given person.42 Data aggregators now possess huge amounts of information about each of us: Corporate data mining links at least 7,000 transactions to each individual in the United States per year—approximately half a million transactions over a lifetime.43 In a world in which firms can identify consumers—as Acquisti’s study shows they increasingly can—firms can link consumers to these digital dossiers and thereby acquire detailed profiles on the interests, habits, income levels, and shopping patterns of those that walk through their doors.

  See HealthGrades, www.healthgrades.com.   See Epinions, www.epinions.com. 41   See Jerry Kang & Dana Cuff, Pervasive Computing: Embedding the Public Sphere, 62 Wash. & Lee L. Rev. 93 (2005). 42   See Daniel J. Solove, The Digital Person: Technology & Privacy in the Information Age 1–2 (2004) (defining the digital dossier). 43   Jason Millar, Core Privacy: A Problem for Predictive Data Mining, in Lessons From the Identity Trail: Anonymity, Privacy and Identity in a Networked Society 103, 105 (Ian Kerr et al eds, 2009). 39 40

BARFIELD PRINT.indd 615

13/11/2018 11:53

616  Research handbook on the law of virtual and augmented reality 2.5.4 Contract terms Finally, in addition to information about places, goods, people, and firms, consumers increasingly have ubiquitous mobile access to information about contract terms. Various websites allow consumers to share information, including reviews and complaints related to contract terms. For example, pissedconsumer.com, complaintsboard. com, ripoffreport.com, complaints.com, consumeraffairs.com, scamfound.com, and others host thousands of consumer comments about contracting problems. Consider just a few examples. In the following, a reviewer posted under the subject heading “Never Never Sign a Contract with Dish Network”: According to Dish TV customer service . . . their contract allows them to raise their rate every year if the [sic] want to[ ]. WHAT????? I cannot find my original contract to dispute this but this is a shady underhanded way or [sic] strong arming extra money out of their customers who are stuck in a contract. . . . DO NOT EVER sign a contract with Dish TV.44

Or consider this review titled “Beware of Contract Fine Print”: MassageEnvy—a massage spa that provides a discount rate for monthly massages with a oneyear minimum membership. If you decide to cancel, please be aware that you must give 30 days notice in writing and your membership continues during that time. So basically you need to visit the MassageEnvy spa exactly 30 days before the end of your membership or otherwise you get charged another monthly credit.45

Thousands of similar reviews contain complaints about being locked into contracts for longer than expected, excessive cancellation fees, and contract terms that mislead consumers about the expected benefits or coverage of a contract. In many cases, multiple consumers have complained about the contract of a given merchant or product. These negative reviews are easy to find. For example, a Google search for “Anytime Fitness” brings up negative reviews in the top five search results. A search for “Anytime fitness contract complaints” brings up complaints posted on pissedconsumer. com, complaintsboard.com, ripoffreport.com, consumercomplaints.com, my3cents.com, scamfound.com, the Better Business Bureau, and a variety of geographically based consumer complaint sites. A search for the same terms just on pissedconsumer.com returns almost 200 negative reviews, which have been viewed thousands of times by searching consumers. Pissed Consumer indexes and sorts these reviews, as well as aggregating the ratings to give Anytime Fitness a “Five Red Star” score. In addition to such consumer reviews, intermediaries provide rankings of consumer goods and services that include scores related to contract terms. J.D. Power & Associates, for example, maintains a website that provides consumer information about a variety of products. On the site, J.D. Power ranks credit cards based on their contractual terms, allocating scores from “among the best” to “the rest.” Other sites provide similar rankings

44   Never Never Sign a Contract with Dish Network, Pissed Consumer (Mar. 20, 2011), http:// dish-network.pissedconsumer.com/never-never-sign-a-contract-with-dish-network-20110320227723. html. 45   Beware of Contract Fine Print, Pissed Consumer (May 28, 2011), http://massage-envy. pissedconsumer.com/beware-of-contract-fine-print-20110528240139.html.

BARFIELD PRINT.indd 616

13/11/2018 11:53

Freedom of contract in augmented reality  617 of credit card contract terms.46 Sites rate other types of contract terms as well, including health care plans, insurance contracts, airline frequent flyer clubs,47 automobile manufacturer warranties,48 and computer warranties.49 It is difficult to judge the quality of such consumer and intermediary reviews, but it is easy to assess the quantity: The past few years have witnessed an explosion of contractrelated information on such review sites and, at least for large firms and popular products, a consumer looking for information about a merchant’s contracts need only do a simple Google search. Such information about contract terms will likely increase over time in both quantity and quality. Augmented reality delivery methods make it more convenient and less costly for consumers to access such information, which will increase demand. Consumers can search for such information on their smartphones. In addition, however, the newest mobile applications use location information to push such information to users without requiring search.50 And some apps—such as SearchReview’s mobile platform—aggregate consumer reviews from thousands of websites, providing a much simplified consumer search experience.51 In addition, mobile computing technologies make it much easier for consumers to supply such information by creating reviews on their mobile devices. The popular website The Consumerist, for example, built a “Consumer Tipster” feature into its mobile application for the iPhone. According to The Consumerist, this is a “free, fun, and simple way to tell the editors of Consumerist.com about any scams, rip-offs [or] shoddy products . . . Use the Tipster app to snap a quick picture of the offending item, write up your comment, and submit to The Consumerist.”52 The Better Business Bureau’s mobile application has a similar feedback feature, as does Pissed Consumer’s iPhone platform.Thus, augmented reality technologies not only allow customers to access information about products and contract terms, but also encourage them to generate and disseminate this information. I now turn to an exploration of how this explosion of accessible information interacts with consumer contract doctrine.

3. THE EFFECTS ON ASYMMETRIC INFORMATION ARGUMENTS FOR DISTRUSTING CONSUMER CONTRACTS Section 2 explored augmented reality technologies, the types of information those technologies can now make available to contracting parties, and some of the ways in which

  See, e.g., CardRatings.com, www.cardratings.com.   See, e.g., WebFlyer, www.webflyer.com. 48   See Cars.com, www.cars.com. 49   See Clearance Club, www.clearanceclub.com. 50   See, e.g., Better Business Bureau, www.bbb.org (providing reviews of local businesses based on location information). 51   See SearchReviews, www.searchreviews.com (aggregating 40 million reviews from over a thousand websites). 52   See Consumer Tipster by Consumer Reports, iTunes App Store, http://itunes.apple.com/us/ app/consumerist-tipster/id443045014?mt=8. 46 47

BARFIELD PRINT.indd 617

13/11/2018 11:53

618  Research handbook on the law of virtual and augmented reality such information permits new forms of low-cost sorting and signaling. Each of these types of information—about places, goods, people, firms, and contract terms—has been available before, but often only at significant cost. Today, these data are more accessible, more often, to more people. The information infrastructure of the economy has shifted underneath the feet of contract doctrine. I now consider the implications of that shift. In the consumer context, we hesitate to enforce contracts as written either because of information asymmetries that create market failure or because of doubts about the parties’ capacity to contract.This section focuses on the first of these two justifications for limiting traditional freedom of contract: the class of cases in which information-related market failures make a transaction suspect by dampening competition in the market for the terms in question and thereby creating opportunity for exploitative terms. In these cases, we distrust that individual choice will maximize welfare and therefore intervene to police contracts. Standard form contracts between firms and consumers—like those at issue in Walker-Thomas—are the best example. After reviewing the arguments for and against judicial distrust of consumer contracts, this Part considers the primary effects that augmented reality will have on this debate. Ultimately, it concludes that an augmented reality will force both judges and scholars to take account of the ways in which digital information is now available to consumers in bricks-and-mortar transactions, and that this convergence of physical and digital space in an augmented reality will create doctrinal pressure to defer to contracts as written. Whether a contracting party has access to a smartphone (or other augmented reality device) may be—or may become—an important element in determining that party’s contractual rights and obligations. 3.1  Information Asymmetries and Freedom of Contract Classical Willistonian contract law assumed that contracting parties were best positioned to arrange their affairs and that judicial intervention into those affairs was largely unwarranted. This aligned with basic welfare economics, which posits that informed parties will reach efficient agreements under optimal conditions. Over the past century, however, the traditional tendency to enforce contracts as written has been under attack—particularly in the context of standard form consumer contracts.53 Courts and consumer advocates are suspicious of such “adhesion contracts,”54 assuming that they contain onesided, proseller terms harmful to consumers. As a result, modern courts are more likely than their classical predecessors to question, alter, or reject a contract’s written terms on grounds of unconscionability or unfair surprise.

53   See Avery Wiener Katz, The Economics of Form and Substance in Contract Interpretation, 104 Colum. L. Rev. 496, 498 (2004) (“As is well known to both students and scholars of contract law . . . for the past one hundred years . . . the historical trend . . . has been to water down . . . formal doctrines in favor of a more all-things-considered analysis of what the parties may have meant in the individual case”). 54   See Friedrich Kessler, Contracts of Adhesion: Some Thoughts about Freedom of Contract, 43 Colum. L. Rev. 629 (1943) (coining the term); Todd D. Rakoff, Contracts of Adhesion: An Essay in Reconstruction, 96 Harv. L. Rev. 1174 (1983).

BARFIELD PRINT.indd 618

13/11/2018 11:53

Freedom of contract in augmented reality  619 3.1.1  The law and economics of standard form contracts By the late 1970s, however, law and economics scholars began to argue that suspicion of standard form contracts might be misguided, even in consumer contracts. Standardized terms create efficiencies by reducing transaction costs—bargaining over every contract we enter daily would be overly cumbersome. Moreover, even if most consumers never read standard form contracts (thereby tempting sellers to include abusive terms), a relatively small cadre of sophisticated and vigilant consumers might suffice to force sellers to include only moderate, efficient terms in their contracts.55 So long as this informed minority creates competitive pressure, merchants will offer consumers efficient terms because failure to do so would result in consumers going elsewhere. As Michael Trebilcock has put it: To the extent that there is a margin of informed, sophisticated, and aggressive consumers in any given market, who understand the terms of the standard form contracts on offer and who either negotiate over those terms or switch their business readily to competing suppliers offering more favourable terms, they may in effect discipline the entire market, so that inframarginal (less well informed, sophisticated, or mobile) consumers can effectively free-ride on the discipline brought to the market by the marginal consumers, although there is the potential for a collective action problem if every consumer attempts to free-ride on the efforts of others in effective monitoring of contract terms.56

Put slightly differently, if consumers have sufficient information to compare terms across different suppliers’ contracts, suppliers will have reason to provide efficient contracts without unreasonable terms.57 “New formalism,”58 or neoformalism, thus uses law and economics to rejustify enforcing contracts as written. Some neoformalists have confined their arguments to commercial contracting contexts, but many of these scholars also doubt the utility or necessity of judicial policing of consumer contracts. Economically, such intervention is only justified when one of two conditions holds: Either the seller has monopoly power and thus the ability to impose burdensome and inefficient terms, or there is no likelihood that consumers can or will inform themselves about such terms in a standard form contract.59 Absent some such market failure, judicial intervention—for unconscionability and even for fraud—is unjustified. So long as markets and parties are reasonably informed, freedom of contract should prevail. 55   See Alan Schwartz & Louis L. Wilde, Imperfect Information in Markets for Contract Terms: The Examples of Warranties and Security Interests, 69 Va. L. Rev. 1387, 1450–52 (1983) (providing this argument about standard form contracts). 56   Michael J. Trebilcock, The Limits of Freedom of Contract 120 (1993) (citations omitted). 57   See Alan Schwartz, How Much Irrationality Does the Market Permit? 37 J. Legal Stud. 131, 136–7 (2008) (arguing that suppliers will provide efficient terms when consumers are informed). 58   See generally David Charny, The New Formalism in Contract, 66 U. Chi. L. Rev. 842 (1999). 59   See generally Alan Schwartz & Louis L. Wilde, Intervening in Markets on the Basis of Imperfect Information: A Legal and Economic Analysis, 127 U. Pa. L. Rev. 630, 666–82 (1979) (making this argument); see also Richard A. Epstein, The Neoclassical Economics of Consumer Contracts, 92 Minn. L. Rev. 803, 804–5 (2008) (“There are . . . two sets of . . . circumstances in which the neoclassical theory accepts that some government intervention may make sense: private monopoly and imperfect information”).

BARFIELD PRINT.indd 619

13/11/2018 11:53

620  Research handbook on the law of virtual and augmented reality This argument’s force, of course, turns on the empirical claim that a sufficient number of consumers can and will inform themselves enough to police standard form contracts and create market discipline. The debate turns on information. The question is whether consumers learn sufficiently to discipline sellers, either through their own direct experiences with a given seller or product, from friends or other consumers, or from third-party intermediary information providers. Over the past decades, many have doubted that consumers can or will inform themselves about standard form contract terms. In most transactions, a consumer has no incentive to read a standard form contract: The consumer most likely will not understand the legal language anyway, and knows that the risk of the contract terms being invoked is very low. Moreover, a small cadre of informed buyers may not appear, because the benefits of processing complex contract terms may not justify the costs.60 3.1.2  The e-commerce complication On the sidelines of this debate, a few contract scholars have asked whether e-commerce transactions conducted online might differ from traditional brick and mortar transactions in ways that change these asymmetric information arguments. Two lines of inquiry have evolved. First, do consumers read electronic standard form agreements more consistently than they read paper contracts? Second, even if they do not, do consumers have access to other sources of information—such as consumer reviews—that might discipline consumer contracts and thus make e-commerce contracts more efficient than their paper counterparts? We consider each argument in turn. In their early work on e-commerce contracting, Robert A. Hillman and Jeffrey J. Rachlinski saw that “[t]he ease with which consumers can compare business practices, including the content of standard forms, suggests that consumers do not need judicial intervention to protect themselves from business abuse.”61 Writing in 2002—early in the commercial internet’s evolution—they naturally focused on the solitary consumer sitting at home in front of a personal computer. E-commerce generally gives a consumer access to a standard form contract in advance of purchase, sufficient time to read the contract, and the privacy to do so away from the pressure of salespeople or other distractions. This increases the likelihood that a consumer will read the contract. Despite these advantages, Hillman and Rachlinski concluded that these new technologies did not radically change consumer contracting: “[C]ourts cannot simply assume that the new tools available to e-consumers will suffice to protect them against exploitation.”62 Others have since taken up the argument over whether consumers read electronic standard form contracts. Not surprisingly, the available empirical evidence suggests that they do not.63

60   See Shmuel I. Becher & Tal Z. Zarsky, E-Contract Doctrine 2.0: Standard Form Contracting in the Age of Online User Participation, 14 Mich. Telecomm. & Tech. L. Rev. 303, 311 (2008) (reviewing literature skeptical of the “informed minority” theory). 61   Robert A. Hillman & Jeffrey J. Rachlinski, Standard-Form Contracting in the Electronic Age, 77 N.Y.U. L. Rev. 429, 478 (2002). 62   Id at 492. 63   See Robert A. Hillman & Ibrahim Barakat, Warranties and Disclaimers in the Electronic Age, 11 Yale J.L. & Tech. 1, 2 (2009) (finding that consumers do not read online terms).

BARFIELD PRINT.indd 620

13/11/2018 11:53

Freedom of contract in augmented reality  621 Section 2 illustrated, however, that since Hillman and Rachlinski’s seminal essay we have witnessed the spread of technologies permitting mass consumer-to-consumer and intermediary reviewing of products, brands, and merchants. A second line of argument has thus recently developed: In e-commerce transactions, consumers have access to a great deal of information beyond just the contract itself. If that information creates competitive pressures on firms to moderate their contract terms, judicial interference in consumer contracts may be less justified in the e-commerce environment: The important question in the consumer transaction context asks whether contract law should hold consumers responsible—by denying them access to protective doctrines such as unconscionability . . . that are based upon bargaining power weaknesses—for their failure to take reasonable, low-cost steps to improve their bargaining power and protect their ability to withhold consent. . . . If a single Internet search would [show problems with a contract], did the consumer lack bargaining power, or just fail to access a cheap and readily available source?64

Put differently, the internet allows consumers to evaluate a broad range of goods prepurchase that previously could only be evaluated postpurchase. Both “experience goods”—which generally can only be evaluated after personal experience with the good—and “credence goods”—which are difficult to evaluate even after experience—may transform into “search goods” which are capable of evaluation prepurchase,65 if the internet aggregates sufficient information about those goods from multiple consumers or watchdog intermediaries. Shmuel I. Becher and Tal Z. Zarsky, for example, have argued that “the online realm creates market conditions and information flows that may allow market forces to assure balanced [standard form contracts].”66 They focus on consumer-to-consumer product reviews and claim that certain indicia of robust consumer-to-consumer information flow should temper courts’ willingness to interfere in consumer contracts. These indicia include that the firm promotes robust consumer conversation about the firm’s products on its website and that there is heavy consumer-to-consumer information flow. Conversely, if the firm frequently changes its standard form contracts, engages in discrimination between consumers by offering one set of terms to the informed minority but another set of more onerous terms to the uninformed majority, or tampers with information exchange between consumers, such actions might justify judicial intervention. This line of inquiry is new, and certainly not settled. There are various empirical questions to be answered, and some early work suggests that online product reviews are not always a good proxy for the quality or fairness of consumer contract terms. At the same time, however, online review sites are still in their infancy, as are mobile applications that both bring their content to brick and mortar transactions and allow consumers to easily augment physical space with review information from their mobile devices. These services are likely to continue to improve in both quality and quantity. In general, therefore, the 64   Daniel D. Barnhizer, Propertization Metaphors for Bargaining Power and Control of the Self in the Information Age, 54 Clev. St. L. Rev. 69, 108 (2006). 65   For discussion of these different types of goods, see generally Phillip Nelson, Information and Consumer Behavior, 78 J. Pol. Econ. 311 (1970) (drawing distinctions between search, experience, and credence goods). 66   Becher & Zarsky, supra note 60, at 345.

BARFIELD PRINT.indd 621

13/11/2018 11:53

622  Research handbook on the law of virtual and augmented reality implications seem clear: Such secondary sources of information have the potential to discipline consumer contracts, even if those contracts generally remain unread. 3.2  Two Effects on Asymmetric Information Arguments This background lays the foundation for considering the consequences of augmented reality for asymmetric information arguments in the context of standard form contracts. To date, contract scholarship has assumed there are two distinct transactional spheres: a set of contracting experiences that occur in the physical world and a separate set of online e-commerce experiences that occur in digital space. The offline consumer is generally assumed to be uninformed and incapable of creating market pressure to discipline standard form contracts; the online consumer may be better informed, and in some instances sufficiently informed to create such pressure, thereby lessening the need for courts to police e-commerce contracts. As we saw in section 2, however, augmented reality technologies collapse the distinction between physical and digital space. This collapse will change asymmetric information arguments for policing consumer standard form contracts. In an augmented reality, consumers in physical space have access to the full panoply of information available in digital space, which may allow them to effectively sort firms, products, and contract terms before entering a contract. This suggests that many more consumer contracts may become efficient, and that judicial resort to unconscionability and other doctrines to police such contracts may be less and less justified. At the same time, however, the cost to sellers of explaining contractual terms to consumers in brick and mortar transactions is also dropping. This may actually have somewhat perverse effects, if sellers use augmented reality explanations to insulate questionable contract terms from judicial scrutiny. Consider each argument in turn. 3.2.1 Consumers sorting sellers: increased consumer access to information in physical space In an augmented reality, a consumer purchasing a lawnmower in a hardware store—or a computer in an Apple store, or an automobile at a local car dealership—has access to the same information about the goods arrayed in the store as another consumer has online. This information-rich environment will likely create doctrinal pressure to enforce standard form contracts against consumers, because it becomes more difficult to argue that the consumer suffered from an information deficit during the transaction. The relevant question may not be whether a given consumer read a standard form contract. Instead, the question may be whether consumers had ubiquitous access to consumer and intermediary reviews that could have indicated problems with the product or contract. If so, there may be sufficient competitive pressure on the seller to discipline its contracts. The doctrinal question of import may thus be whether a given consumer had a smartphone (or other such device) at the time of contracting. Rather than focusing on educational background or even literacy, courts may have to begin to wrestle with the reality that technologically enabled consumers have an advantage over those without access to such information. Courts may begin to differentiate between connected and

BARFIELD PRINT.indd 622

13/11/2018 11:53

Freedom of contract in augmented reality  623 disconnected consumers—between those with access to digital information about their transactions and those without such access. Courts are also likely to look closely at the actual information available to consumers in a given context. As section 2 explored, we are at the beginning of an augmented reality and the collapse of digital and physical space. A huge amount of new information is now available to consumers, but it is not perfectly comprehensive. If information about a given product or contract is unavailable, consumers’ ability to sort decreases and firms’ temptation to include oppressive terms in their contracts increases. The point is that contract scholarship to date has distinguished between e-commerce or online transactions and brick and mortar, traditional consumer contracts. As augmented reality technology makes information available everywhere, that distinction is less and less relevant. The distinction that makes a difference is that between those with access to augmented reality technologies and those without. 3.3  Three Counterarguments and Their Implications To this point, I have considered the strongest arguments that information asymmetries will be dampened in an augmented reality. These arguments will likely raise several legitimate objections. This subsection briefly considers three counterarguments: (1) that consumer review and intermediary information is of insufficient quality to aid consumers in exerting market pressure on firms over harsh contract terms; (2) that consumers will never be sufficiently motivated to learn about consumer contracts, even in an augmented reality; and (3) that such information will cause information overload, thereby raising—rather than lowering—transaction costs for consumers. 3.3.1  Questionable information quality The first likely objection to these arguments is that both consumer-generated and intermediary-generated reviews may be insufficient to discipline firms with consumerunfriendly contract terms. In other words, such reviews may not be an accurate proxy of contract quality, and therefore consumer reliance on such reviews in an augmented reality may change certain aspects of product markets but not affect consumer standard form contracts. There is an emerging technical literature about online consumer review mechanisms and how to ensure or improve review accuracy and quality.67 Various general problems exist: inaccurately negative reviews submitted by competitors; inaccurately positive reviews planted by firms about their own products or services; the potential for reviews to skew towards the negative because only disgruntled consumers will be motivated to post reviews; the difficulty for a given user of determining which reviews to attend to, given that a particular reviewer may have very different tastes or quality standards than the user; and how to best aggregate and simplify large numbers of reviews to accurately inform consumers without requiring them to read hundreds or thousands of detailed comments.

67   See, e.g., Bing Liu, Sentiment Analysis and Subjectivity, in Handbook of Natural Language Processing (2d ed., Indurkhya & Damerau, eds.) (2010); B. Pang & L. Lee, Opinion Mining and Sentiment Analysis, 2 Found. & Trends in Info. Retrieval 1 (2008).

BARFIELD PRINT.indd 623

13/11/2018 11:53

624  Research handbook on the law of virtual and augmented reality Certain additional problems are relevant to the debate over consumer contract terms. In particular, some early empirical work suggests that product reviews—focused on whether consumers favored or disfavored a given product or service—may not necessarily be contract reviews. In other words, product reviews may not be accurate proxies of whether the contract contains proconsumer terms.68 The reason is simple: A consumer’s online review of a product is likely to be multidimensional and judge that product based on many variables, such as product aesthetics, durability, the shopping or repair experience, whether the product lived up to expectations, the usefulness of the product, and, perhaps, the consumer’s experience with the product’s contract terms. The mixed nature of such reviews makes it difficult for them to serve as pure proxies for contract quality. These are serious concerns, and I do not have a dispositive response to all of them. They raise empirical questions about the utility of online consumer and intermediary reviews, and about how sophisticated the market for such review information is likely to become. To this point I have largely presented the optimistic assumption that such information will continue to improve in quantity and quality as these young technologies mature. Consumer review services, for example, may begin to offer easy means for consumers to rate different aspects of a product separately—giving a score to each of the dimensions mentioned previously, for example, or at least indicating which dimension was most relevant to the consumer’s product experience. Such separation would allow a review site to display a scorecard about a given product or service that indicates consumers’ perceptions of each variable individually. Consumers’ experience with contract terms seems likely to be one unique aspect of such a separation strategy. In addition, as noted in section 2, many existing reviews already focus on contract terms. Review services could easily begin to search for and to categorize reviews based on keywords, such as “contract,” thereby dampening the problem of multidimensionality without requiring additional work by consumers. In addition, reviews are likely to become more sophisticated as online review services integrate with social networks. The travel-related review website TripAdvisor, for example, already integrates with the social network Facebook, allowing users on TripAdvisor to see whether their friends have posted reviews about a given hotel or location. This at least partially overcomes the objection that online reviews are unhelpful because individuals don’t know whether other consumers share their tastes and preferences. Further, quantity and scale can to some extent ameliorate concerns over the quality or accuracy of any given review. Obviously, 10,000 inaccurate reviews are no better than 10, but 10,000 reviews are unlikely to all be planted or skewed. Instead, the large volume can give a consumer a general sense of the problems with a given product or service, dampening the effects of any single review that is overly positive or negative. Ultimately, the empirical questions raised cannot be answered here. The solutions suggested above seem likely to develop as consumer review services evolve. We are only now beginning to experiment with the possibilities of consumer-to-consumer sharing of 68   See, e.g., Nishanth V. Chari, Disciplining Standard Form Contract Terms Through Online Information Flows: An Empirical Study, 85 N.Y.U. L. Rev. 1618, 1622 (2010) (suggesting that positive online consumer rankings of products often correlate with more pro-seller standard form contract terms and, therefore, that online rankings do not serve as useful consumer information as to the nature of such terms).

BARFIELD PRINT.indd 624

13/11/2018 11:53

Freedom of contract in augmented reality  625 product information; it is too early to judge these offerings as imperfect when many are only in their infancy. In addition, it is worth noting that my argument does not depend on the market for consumer standard form contracts becoming perfectly informed. I have simply made the related claims that this market is becoming better informed as the cost of signaling and sorting drop—moving towards a perfect market, even if not attaining perfection. Even at this very early stage, however, augmented reality technologies bring new information to consumers that has previously been unavailable, suggesting that our approach to standard form contracts entered into in brick and mortar transactions may change. 3.3.2 Consumer motivation Second, one could certainly object that consumers in an augmented reality are no more likely to be motivated to read, watch, or listen to contract-related information than they are when confronted with paper contracts. Put differently, no matter how augmented such information becomes, consumers will be indifferent to it. This may have a somewhat perverse effect: If consumers remain uninformed despite these additional sources of information, a formalist turn by the judiciary in contract cases will be unwarranted because firms will be able to insert exploitative terms into their contracts and then to insulate those terms via augmented reality technologies that go unused by the consumer. Robert Hillman has made a similar argument in the context of mandatory website disclosure of e-commerce contract terms.69 His concern is that requiring disclosure of contractual terms on a merchant’s website—so that consumers could read such terms prior to purchase—might not increase consumer information (because consumers would still not read them) and might instead merely make those terms, however unconscionable, more enforceable under the theory that consumers had an opportunity to become informed and chose to ignore that opportunity. Procedural unconscionability may be difficult to establish if consumers had an opportunity to read the terms prior to contracting. He calls this a “legal backfire.” There are at least two possible responses in this context, one descriptive and one somewhat normative. The descriptive argument is that augmented reality technologies are likely to make it easier and easier for consumers both to access and to generate such information, as these information channels become embedded into our surroundings and available through multiple means with almost no effort. Imagine, for example, that consumers could simply set their mobile devices to inform them of negative consumer review information—including, perhaps, setting a preference to be informed about negative reviews of contractual terms. The consumer’s mobile device could then interact with physical products as the consumer shopped, retrieving and analyzing information about those products on the consumer’s behalf. If a given product triggered the consumer’s preset preferences, the device could shake, change color, or project a warning label onto the physical good to warn the consumer that the product raised concerns. The consumer would not have to do much of anything; the technology would make such information part of the consumer’s shopping experience almost automatically.

69   See Robert A. Hillman, Online Boilerplate: Would Mandatory Website Disclosure of E-Standard Terms Backfire?, 104 Mich. L. Rev. 837, 840 (2006).

BARFIELD PRINT.indd 625

13/11/2018 11:53

626  Research handbook on the law of virtual and augmented reality In such a world, the argument that consumers will be insufficiently motivated to learn of contract terms seems mostly a remnant of our realization that consumers will not read paper contracts. That motivational hurdle, however, is outsized compared to the minimal motivation that may ultimately be required to become informed in an augmented reality. If consumers do not avail themselves even of these almost costless means to become informed, the normative response may be to ask: At what point do we hold consumers responsible for understanding contracts when we live in an augmented reality in which digital information about products and terms is available prior to purchase and new sources of information (such as video explanations) can be ubiquitously available to consumers? I do not tackle this question here—it can wait until these evolutions occur. My hunch, however, is that courts and commentators are likely to find that these new sources of information justify holding consumers to their contracts. There is a limit to the realist or contextualist tendency that has led contract law away from formalism in the last century, and augmented reality is likely to reveal that limit. 3.3.3 Information overload A third possible objection runs in the opposite direction from the consumer motivation concern: Perhaps in an augmented reality, consumers will become so saturated with information that it will become unhelpful. In other words, although augmented reality might lower certain transaction costs—related to search and data aggregation, for example—it may raise others. The effect may be that consumers become no better informed than at present about consumer contracts. Again, I cannot dispose of this concern with a simple argument. It is far too early in the evolution of augmented reality technologies to know whether they will lead to an environment in which consumers are so overwhelmed with information that all information becomes useless. It seems likely to me that instead, products and services will evolve to simplify and organize information for consumers so that it remains useful, and will also give consumers the option to choose when to be pushed information during a brick and mortar transaction and when to keep that information from view. I am particularly interested in the ways in which scoring mechanisms—such as the five stars commonly used in consumer reviews—are already reducing massive amounts of data to userfriendly formats. A consumer standing in front of five lawnmowers in a hardware store will want distilled information about those five products or brands, not unrelated information about the power drill on a nearby shelf or overly complex information about each lawnmower’s engineering specifications. But it is relatively easy to imagine efficient ways for augmented reality technologies to communicate condensed, useful information to that consumer— perhaps the consumer might point a mobile device at each lawnmower, and the device’s screen will flash red if one of the products has serious product defects or contractual problems. What information the device displays might depend on preferences the customer input into the device. Perhaps the device is set to warn only of serious problems with a given product, and the consumer can therefore shop safe in the assumption that there are no issues to be concerned about unless the device notifies the consumer otherwise. Whatever the solution, it seems likely that technologies will evolve to help consumers better assimilate the increased amounts of information available in an augmented reality. This may be an unsatisfying answer for some, or may seem overly optimistic about technological solutions to social problems. Those are fair criticisms. At the same time,

BARFIELD PRINT.indd 626

13/11/2018 11:53

Freedom of contract in augmented reality  627 it is, again, early in the evolution of these technologies. For now, suffice to say that the information overload objection is at this point theoretical and that there are plausible, practical ways to solve it in the future if need be.

4. THE EFFECTS ON BOUNDED RATIONALITY ARGUMENTS FOR DISTRUSTING CONSUMER CONTRACTS Let me summarize the argument to this point. Mobile computing augments physical space with digital information about places, goods, people, firms, and contract terms. This information has become rich and ubiquitously available to consumers during brick and mortar transactions. Although distrust of standard form consumer contracts is strong, such distrust rests, in part, on the argument that consumers do not read standard form contracts and have insufficient information to evaluate those contracts prior to purchase. In an augmented reality, however, consumers have access to the internet everywhere, and can easily search for relevant information to make purchasing decisions. In addition, increasingly sophisticated mobile applications can now push such information to consumers based on location or other variables, making it even easier for consumers to evaluate sellers and their contracts. This should discipline firms to reform onesided contracts. All of this suggests that in an augmented reality, courts may again begin to enforce consumer contracts as written. In addition to asymmetric information arguments, however, contract scholars have pursued a second line of argument in questioning standard form consumer contracts. The economic theory of contract assumes that parties have the capacity to make rational, informed decisions. However, if this assumption fails, there is no reason to protect the outcome of the parties’ choices: We can only assume that the outcome will be welfare maximizing if the parties have contractual capacity. Problems with capacity are thus the second traditional justification for limiting freedom of contract. Contract doctrine has long protected certain classes of parties due to concerns about capacity. Infirmity, minority, and intoxication are the traditional examples; lack of sophistication, education, or knowledge are more modern ones. More recently, behavioral law and economics scholars have begun to argue that cognitive heuristics, biases, and errors sometimes lead contracting parties to make faulty choices.70 Although not fully incapacitating, these justifications for intervening in contractual arrangements rest on a similar argument—that consumers are incapable of fully rational choice and thus courts should police consumer contracts. This section argues that the convergence of physical and digital space in an augmented reality may limit these bounded rationality justifications for intervening in consumer contracts. This will occur for two primary reasons: (1) because increased information may inherently mitigate bounded rationality; and (2) because augmented reality technologies may permit low-cost debiasing interventions to directly dampen the effects of bounded rationality in consumer contracting.

70   See, e.g., Shmuel I. Becher, Behavioral Science and Consumer Standard Form Contracts, 68 La. L. Rev. 117, 178 (2007) (“[B]ehavioral insights demonstrate that people sometimes make bad choices even where good information is available”).

BARFIELD PRINT.indd 627

13/11/2018 11:53

628  Research handbook on the law of virtual and augmented reality 4.1  Bounded Rationality and Freedom of Contract In the past decade, contracts scholars have turned to behavioral law and economics findings to question the efficiency of standard form contract terms. These arguments have been described as part of a “new realism” in legal theory,71 because at a general level they are arguments for paternalistic intervention into parties’ contracts. Essentially, this literature justifies judicial intervention on the grounds that errors in consumers’ decision making make it possible for sellers to include exploitative contract terms in their standard form contracts. This is the bounded rationality rationale for limiting freedom of contract. Melvin Eisenberg launched this approach to standard form contracts in his seminal 1995 article “The Limits of Cognition and the Limits of Contract.”72 Eisenberg argues that the standard economic account of contracting assumes that consumers have the ability to weigh the costs and benefits of contract terms and to make a rational choice that will maximize their utility. He then reviews the literature on three major cognitive biases or heuristics that may limit the ability of consumers to act in this rational manner. First, consumers are subject to bounded rationality: They cannot process all available information as a computer would, but must instead selectively attend to and incorporate information into their decisions. Second, consumers are subject to optimism biases. In general, psychological studies have shown that adults are systematically overoptimistic and tend to downplay (or mispredict) the likelihood of negative future events. Third, consumers are subject to what is known as the availability heuristic.73 This leads them to place disproportionate weight on easily available information when making decisions and to underemphasize less available—but perhaps more relevant—information. Thus, consumers may ignore low-probability risks and give too little weight to future costs and benefits as compared to present costs and benefits. Using this bounded rationality research as a foundation, Eisenberg argued that judicial intervention may be justified in cases such as Williams v. Walker-Thomas Furniture Co.74 He recognized that it might be rational for a consumer like Ms Williams to ignore a standard form contract. The costs of reading the contract are high, as are the costs of trying to understand its terms; the potential benefits are low; the risks of actually needing to fall back on the contract are—or seem—very low. “Rational ignorance” may therefore prevail, because consumers have limited information processing ability and make choices about where to invest their attention and cognition. Other contract scholars have followed Eisenberg’s line of inquiry. Russell Korobkin, for example, similarly argues that “the reason form terms deserve scrutiny is that buyers

71   See Victoria Nourse & Gregory Shaffer, Varieties of New Legal Realism: Can a New World Order Prompt a New Legal Theory? 95 Cornell L. Rev. 61, 77 (2009) (“Behavioral economics represents a frontal assault within economics itself on the simplifying assumptions of neoclassical law and economics”). 72   Melvin Aron Eisenberg, The Limits of Cognition and the Limits of Contract, 47 Stan. L. Rev. 211 (1995). 73   See id at 220 (“When an actor must make a decision that requires a judgment about the probability of an event, he commonly judges that probability on the basis of comparable data and scenarios that are readily available to his memory or imagination”). 74   350 F.2d 445 (D.C. Cir. 1965).

BARFIELD PRINT.indd 628

13/11/2018 11:53

Freedom of contract in augmented reality  629 are not fully rational, but rather make decisions in a boundedly rational manner.”75 Korobkin focuses on the psychological finding that people often focus on salient information at the expense of less salient, but relevant, information. In the standard form contract context, he argues that consumers may focus heavily on more salient terms, such as price, while ignoring other less salient terms, such as boilerplate arbitration clauses or financing terms. Sellers can therefore include inefficient and self-serving boilerplate terms, because consumers will ignore them. Korobkin argues that procedural unconscionability analysis should analyze a term’s salience, thereby seeking to counteract this effect. Like Eisenberg, Korobkin also applies his theory to Williams v. Walker-Thomas. He argues that buyers like Ms Williams are unable to factor all of the various terms in a standard form contract into their decisionmaking, or to compare all of those variables against competing terms from other sellers. Thus, there is little competitive pressure on sellers to include only efficient and reasonable terms, and there is no reason to assume that such consumer contracts are welfare maximizing. In addition, like Eisenberg, Korobkin considers the effect of optimism biases and overconfidence. He argues that Ms Williams likely underestimated the risk of future default and overestimated her ability to control that risk going forward. Again, such biased reasoning would lead Ms. Williams to discount the importance of the crosscollateralization clause, thus opening the door for the seller to include the seemingly oppressive term. Oren Bar-Gill has likewise argued that optimism biases may undermine contractual capacity: The force of the freedom of contract argument . . . is significantly reduced when one (or both) of the parties to the contract holds inaccurate perceptions of the future. The freedom of contract paradigm is based on the presumption that contracting parties correctly anticipate their future actions and thus the future consequences of the contract they have signed. Without an accurate perception of the future, freedom of contract cannot defend future-oriented contracts.76

He focuses on consumer credit card transactions and argues that consumers’ overoptimism leads them to underestimate the likelihood of negative future events related to their credit. As a result, consumers ignore oppressive boilerplate terms in credit card contracts that address adverse future contingencies in the event that a consumer fails to pay. Credit card companies take advantage of this imperfect decisionmaking by including oppressive long-term rates and fees—which consumers ignore—but attracting consumers with consumer-friendly short-term pricing such as teaser rates. Like Eisenberg and Korobkin, Bar-Gill considers the implications of his theory for Williams v. Walker-Thomas. His account differs somewhat: Even if Williams read the relevant term and understood its formal implications, she might still have underestimated the practical importance of this clause. Williams, when making the

75   See Russell Korobkin, Bounded Rationality, Standard Form Contracts, and Unconscionability, 70 U. Chi. L. Rev. 1203, 1207 (2003) (“[T]he reason form terms deserve scrutiny is that buyers are not fully rational, but rather make decisions in a boundedly rational manner, and . . . this provides sellers with an incentive to draft non-salient contract terms to their own advantage, whether or not such terms are efficient”). 76   Oren Bar-Gill, Seduction by Plastic, 98 NW. U. L. Rev. 1373, 1415 (2004).

BARFIELD PRINT.indd 629

13/11/2018 11:53

630  Research handbook on the law of virtual and augmented reality early purchases, may have underestimated the likelihood of purchasing additional items from the same seller, or she may have naively believed that she would never miss a payment. Due to the underestimation bias, Williams may have been insufficiently sensitive to the inclusion of the repossession clause.77

Thus, the underestimation bias allowed the Walker-Thomas Furniture Company to include an oppressive term related to future repossession, because Williams—like most consumers—was likely to underestimate the importance of that term. 4.2  Two Effects of Augmented Reality on Bounded Rationality Arguments These arguments based in behavioral law and economics are powerful, and they have captured the imagination of contract scholars over the past decade. They seem to justify intervention in consumer contracts from within economics by calling into question the basic economic premise that contracting parties can rationally assess their contractual opportunities and make welfare maximizing decisions. How do these bounded rationality arguments fare in an augmented reality? There are two primary consequences of augmented reality to consider here: the possibility that the ubiquitous availability of information will inherently or indirectly debias consumers, and the possibility of using augmented reality technologies to purposefully and directly debias contracting consumers. I consider each in turn below. 4.2.1  Sorting and debiasing: general debiasing through increased information On the one hand, it is doubtful that augmented reality technologies will contribute to a general debiasing of consumers. Many of these biases are difficult to correct.78 For example, simply explaining the overconfidence bias is rarely sufficient to debias consumers.79 Moreover, debiasing research has generally found that merely providing subjects with general information about risks and probabilities does not dampen these behavioral effects. Therefore, a diffused increase in information about consumer transactions seems unlikely to counteract the overoptimism, availability, and salience biases previously discussed. On the other hand, an augmented reality may help to debias consumers in a somewhat indirect way. As discussed in sections 2 and 3, in an augmented reality, consumers are likely to have access to increased information about products and sellers at the point of contracting because they can easily access product reviews from other consumers and from intermediaries. The availability of such product and merchant reviews has debiasing implications. Consider the salience problem. Korobkin argues that consumers generally focus on only a few salient characteristics of a product, such as price.They are thus unlikely to attend to the detailed contract terms in a standard form contract. Augmented reality, however,

  Id at 1432.   See, e.g., Jeffrey J. Rachlinski, A Positive Psychological Theory of Judging in Hindsight, 65 U. Chi. L. Rev. 571, 586 (1998) (discussing the difficulty of correcting certain biases). 79   See, e.g., Baruch Fischhoff, Debiasing, in Judgment under Uncertainty: Heuristics and Biases (A. Tversky & D. Kahneman eds, 1982) at 422 (discussing debiasing techniques). 77 78

BARFIELD PRINT.indd 630

13/11/2018 11:53

Freedom of contract in augmented reality  631 may provide consumers with salient information that simplifies and counteracts this tendency. For example, although consumers may not read a standard form contract and understand all of its terms, they might focus on the fact that the average consumer gave that product only one out of five stars on Amazon.Or a consumer might pull up mostly negative consumer reviews that highlight problems experienced with the product, thereby making salient the likelihood of needing to depend on the warranty. Finally, having access to consumer review information at the point of sale may bring to the consumer’s attention examples of other consumers who had to resort to the contract and discovered its oppressive or onesided terms. Section 2 explored examples of online reviews.Consider just one more example in the context of this discussion of salience. A consumer posted a review about Sirius satellite radio on pissedconsumer.com. The review highlights problems with Sirius’s contracts: I signed up for 1 year of sirius radio in April of 2010. I did not want it to continue, so I let it expire without renewing. The radio stops working (makes sense). Then I start getting statements saying I’m past due . . . Turns out when you sign up for one year of service it’s actually open ended! They keep charging, you have to spend a few hours calling . . . to cancel your 1 year contract. Oh, and if you try to cancel early there’s an early termination fee . . . I don’t know if a 1 year contract that they decide is open ended is legal, but it’s defiantely [sic] unethical and sleazy . . . Stay away!80

This sort of virulent consumer review is highly salient—certainly more salient than the dry legal language of standard form contracts. If a given product or merchant received a ­sufficient number of such reviews, an informed consumer would almost certainly take notice. In addition, mobile applications often aggregate and simplify reviews of consumer products using rating systems, which also counters the availability and salience problems noted by Korobkin and others. Pissed Consumer’s mobile application, for example, uses a “Five Red Star” system. The more red stars a product receives from other consumers, the higher the volume of consumer complaints. Likewise, the Better Business Bureau’s mobile application grades merchants on the standard academic A+ to F scale. These grades are based on the Bureau’s complaint history with the business, the type of business, time in business, availability of background information about the business, licensing or other governmental actions against the business, advertising issues, or failure to honor commitments to the Bureau. The Better Business Bureau also accredits merchants, and the mobile application indicates in a very simple—and salient—graphical way which local businesses have received their accreditation. As these services mature and expand, such ratings systems are likely to become more sophisticated. In particular, they may begin to differentiate between types of complaint and make it even easier for a consumer to search for complaints about a seller’s contract terms as opposed to other aspects of the seller’s goods or services. Even in their current form, however, these simplified ratings services may serve to counteract bounded rationality in consumer purchasing decisions by producing highly salient information that consumers can easily digest.

80   Sirius Keeps Billing after Contract Expires! Pissed Consumer (May 31, 2001), http://sirius-­sat​ ellite-radio.pissedconsumer.com/sirius-keeps-billing-after-contract-expires-20110531240638.html.

BARFIELD PRINT.indd 631

13/11/2018 11:53

632  Research handbook on the law of virtual and augmented reality In addition to ameliorating salience effects, augmented reality may also counteract consumers’ overconfidence about the future and apparent tendency to underestimate the probability of negative future events. Reading online consumer reviews highlights the obvious but important fact that sheer volume of experience uncovers even very low probability contingencies. When thousands of consumers have experience with a product or merchant, tens or hundreds will stumble into the unlikely, but inevitable, contingencies. If the seller’s standard form contract took advantage of the fact that consumers were unlikely to think about such low probability contingencies ex ante—by including a onesided term about such contingencies—ex post that onesidedness will be discovered as it is triggered by a given consumer’s experience. Later consumers reading an online review will therefore be on notice of that possible contingency and will likely factor it into their calculus about whether to purchase the product or service in question. Even if a given consumer chooses not to read every online review—and thus fails to learn about all of the discovered contingencies—feedback about such low probability events gets factored into the aggregated ratings on these sites. This creates a sorting proxy that may counter the availability heuristic. Consumers may not need to factor future contingencies into their ex ante evaluation of a product or contract if the electronic market of reviews does this for them. At one level, this suggests that newer products and consumer contracts deserve greater judicial scrutiny than their more established counterparts. Early adopters who purchase without the benefit of this electronic hindsight may stumble into one-sided terms. Later consumers should not—or, at least, should be on notice and able to factor the costs and benefits of such terms into their decisions. At another level, in an augmented reality, sellers should begin to anticipate these effects and realize that online information exchange will warn consumers away from onesided contracts. As the standard law and economics model predicts, they should therefore avoid including one-sided terms to begin with. 4.2.2  Signaling and debiasing: targeted debiasing by firms or regulators In addition to bringing contract-relevant information to the point of sale, augmented reality technologies allow firms and regulators to purposefully try to debias consumers. In high-value transactions, sellers may want to address bounded rationality problems directly to insulate their standard form contracts from later judicial attack. Alternatively, regulators may want to use augmented reality technologies to reach consumers at the point of sale and to nudge them to make better decisions.81 Consider this second mitigating effect on the bounded rationality justification. Targeted information can in some circumstances help to debias consumers. As Linda Babcock has noted, “In the literature on debiasing, one type of intervention stands out as effective against a wide range of biases. This involves having subjects question their own judgment by explicitly considering counterarguments to their own thinking.” 82 Warning 81   See generally Richard H. Thaler & Cass R. Sunstein, Nudge: Improving Decisions About Health, Wealth, and Happiness (2008). 82   See, e.g., Linda Babcock et al, Creating Convergence: Debiasing Biased Litigants, 22 Law & Soc. Inquiry 913, 916 (1999) (finding that explaining the overconfidence bias to mock litigants and urging them to consider the weaknesses of their case helped to mitigate self-serving biases).

BARFIELD PRINT.indd 632

13/11/2018 11:53

Freedom of contract in augmented reality  633 labels on cigarettes and alcohol are an example, as are risk disclosures to consumers prior to entering into credit or other sensitive transactions. In general, requiring subjects to consider contrary evidence has had greater success in countering overoptimism biases than merely providing generalized information about the biases themselves. Christine Jolls and Cass R. Sunstein argue that debiasing efforts can use one aspect of bounded rationality—the availability heuristic—to counteract another—overoptimism. “[B]ecause making an occurrence available to individuals will increase their estimates of the likelihood of the occurrence, availability is a promising strategy for debiasing those who suffer from excessive optimism.”83 The key is to expose subjects to a “concrete instance of the occurrence.”84 This tends to “render the incident in question available in a way that can successfully counteract optimism bias.”85 Jolls and Sunstein argue that one could use the law to debias consumers by requiring manufacturers to provide examples of the consequences of harm-producing uses of their products, rather than merely providing statistics or a generic warning label: “[I]n the consumer safety context the law might require that the real-life story of an accident or injury be printed in large type and displayed prominently so that consumers would be reasonably likely to see and read it before using the product.”86 This is targeted debiasing: trying to influence a consumer’s decision by exposing the consumer to particularized information intended to change the consumer’s probability assessments. Augmented reality offers novel methods for such debiasing. As discussed in section 3, augmented reality is rapidly lowering the cost of distributing video, for example, as well as images and audio. This makes it possible to create very vivid messages and deliver them to consumers prior to contracting. Computer scientists are already exploring various methods of nudging consumers using mobile technology. Researchers have, for example, used ambient display installations to influence workers’ decisions about whether to use stairs or elevators, leading to healthier choices often without workers’ conscious awareness that their behaviors are changing.87 Similarly, the “Augmented Shopping Cart” includes an ambient handlebar display connected to a bar code scanner: When a shopper scans an item before putting it in the cart, the cart’s handlebar changes colors to communicate nonobvious environmental, nutritional, or other information to the shopper.88 This can alter the consumer’s decisionmaking.89 Mobile phone-based applications have been used to promote smoking

83   Christine Jolls & Cass R. Sunstein, Debiasing Through Law, 35 J. Legal Stud. 199, 210 (2006). 84   Id. 85   Id. 86   Id. at 213. 87   See Yvonne Rogers et al, Ambient Influence: Can Twinkly Lights Lure and Abstract Representations Trigger Behavioral Change? UbiComp 2010, at 261 (Sept. 2010). 88   See Jon Bird et al, The Augmented Shopping Trolley: An Ambient Display to Provide Shoppers with Non-Obvious Product Information, PINC 2011 (May 8, 2011) (describing project); Vaiva Kalnikaite et al, How to Nudge In Situ: Designing Lambent Devices to Deliver Salient Information in Supermarkets, UbiComp 2011 (Sept. 2011) (finding that the augmented “shopping handle generated a significant nudge effect, influencing what products people chose and placed in their shopping trolleys”). 89   See, e.g., Jasminko Novak, Ubiquitous Computing and the Socially-Aware Consumer—

BARFIELD PRINT.indd 633

13/11/2018 11:53

634  Research handbook on the law of virtual and augmented reality cessation,90 dieting, exercise, medication management, environmentally friendly transportation choices, and other desired social practices.91 One can imagine a variety of creative mechanisms to debias consumers in the contracting context. Finally, one last comment on debiasing in an augmented reality. There is the risk of information overload if all consumer goods become saturated with digital information. As Colin Camerer puts it, even in the analog world, “[w]hen hammers start to sprout warnings of the danger they pose to thumbs, and ladders of the risk of falling, additional information confers ever smaller benefits and can actually backfire if it distracts consumers from more worthy warning messages.”92 In augmented reality, this could become comical: If every consumer good has augmented reality video presentations associated with it, consumers will ignore all such information—there is no time to watch an explanatory video about a box of TicTacs. Computer scientists and ambient informatics researchers are already confronting this problem when considering the effects of augmented reality technologies on information flow to consumers. One study, for example, examined several modalities—star rating systems, text blocks, and videos—for communicating product review information to consumers via mobile phones.93 It found that consumers prefer simplified, quick information modalities, such as star ratings, rather than more complex, time-consuming text or video presentations. These are very new problems, and this is preliminary research. Designing appropriate augmented reality technologies that achieve this balance will present both technological and regulatory challenges. Regulatory intervention might become necessary at some point to control the use of augmented reality technologies for debiasing or “warning label” purposes. Consumer protection law should take account of these technologies and begin to consider how best to employ them. This is not a contract law problem per se, but rather a consumer protection problem of how best to use these new methods of information dissemination.

5. CONCLUSION Doctrinally, we have come full circle. Traditionally, courts deferred to contracts as written, believing that consumers’ consent to those contracts justified enforcement. As standard form contracts proliferated, courts realized that consumers had insufficient information Support Systems in the Augmented Supermarket, Proc. of the First Int’l Workshop on Soc. Implementation of Ubiquitous Computing (2005) (examining use of technology to nudge consumers toward socially desired decisions). 90   See Robyn Whittaker et al, Mobile Phone-Based Interventions for Smoking Cessation, 4 Cochrane Library (2009) (reviewing literature). 91   See, e.g., Parisa Eslambolchilar et al, Nudge & Influence through Mobile Devices, MobileHCI 2010, at 527 (Sept. 2010) (reviewing literature and noting that “[t]he use of digital technology is central to make behavioral changes”). 92   Colin Camerer et al, Regulation for Conservatives: Behavioral Economics and the Case for “Asymmetric Paternalism,” 15 U. Pa. L. Rev. 1211, 1235 (2003). 93   See Felix von Reischach et al, An Evaluation of Product Review Modalities for Mobile Phones, MobileHCI 2010 (Sept. 2010).

BARFIELD PRINT.indd 634

13/11/2018 11:53

Freedom of contract in augmented reality  635 to evaluate those forms, and that firms could therefore include oppressive or onesided terms. Courts began to distrust standard form consumer contracts and to police them using doctrines such as unconscionability. Today, however, we live in a connected digital world. The doctrines born out of judicial distrust for standard form contracts were created before computers, the internet, or wireless technologies—before consumers had constant realtime access to information about the places, goods, people, firms, and contracts around them. In short, critical aspects of contract law were designed to resolve the problems of an economy in which contracting parties knew relatively little about each other, the goods they were trading, or the markets in which they operated. By contrast, today’s information technologies increasingly create an augmented reality in which digital and physical spaces are profoundly interwoven, and in which such types of information are ubiquitously available. Consumers can now sort firms and their contracts at lower cost and more efficiently, and firms can now signal their consumer-friendly contract terms to consumers. As our augmented reality develops, courts will—or should—begin to take notice of these changes to the economy that suggest changes to contract doctrine. Contract theory must also take account of these changes to the economy’s basic information infrastructure. Technological development can fundamentally change transaction costs. This may not occur often—indeed, changes profound enough to alter transaction costs in a significant way may occur only once in a generation. But we are experiencing such a shift, and this illustrates the importance to contract theory of recognizing how contract doctrine is contingent on underlying technological and economic conditions.

SOURCE Williams v. Walker-Thomas Furniture Co., 350 F.2d 445 (D.C. Cir. 1965).

BARFIELD PRINT.indd 635

13/11/2018 11:53

19.  Law and property in virtual worlds Wian Erlank

1. INTRODUCTION There is something intrinsically alluring about the thought of being able to own something that is not there.1 The purpose of this chapter is not to focus on or deal with the underlying maze of theoretical and practical complexities associated with virtual worlds or virtual property,2 but readers are encouraged to consult the references mentioned in the footnotes if they wish to immerse themselves in the theoretical rather than virtual world. In essence, the purpose of this chapter is to serve as an overview of virtual property and how the law deals with, specifically, property law issues arising from this. What then is virtual property, one may ask? As with most areas of the law, a universally accepted definition is hard to come by, but this chapter will aim to provide some clarity on the issue. Because the term “virtual property” has a very subjective element,3 the concept needs to be narrowed down for the purpose of this chapter. When the term is mentioned, some people immediately think that virtual property is property which does not exist.4 However, this is incorrect: Virtual property can be equated to immaterial

  See below for discussion of the fact that virtual property is not nonexistent.   For an indepth look into these complexities, see in general Wian Erlank, Property in Virtual Worlds (LLD dissertation, University of Stellenbosch 2012). For a broad background about virtual property see in general Joshua A.T. Fairfield, Virtual Property, 85 Boston University Law Review 1047–1102 (2005); Charles Blazer, The Five Indicia of Virtual Property, 5 Pierce Law Review 137–61 (2006); Kevin E. Deenihan, Leave Those Orcs Alone: Property Rights in Virtual Worlds, SSRN (http://ssrn.com/abstract=1113402) 1–51 (2008); F. Gregory Lastowka and Dan Hunter, The Laws of the Virtual Worlds, 92 California Law Review 1–74 (2004). 3   The same goes for one’s understanding of the concept of “property.” For a discussion on the different meanings of “property”, see Erlank, Property in Virtual Worlds 212; Wian Erlank, Determining the Property of Virtual Property, in Property Law Perspectives III 87–98 (Ann Apers et al eds, Intersentia 2014); Erlank, Determining the Property of Virtual Property 88; Cribbet et al, Property: Cases and Materials, 8th ed. (Foundation Press 2002) 2; Nelson et al, Contemporary Property (West Publishing Co 1996) 2–5; Kevin Gray, Property in Thin Air, 50 Cambridge Law Journal 252–307 (1991); Hanri Mostert and Anne Pope (eds) The Principles of the Law of Property in South Africa (Oxford University Press 2010) 4–5; P.J. Badenhorst, J.M. Pienaar, and H. Mostert, Silberberg and Schoeman’s The Law of Property, 5th ed. (Butterworths 2006) 1; Kevin Gray and Susan Francis Gray, Elements of Land Law, 5th ed. (Oxford University Press 2009) 86­­–90; Rheehuis and Heisterkamp, Pitlo: Goederenrecht: Deel 3 van Het Nederlands Burgerlijk Recht (Kluver Deventer 2006) 1–4, 18; Jan Wilhelm, Sachenrecht, 2nd ed. (Jan Wilhelm, Walter de Gruyter 2002) 1–4, 38–42; Hans J. Wieling, Sachenrecht, 5th ed. (Springer 2007) 3–5; Mark P. Thompson, Modern Land Law, 4th ed. (Oxford University Press 2009) 1, 5–10. 4   Semantically speaking, the explanation can become difficult. Virtual property does exist, but it is intangible; the same goes for other forms of intangible property such as those defined as objects of intellectual property (patents, copyright) or even property law in general—such as bank 1 2

636

BARFIELD PRINT.indd 636

13/11/2018 11:53

Law and property in virtual worlds  637 property or even intellectual property when it comes to the question of existence,5 since it is something that one cannot touch, cannot taste, and cannot pick up and take home.6 However, it is still property, and it still exists.7 Even though things like website addresses (URLs)8 and email addresses are often classified under intellectual property, it is more correct and more suitable for them to be reclassified as virtual property.9 One could even venture to include certain other accepted immaterial property objects such as bank accounts, stocks, options, and derivatives under the umbrella of virtual property.10 Indeed, one could go so far as to include digital goods, such as digital versions of books (ebooks),11 computer or smartphone programs or apps,12 television series, and movies,13 as well as digital music (albums and tracks),14 as objects of virtual property.15 However, the focus in this chapter will be on the types of virtual property found inside

accounts and stocks and shares, among other things. Virtual property falls into this same category, being intangible but not nonexistent.  5   See in general Erlank, Property in Virtual Worlds 258–64.  6   The aspects of intangibility and the (in)ability to touch, smell and feel virtual property are changing with developments in virtual reality equipment. It is quite possible that we will soon be able to interact with virtual objects to such an extent that they will be indistinguishable from real-life objects. This can be referred to as “crossing the conceptual barrier”. See the discussion of immersion as a key concept in the virtual world environment later in this article, as well as Erlank, Property in Virtual Worlds 43, 265–7.  7   In other words, it is still property in the classical or generic sense of the word. It forms part of a person’s patrimony and (depending on its allocation as a right or an object) can be seen as being the virtual equivalent of a thing or a chattel in the real world. See Erlank, “Determining the Property of Virtual Property” 87 fn 3; Erlank Property in Virtual Worlds 253–308. It must also be stressed that virtual property is not a subset of intellectual property. While there may be intellectual property rights attached to virtual property, the property itself is not a form of intellectual property. See J.H.M. (Sjef) Van Erp Servitudes: The Borderline between Contract and (Virtual) Property, in Towards a Unified System of Land Burdens (J.H.M. (Sjef) Van Erp & Bram Akkermans eds, Intersentia 2006) 1513–34; Erlank Property in Virtual Worlds 204–6, 258–64.  8   A Uniform Resource Locator (URL) is a name or series of easily readable characters which point to a specific internet protocol (IP) address. The URL can be associated with differing IP addresses and is usually easy to remember. For general information see Wikipedia Contributors date unknown http://en.wikipedia.org/wiki/Uniform_Resource_Locator.  9   See Erlank, Property in Virtual Worlds 261; Lelanie Bekker, Protection of Virtual Property in Terms of the South African Constitutional Property Clause, 2 European Property Law Journal 183–93 (2013) 184. See in general the discussion about various objects of virtual property: Fairfield, Virtual Property 1049, 1052; Lastowka and Hunter, The Laws of the Virtual Worlds 29. 10   See Van Erp, Servitudes 4; Fairfield, Virtual Property 1049, 1055; Erlank, Virtual Property— Protection and Remedies in Akkermans and Ramaekers (eds) Property Law Perspectives (Intersentia 2012) 10. 11   Wian Erlank, Books, Apps, Movies and Music—Ownership of Virtual Property in the Digital Library, 2 European Property Law Journal 194–5, 200–207 (2013). 12   Id 194–5, 198. 13   Id 194–5, 198, 210–11. 14   Id 194–5, 209, 212. 15   In terms of virtual property, form follows function. This means that if the virtual object was designed to mimic or fulfil the same function in a virtual world as it would have in the real world, it should be accepted as an object of virtual property. See Wian Erlank, Acquisition of Ownership Inside Virtual Worlds, De Jure, 770–82 (2013) 771; Fairfield, Virtual Property 1049. Also see in general, Erlank, Books, Apps, Movies and Music 194–212.

BARFIELD PRINT.indd 637

13/11/2018 11:53

638  Research handbook on the law of virtual and augmented reality virtual worlds—specifically, the types of objects (or things) that mimic their real-world counterparts.16 With regard to the other types of virtual property mentioned above, these topics are dealt with elsewhere.

2.  VIRTUAL WORLDS Before investigating the concept of virtual property in more depth, it will be beneficial to briefly mention two issues that are exceptionally important for one’s understanding of virtual property. The first issue refers to the origin of the concept of virtual property.17 The second issue refers to the importance of virtual property as a new object of property law.18 Virtual property originates from and can be found inside virtual worlds. This leads to the question about the definition of a virtual world. In plain, everyday language one could define a virtual world as an alternative nonphysical world, in contrast to the real, physical world we live in.19 However, for the purpose of this chapter the definition of virtual worlds developed by Bartle will be used.20 Bartle states that “virtual worlds are computer-moderated, persistent environments through and with which multiple individuals may interact simultaneously.”21 In order to understand Bartle’s definition, it will be broken down into its base elements and discussed below. One should take note that a number of these prerequisite elements, or essentialia,22 have correlations with requirements of virtual property,23 which will be discussed later in this chapter.24 The first requirement is that the virtual world should be computer moderated.25 A computer moderated environment means that a computer controls everything about the virtual environment. This is usually achieved by means of the program or the programmed

  A definition of virtual worlds is discussed in the next section.   For a more in-depth discussion of the history of and development of virtual property and virtual worlds see Erlank, Property in Virtual Worlds 17–42; Lastowka and Hunter, The Laws of the Virtual Worlds 14–29. 18   In other words, the fact that virtual property is accepted, and protectable, as a discrete object of real-world (private) property law, as well as an object of constitutional law. See Erlank, Property in Virtual Worlds; Bekker, Protection of Virtual Property 192–3. This should not be confused with so-called objects of “new property” that are generally accepted to be socioeconomic rights. See Charles A. Reich, The New Property, 73 Yale Law Journal 733–87 (1964). 19  Erlank, Property in Virtual Worlds 46. 20  Bartle, Pitfalls of Virtual Property (Richard Bartle, The Themis Group Durham 2004) (http://www.themis-group.com/uploads/Pitfalls%20of%20Virtual%20Property.pdf) 2. See also Lastowka and Hunter, The Laws of the Virtual Worlds 5, who focus on persistence and dynamics as essential characteristics of a virtual world. 21  Bartle, Pitfalls of Virtual Property 2. 22   See also Edward Castronova, Virtual Worlds: A First-Hand Account of Market and Society on the Cyberian Frontier, CESifo Working Paper No 618 (2001) 6; Erlank, Property in Virtual Worlds 47. 23  Fairfield, Virtual Property 1053–5; Erlank, “Determining the Property of Virtual Property” 89–91. 24   For a more in-depth discussion see Erlank, Property in Virtual Worlds 253, 272–85. 25  Bartle, Pitfalls of Virtual Property 2, Erlank, Property in Virtual Worlds 47. 16 17

BARFIELD PRINT.indd 638

13/11/2018 11:53

Law and property in virtual worlds  639 code. So, what does this mean in terms of virtual property? It means that all in-game aspects that have anything to do with virtual property are managed automatically or (in some circumstances) even autonomously by the computer, without the need for real people to have any sort of personal interference in the workings of the virtual world. One must also remember that the computers that manage these virtual worlds are not the same as one’s local desktop or laptop computer at home. These computers are typically arranged in a large array or matrix and linked to each other.26 The next requirement is that the virtual world must be persistent.27 In this sense persistence means that the virtual world should exist continually and without interruptions, specifically with regard to electricity and network connectivity. It is essential that the virtual world must always be available for the player to interact with. As soon as a virtual world becomes inactive or is switched off, it becomes a theoretical world which really does not exist, as opposed to a virtual one.28 This requirement of persistence is one of the chief reasons why virtual worlds are so addictive,29 because, from a player or user’s point of view,30 it would mean that he or she is missing out on things that happen in the virtual world when they are not logged in and taking part in it. The next requirement relates to the virtual-world environment.31 Under normal circumstances a virtual-world environment would be representative of one’s own real world, or at least be familiar and recognizable to one in some way or another. This is so that the world and its environment should be immersive and persuasive for the player.32 In a modern virtual world the environment is usually represented as being three-dimensional in nature, and one can usually navigate through the virtual world either from a first or third person perspective.33 The level of immersion is steadily increasing because of the improvement in both computer processing power and graphics card and gaming console capabilities.34 This increasing immersion can be seen in the latest developments in virtual reality gaming systems—Facebook’s Oculus Rift, Samsung’s Gear VR, and Google’s Cardboard. One should also note the importance of PlayStation VR,35 which brings a

  This collection of computers is also known as a server farm.  Bartle, Pitfalls of Virtual Property 2; Erlank, Property in Virtual Worlds 48. 28   See the discussion on the development of virtual worlds in Erlank, Property in Virtual Worlds 17–42, 48–51. 29   See Erlank, Property in Virtual Worlds 137 for more on the addictive nature of virtual worlds as well as social media addiction—where the afflicted person needs to be constantly online lest he or she misses something. 30   For purposes of this article and in fact, most discussions of virtual worlds, the words “player” and “user” are interchangeable. 31  Bartle, Pitfalls of Virtual Property 2; Erlank, Property in Virtual Worlds 51. 32   See Erlank, Property in Virtual Worlds 19, 25, 265–8 for a more in-depth discussion of the aspect of immersion (similar to “crossing the conceptual barrier” and “suspension of disbelief ”) and its effect on the player. 33   In other words, a first person perspective gives the player the impression that he or she is in fact viewing the virtual world straight out of the avatar’s eyes. From a third person perspective, the player will usually have an isometric topdown view of the virtual world environment where they can see the avatar in front of them from the same perspective that a bystander or third person could view the avatar. 34   See Erlank, Property in Virtual Worlds 43–6. 35  www.playstation.com/en-gb/explore/playstation-vr/. 26 27

BARFIELD PRINT.indd 639

13/11/2018 11:53

640  Research handbook on the law of virtual and augmented reality very sophisticated VR system to the masses of people who already own a PlayStation 4. Virtual worlds are also increasingly found in the use of so-called augmented reality (AR) programs and equipment such as the Layar program or the Google Glass spectacles. Additionally, there are the possibilities promised by startup tech companies such as Magic Leap,36 who work with visionaries such as Neal Stephenson, who sparked the imagination and possibilities of the virtual worlds with what he referred to as the “metaverse” in his 1992 scifi novel Snow Crash.37 Magic Leap’s AR/mixed reality and Microsoft’s holographic system,38 the HoloLens,39 represent a major step forward in the technology associated with virtual and augmented reality-based systems. Where VR and AR technologies mix, these are often referred to as mixed reality (MR) systems. Where one’s avatar would have been represented on the screen as a very abstract twodimensional (2D) collection of basic pixels 20 years ago,40 the modern version of the same avatar in the same game would consist of numerous layers of computer engineered visual wizardry to make the avatar look as close to real life as possible, in three dimensions (3D); now, with modern VR systems, the avatar not only looks real but is also (according to the perceived immersion of the player into a VR system) actually in the same environment as the player. Therefore, modern developers strive to make the worlds as photorealistic as possible. The choice of gaming environment may also steer away from the norm in order to recreate fantastical or totally fictional environments. One example of this would be a game that represents the Star Trek universe.41 If such a virtual world is to be successful, the developers would need to make it into as accurate a representation of the Star Trek universe as possible, in order for it to be easily recognizable to fans. In essence, the choice of environment is dependent on the subject matter of the game or the virtual-world backstory. It is in this category that virtual property becomes very important. The property (objects) or things that are available for use by the player in the world will help to create a sense of familiarity and believability for the player in the virtual world.42 These virtual property items will almost always be instantly recognizable and will be based on real-world items of property, such as an inn or hotel. The player immediately and instinctively identifies these things as property and not virtual items. The virtual property is also designed with the specific intention that it should be able to emulate its real-world counterpart in both form and function.43 So, instead of thinking that one is going into a “virtual inn,” one enters an inn and meets the innkeeper, not an “interpretation of computer data and bits.” Imagine what a virtual world based on the Star Wars films would be without a lightsaber!  www.magicleap.com/press/press_release5.pdf.   Neal Stephenson, Snow Crash (Penguin Books 1992). 38   www.magicleap.com/#/home. See also Wired, “The Untold Story of Magic Leap, the World’s Most Secretive Startup”, www.youtube.com/watch?v=BLkFWq_ipCc. 39   https://www.microsoft.com/en-us/hololens/hardware. See also www.youtube.com/watch​?​v​=​N​ w​Y-6sQDYnk. 40   An avatar is the character that represents you in the game. Seen in a different way, an avatar is the virtual-world personification of you. 41   This virtual world is called Star Trek Online. Other distinct “universes” include ones from Star Wars (Star Trek: The Old Republic), Harry Potter (Harry Potter Online), and Lord of the Rings (The Lord of the Rings Online), among others. 42   That is, contribute directly to the level of immersion, as discussed above. 43   See Erlank, Acquisition of Ownership inside Virtual Worlds 771; Erlank, Virtual Property 11. 36 37

BARFIELD PRINT.indd 640

13/11/2018 11:53

Law and property in virtual worlds  641 The next element refers to interaction.44 If a player is unable to interact with the virtual world, then the world loses its virtual-world status and merely becomes an abstract artwork or movie. Interaction usually takes place by means of the player manipulating his or her avatar in such a way that it can interact with other avatars and virtual property objects. There is a tangible relationship between the level of immersion and believability in the virtual world for the player, on the one hand, and the level of interactivity, on the other.45 As such, a big part of interactivity lies in the level of manipulability and destructibility of virtual-world items, that is, virtual property objects.46 Gamers seem to find it more fun to be able to interact with and destroy things in a virtual world.47 This immediately makes one think of one of the perks (competencies) of ownership: One is able to destroy one’s own property. Interaction is important not only for the player–avatar–object relationship, but also for the social relationships that players have with other players.48 Players can interact by going on quests together, talk, eat, and even kill each other (in the virtual sense).49 With AR and VR, interactivity is taken to a new level. While a player would previously manipulate his or her avatar as well as items in a virtual world by means of a keyboard and mouse combination, or even a joystick controller, with AR and VR the manipulation and interaction is much more “hands on.” Either a camera tracks the user’s gestures in the real world and translates them into the virtual world, or, at the very least, much more sophisticated controllers are used, in that they often necessitate actual comparable physical movements in the real world to interact with the virtual objects. This level of immersion is comprehensively facilitated by such VR/AR peripherals as the PlayStation Move and Aim controllers (used in conjunction with the PlayStation Camera, which is 3D-depth sensing);50 as well as the Oculus Rift’s Touch controllers and Oculus Sensors.51 The more immersive and intuitively similar the interaction with a virtual-world environment and objects, the less a person needs to consciously suspend their disbelief. That is, they do not need to make any effort to consciously ignore the fact that the virtual world and property is not “real.”

44   See Bartle, Pitfalls of Virtual Property 2, Erlank, Property in Virtual Worlds 52. It is, however, quite possible to argue that a player or user could possibly acquire and or assert property rights to virtual property in a single user environment, such as in a single player VR application where a player is able to create new items either through their own labor (Lockean labor theory) or by other means, such as the purchase of in-game items. This interpretation of virtual worlds relies on the multiplayer aspect of virtual worlds, but it is quite possible to argue for the existence of virtual property and virtual worlds from a different conceptual perspective that does not rely on multiplayer interaction. 45   See the discussion above. 46  Erlank, Property in Virtual Worlds 53. 47   Id fn 95. 48   For a documentary perspective on the social aspects of virtual worlds, you could look at the documentary film called Second Skin (2008), also referred to later in this article. See Pure West Films 2008 http://secondskinfilm.com. 49   See Erlank, Property in Virtual Worlds 54–7, 70–74, 137–41 for a discussion about the social aspects of virtual worlds. 50   Sony “PlayStation VR accessories,” www.playstation.com/en-gb/explore/playstation-vr/acces​ sories/. 51   Oculus “Accessories,” www.oculus.com/accessories.

BARFIELD PRINT.indd 641

13/11/2018 11:53

642  Research handbook on the law of virtual and augmented reality The last element is that there must be participation by multiple individuals.52 This is a key component of a virtual world, and also for the existence of virtual property. If there was no opportunity for multiple people to take part in the virtual world at the same time, it would mean that it is just a normal game that one plays by oneself on a PlayStation or Xbox. It would not be a virtual world. As such, the individuals who join up to participate in a virtual world behave in much the same way as they would have in real life; the only difference is that they now do it in a virtual environment. It is because of this multiplayer element that most of these virtual worlds are referred to as Massively Multiplayer Online Role Playing Games (“MMORPGs”). Many of these virtual worlds give one the ability to customize, in almost all aspects, the way one’s avatar looks. One may be able to choose an avatar’s race, allegiance, skin and hair color, facial features, body type, sex, clothing; even its moral values. Appearance in virtual worlds goes even further than just changing the clothes or race of one’s avatar. Many virtual worlds make provision for one to use body language to show one’s avatar’s feelings. Avatars can (be instructed to) yawn, tap a foot in impatience, or shrug, among other things. Because of this, players not only form a bond with their own avatars, but are also able to recognize other players by the appearance of their avatars. In other words, a person-as-player becomes so immersed in the virtual world that the other player’s avatar becomes as real as if the first player knew the second in real life. This is even more so with VR and AR, where a player’s avatar can mimic their actual body movements in real time in the virtual world. One’s avatar can also now be photorealistically immersed in the virtual world. However, while most users tend to identify strongly with their own avatars, the more realistically an avatar represents one’s actual self, the more one finds that there is a reason for real-world law to intrude into the virtual world. One such area in which things can, and do, go wrong is cyberbullying and virtual assault and rape. These issues are discussed elsewhere in this book, so I will not go into more depth in this chapter. Suffice to say that the more one identifies personally with one’s avatar, the more immersive the experience is for the user, and of course the more one tends to take an attack on one’s avatar personally. The last aspect to touch on here is the fact that all virtual worlds make provision for players to communicate with one another.53 Some of the very old virtual worlds still have speech bubbles appearing over the heads of avatars in a comicbook fashion,54 showing what they are saying. But as technology and the speed of internet connections have advanced, players are able to make use of voice over internet protocol (VoIP) technology to chat to each other in real time in the same fashion as one would make a normal telephone call. All of this is designed and incorporated into the game to further enhance the social aspects of the game. Additionally, even outside of the “normal” virtual worlds, the gaming platforms themselves run massive de facto virtual worlds, at least in terms of participation of multiple individuals—where the focus is on communication between players. Both Sony and Microsoft (to name but two) focus on their social platforms, the PlayStation Network (Sony) and Xbox Live (Microsoft), which facilitate multiuser com-

 Bartle, Pitfalls of Virtual Property 2, Erlank, Property in Virtual Worlds 54.  Erlank, Property in Virtual Worlds 55–6. 54   See a discussion of the history of virtual worlds in Erlank, Property in Virtual Worlds 17–43 and in Lastowka and Hunter, The Laws of the Virtual Worlds, 14–29. 52 53

BARFIELD PRINT.indd 642

13/11/2018 11:53

Law and property in virtual worlds  643 munication and participation over and above any discrete game that a player is running at any given time. Apart from the examples previously mentioned, some of the most well-known virtual worlds are World of Warcraft, Everquest, and Second Life. While these virtual worlds have been extensively featured in academic discourse over the past few years, one should also not forget to mention the newer generation of “classical” virtual worlds, such as Final Fantasy XIV, Elder Scrolls Online, Destiny, Neverwinter, DC Universe Online, Star Trek Online, and Trove—to name but a few. From the preceding discussion, it is evident that participation in an online virtual world is not a lonely affair which is left to a select (and usually clearly defined) few,55 but has already become a way of life for a vast majority of the population. In fact, many people find that they are able to participate better in social circles by means of virtual worlds than is the case in real life. For example, in the virtual world of Second Life, there is an island that was created with the specific intent to house players (via their avatars) who suffer from Asperger’s Syndrome.56 The virtual environment gives them a safe space in which they can practice their social skills in nonthreatening situations.57 Another relevant example is the fact that many educational institutions are giving lectures in virtual university classrooms in Second Life,58 and more than 300 universities make use of Second Life as an educational tool,59 enabling students from all over the world to see and participate in the lectures via the virtual-world interface.60 It is clear from these two examples that virtual worlds can also be used for serious and educational purposes and should not be regarded simply as games.61

3.  SOME ECONOMIC ASPECTS AND PROPERTY THEORY Before moving on to the specifics of virtual property, a few statistics about virtual worlds and the property contained therein should be mentioned. Castronova describes   In other words, it is not the sole domain of social outcasts and geeks anymore.   Asperger syndrome can be defined as “[a]n autistic disorder most notable for the often great discrepancy between the intellectual and social abilities of those who have it.” www.medicinenet. com/script/main/art.asp?articlekey=9675. 57   As an example, one of the private islands in Second Life is called “Brigadoon.” This is a safe, virtual environment where people with Asperger’s syndrome and their caregivers can interact. http://abcnews.go.com/Technology/FutureTech/story?id=1019818. 58   For example, both the University of Hawaii and East Carolina University offer courses set inside Second Life. See www.hawaii.edu/secondlife/; www.ecu.edu/cs-itcs/secondlife/; http:// usatoday30.usatoday.com/news/education/2007-08-01-second-life_N.htm. 59  http://usatoday30.usatoday.com/news/education/2007-08-01-second-life_N.htm. 60   This is an important aspect for the provision of distance-based education. The fact that there is a familiar three-dimensional environment, in which students can interact with the lecturer as well as other students, makes it more conducive to bridging the distance gap. 61   Many other examples exist. Among others, virtual worlds and the associated technologies are used for military training (insurgency and tactics as well as community interaction in hostile areas), pilot training (flight simulators), and medical training (surgical and clinical training), and even NASA uses virtual reality simulations to train astronauts for things such as repair missions and space walks. 55 56

BARFIELD PRINT.indd 643

13/11/2018 11:53

644  Research handbook on the law of virtual and augmented reality the economy of Everquest’s Norath in the same way as if it were a normal real-world economy.62 He provides statistics relating to such activities as production, labor supply, income, inflation, and foreign and currency exchange. He notes that by the late summer of 2001, Everquest had a subscriber base of more than 400,000 players and showed growth of more than 10 percent over two quarters.63 At that stage Sony’s monthly revenues from Everquest alone amounted to approximately $3.6 million! Bear in mind that this was 16 years ago and only in relation to one virtual world. At that stage, the average Everquest player spent an average of 20 hours a week playing the game. A documentary, Second Skin, dealing mostly with the social and interrelationship aspects of virtual worlds was released in 2008.64 This documentary provides slightly more up to date statistics, from 2007. What this brief example boils down to is that virtual worlds equal money, both for the developers (creators) and for the participants (users/players). They are heavily inhabited (by real-world players/users/inhabitants) for whom the virtual world(s) do not seem trivial. In Castronova’s study he concludes that even though the behavior that is elicited by players might seem to be a matter of addiction,65 there must be some other factor responsible for their behavior. He finds that a player in a virtual world experiences an alternative reality or a “different country” in which he or she may choose to spend most of his or her life as he or she wishes, free from the constraints of reality.66 In the competition between Earth and the alternate virtual world, Earth then seems to be the less appealing option.67 Bearing in mind the close connection between users of virtual worlds and their avatars, the worlds themselves, the economic value contained, and how much users tend to identify with their avatars as virtual-world personifications of themselves, it should come as no surprise that many people tend to identify with their virtual-world property in the same way as with their real-world property.

4. VIRTUAL PROPERTY AS A DISCRETE OBJECT OF PROPERTY LAW As mentioned at the start of this chapter, virtual property is often (incorrectly) categorized as a subset of intellectual property. While a virtual property object could have intellectual property rights attached to it, it is not correct to say that it is purely limited to being an object of intellectual property law. This does not in any way affect the interests of existing intellectual property rights holders.68 To explain this, one could use the following example  Castronova, Virtual Worlds.   Which was up from about 320,000 subscribers in March 2001. See Castronova, Virtual Worlds 9. 64   See www.imdb.com/title/tt1183696/?ref_=fn_al_tt_1; http://secondskinfilm.com. Please note that this is the 2008 edition of Second Skin. Other films by that same title may not be suitable for family viewing. 65  Castronova, Virtual Worlds. 66   Id 10. 67   Id 10. 68   This does not in any way affect the interests of existing intellectual property rights holders. See Erlank, Books Apps Movies and Music 210. 62 63

BARFIELD PRINT.indd 644

13/11/2018 11:53

Law and property in virtual worlds  645 of a book that one purchased at the local bookshop. After purchasing the book, you can consider yourself to be the owner of the book. This means that it is “yours.” You can (and hopefully will) read and enjoy the book. Afterward you are able to abandon (discard/throw away) the book, sell it, destroy it, encumber it (by pawning it), or lend it to someone. These are some of the competencies of ownership that everyone assumes as naturally occurring. While one is now owner of the book (the physical, tangible paper object itself), one is not the owner of the intellectual property rights attached to the work. You may not copy it, distribute it, change it, and so on. The same applies (or in any case should apply) to digital or virtual versions of the book. If the only difference between the actual book on your bookshelf and a virtual ebook is the tangibility or intangibility of the object, why should you try to remove the competencies of ownership from the virtual version by negating the fact that it too is an object of property law, and purely subject to intellectual property licensing regimes? I deal with this issue in much more depth in other publications. Moringiello explains how and why it is problematic to equate virtual property and intellectual property.69 Typically, when lawyers hear the term virtual property, they immediately think of intellectual property. Courts tend to think in the same way. To illustrate this, Moringiello discusses the US domain name case of Dorel v. Arel,70 which dealt with the question: Can a judgment creditor reach a domain name by means of garnishment?71 The problem was that under the applicable statute that applied, only a property right could be garnished. In order to make the outcome fit the facts, the court looked to trademark law as a subset of IP law for help. The court then concluded that if a domain name is eligible for trademark protection, it is seen as property and therefore also garnishable; if not, then it is not property and hence not garnishable. Moringiello explains the problem in the following way: If a domain name is eligible for trademark protection, it cannot be transferred without the goodwill of the business to which it is attached, and as a result is of nearly no use for the creditors. On the other hand, a generic domain name which is incapable of being protected by trademark can be extremely valuable to creditors, because it can be transferred for a substantial amount of money.72 In other words, it is clear that trying to squeeze in a domain name under the category of a trademark meant it lost all its value to the creditors. Thus, one needs to look at redefining it as a virtual property right that is capable of subsisting on its own. Now, that aside, the problem to identify here is the fact that disputes that take place only in a virtual world tend to mirror disputes which take place in the real world. As such, they tend to be solved in the virtual world in a manner analogous to the way in which the real world would have dealt with

 http://lawprofessors.typepad.com/property/virtual_property/.   Dorel v Arel 60 F. Supp. 2d 558 (E.D. Va. 1999). 71   For civil law readers: garnishment here could be equated to seizure or attachment. 72   To understand this, you need to bear in mind that the company has gone bankrupt. Once a company has been declared bankrupt or even if rumors about such an incident start, you will find that the goodwill pertaining to the company quickly dissipates. Think of the runon case of DSB bank, which caused its collapse. After the bank went under, no one would have been interested in buying its domain name. On the other side of the coin, if the company in question had a generic domain name such as business.com, it could have made a tidy sum of money. Business.com recently sold for $7.5 million. 69 70

BARFIELD PRINT.indd 645

13/11/2018 11:53

646  Research handbook on the law of virtual and augmented reality them.73 To be able to deal with disputes relating to virtual-world assets in the real world, one should conceptualize it starting from the fact that it only deals with intangibles. Moringiello sums up by saying that we ought to put aside tangibility as a determining factor in viewing virtual property, and rather concentrate on the other aspects of property rights.74 Moringiello also wrote a paper in which she argues that the principle of numerus clausus should be applied to virtual property.75 The numerus clausus (literally meaning “closed number”) principle is a central tenet of many countries’ property law systems—usually those in the civil law tradition. It refers to the concept that there are only a limited (closed) number of accepted (or real, as in realty) property rights. She argues that because the questions that relate to the legal nature of digital assets in virtual worlds tend to mirror the questions that arise in connection with intangible rights in general, this leads to understandable confusion about the classification of these rights as contractual or property rights. By applying the principles of numerus clausus to virtual property, the courts that have to deal with disputes relating to virtual property should be able to resort to mandatory property types or categorizations. In this way, people (such as virtual-world developers) could be limited in their ability to muddy the waters of property,76 by customizing such rights by means of the contracts that are usually created by the electronically presented click-wrap forms and End User License Agreements (EULAs).77 Another approach to solving the problem that courts will face when having to adjudicate matters dealing with virtual property makes use of the application of a nuanced test.78 In brief, this test is used to determine, first, if a virtual object should be accepted as an object of virtual property; second, if such virtual property object could be regarded as a form of real-world property; and last, whether such virtual property object should be protected as an object of real-world property law.79 Up to now I have discussed what virtual worlds are, and how players interact with them. I have also looked at the interaction between the player, the virtual world, and virtual property. But now the following critical question should be addressed. I think we can all agree by now that virtual property is important—but what is it? Is it a contractual relationship? Is it a right? If so, what type of right? Could it be a property right? To help with these questions I would like to briefly refer to Fairfield’s classification of

  Hence, as discussed above, form follows function.   An interesting aside here is the fact that even the media has started to refer to URLs as virtual property. In a Reuters article, it was mentioned that General Motors have “roped off a bit of virtual real estate” which they could use in the event that they would seek bankruptcy protection. They registered gmrestructuring.com and gm-restructuring.com in early April 2009. See www. reuters.com. 75   Juliet M. Moringiello, Towards a System of Estates in Virtual Property, SSRN https:// papers.ssrn.com/sol3/papers.cfm?abstract_id=1070184 (2007). For a discussion of the application of the numerus clausus as a principle of the “law of virtual things” see Erlank, Property in Virtual Worlds 313–15. 76   See in general Erlank, Property and Sovereignty in Virtual Worlds. 77   For more on how EULA’s affect virtual property see Erlank, Property in Virtual Worlds 98–119; Erlank, Determining the Property of Virtual Property 91. 78  Erlank, Property in Virtual Worlds 389–407. 79   See Erlank, Virtual Property 15–19; Erlank, Property in Virtual Worlds 389–407. 73 74

BARFIELD PRINT.indd 646

13/11/2018 11:53

Law and property in virtual worlds  647 virtual property.80 In his article entitled “Virtual Property,” Fairfield claims that virtual property needs to have three characteristics (which we could also call essentialia). These essentialia are that virtual property must be rivalrous, must be persistent, and must have interconnectivity.81 Do these requirements sound familiar? If they do, it is because they are similar to the requirements (discussed earlier) that are needed for a virtual world to exist. Of course, one can decide to add to these characteristics, but for brevity’s sake I will look only at the requirements mentioned by Fairfield.82 The rivalrousness of virtual property refers to the fact that only one person owns and controls the property.83 Because of this he or she is able to exclude others from doing the same. We could also just call this the competency of exclusivity. The rivalrousness of such property could be equated to what is called scarcity in economic terms, and is what creates the economic value of the property. The second characteristic is persistence,84 which goes hand in hand with the persistence requirement for virtual worlds. If a piece of property is persistent, it means that it does not cease to exist, or disappear permanently, when a virtual-world citizen logs out of the virtual world and turns his computer off. If the property was not persistent, then no one would be willing to invest in the acquisition of the property and it would become valueless. This brings us to the final characteristic: interconnectivity.85 This is the same as the virtual-world requirement of interaction. Interconnectivity enables multiple players to be online and interact within the same virtual world, making use of the same common resources. It must be possible for all of this to occur simultaneously. The fact that people are able to visit and use the same virtual property immediately increases the property’s value. One would also want other avatars to be able to see and experience the new house that one has created in Second Life. They can come and visit and interact with one’s property by sitting on the couch and pouring themselves a virtual glass of wine. With no interconnectivity, only you would be able to see and experience your property. The concept of “ownership” is a problematic one inside virtual worlds and it is often easier (and probably more correct) to describe players’ property rights toward the things they “own” in virtual worlds as possessory.86 The reason for this is that the right of ownership is normally inferred from the fact of possession. Because a player is not enabled by the game’s code to possess an item “belonging” to another player, the mere fact of possession often equates to ownership inside the virtual world.87 In many virtual worlds,

 Fairfield, Virtual Property 1047.   For a more in-depth discussion of the characteristics of virtual property see Erlank, Property in Virtual Worlds 272–85. 82   For the analysis, I will use Barfield’s discussion. Woodrow Barfield, On Money, Taxes, and Property in Virtual Reality 13 Virtual Reality 37–9 (2009). 83  Fairfield, Virtual Property 1053; Blazer, Five Indicia 143; Erlank, Property in Virtual Worlds 275. 84  Fairfield, Virtual Property 1054; Blazer, Five Indicia 144; Erlank, Property in Virtual Worlds 278. 85  Fairfield, Virtual Property 1054–5; Blazer, Five Indicia 145; Erlank, Property in Virtual Worlds 280. 86   James T.L. Grimmelmann, Virtual Worlds as Comparative Law, 49 NYL. Sch. L. Rev. 147–84 (2004) 151. 87   This is a basic rule of thumb and certain virtual worlds do in fact allow some form of leasing or lending, which means that they distinguish between ownership and some holders (or 80 81

BARFIELD PRINT.indd 647

13/11/2018 11:53

648  Research handbook on the law of virtual and augmented reality property rights attach to a possessor of an object for as long as the item is carried around by the avatar.88 As soon as the avatar drops the object,89 this constitutes abandonment and the item immediately becomes res derelicta.90 The next person to pick up the item will become the new owner.91 The example mentioned here is a very simplistic use of the ownership concept in a virtual world and many worlds are far more sophisticated in their approach to the ownership concept, as well as in their implementation of it via code. This is due to the pressure on developers to implement more detailed code-based property rules and not only a simple “possession is all” rule.92 An example of a more complex code-based property system is that of virtual homeownership.93 The default rule in virtual worlds is that all players have access to all areas at all times. The next, more developed rule is to allow only the owner of a house into his virtual home, which is counterintuitive if a homeowner wishes to invite guests over without giving them access to roam freely around the premises. Since many of the virtual worlds have socialization at the center of their existence, the ability to invite others into your personal space and have them interact with you and your possessions are of paramount importance. To solve this dilemma, developers have created a nuanced system of levels of exclusionary capacity, which they build into the code-based property system. Grimmelmann refers to this (making use of English property law terminology) as a “virtual fee simple” with a new estate carved out called the “right to visit.”94 This so-called right to visit is a perpetual, nontransferable right and is subject to revocation by in wider terms, possessors), therefore not always equating possession with ownership. However, one should remember that this rule of possession equating ownership is the de facto code based rule with which a virtual world will be programmed. It also forms the foundation of the property systems in virtual worlds that do make a distinction. One should bear in mind that insofar as the system is regulated by the game code, the distinction between ownership and lease is just a more nuanced implementation of the game code. 88   This is often referred to as “equipping” an item. See http://everquest2.com/manual /EQII_ Manual.pdf 16. 89   Or “unequips” the item. In certain instances, an object is “bound” to the user and can never be transferred to another player; the act of abandonment or “dropping” results in the item’s demise and removal from the virtual world. 90   One should bear in mind that the fact that a legal system (real or virtual) acknowledges abandonment does not mean it equates ownership with possession. All legal systems will in some cases equate possession with ownership by means of original acquisition of ownership, while making a distinction for other instances where ownership and possession is split by, for example leasing, lending, pawning, and even servitudes. If virtual worlds recognize such instances they are similar to real worlds. However, this is the basic rule that will be coded into a virtual world on which developers can build and create more nuanced systems of complexity to mimic the real world. 91  Grimmelmann, Virtual Worlds 151. For example, Ultima Online’s Playguide describes the procedure and consequences of dropping items as follows: “You can remove items from your inventory and drop them almost anywhere on the screen near your character. If, for some reason, the item can’t be dropped in the location you’ve selected, either an error noise will sound or the item will return to your inventory. A dropped item will stay where it is until it deteriorates naturally or someone picks it up. Dropped items don’t tend to stay around for long.” www.uoherald.com/ node/115. 92  Grimmelmann, Virtual Worlds 151. 93   See also id. For an example of how a developer deals with this issue, see http://everquest2. com/manual /EQII_Manual.pdf 26. 94  Grimmelmann, Virtual Worlds 152.

BARFIELD PRINT.indd 648

13/11/2018 11:53

Law and property in virtual worlds  649 the owner of the house at any stage, making it a typical precarious right. When applied to civil law, it could be described as a “virtual ownership right” with a new entitlement (or stick in the bundle of ownership that is carved out for eligible visitors) called the “right to visit.”95 This right to visit has a measurable effect on the ability of both owner and visitor to interact with their own property as well as the property of others. In the example above about ownership being lost by the dropping of an item, the rules of the acquisition of the object in question were clear. However, if one were to integrate that example into the context of a virtual home in a virtual world, a number of problems arise. One such problem (and a main reason why any player would wish to own a home, whether virtual or real) would be to have a secure environment where they could store their collected movable virtual items without losing their property rights in it.96 Apart from this need to store movables, homeowners would also like to have guests over for functions and other social visits. However, what would be the consequences of inviting visitors to a player’s house? Would an item dropped by a guest become the property of the homeowner, or would any guest that picks up or uses an item in the house become the new owner of that item? It is clear that in this case, the general rule of possession equating virtual ownership is not satisfactory. In order to solve this problem, the nuanced system of exclusion mentioned above incorporates a number of new options. A homeowner can choose between certain categories of access that he or she wishes to assign to a visitor. Depending on the category that a homeowner assigns to a visitor, that visitor will not only be able to access certain areas of the home, but the visitor’s interaction with the items in the home will also be defined. For example, one visitor might have access to enter the house, but not to pick anything up. Another visitor might have permission to pick items up in the house but not to remove the item, while yet another visitor might be given circumscribed permission to carry off any item in the home.97 The above example shows how virtual homeownership tends to mimic real-world property frameworks as well as how it underscores a number of the main property theories, including property as exclusion and as a catalyst for human flourishing, as well as the interplay between the private right to exclude and the public interest to prevent exclusion. That is, it clearly illustrates the social function of encouraging private ownership.

95   This will only happen in a virtual world, since the level of control that is given by the game code will mean that a license is worth a lot more than in the real world. In the real world, it is questionable whether a precarious right like this amounts to carving out a stick since in English law this will probably be a called a revocable license. However, as applied to the content of virtual worlds this makes for a good analogy. 96   Even if items are not taken by other players, they are subject to “decay,” as is explained in the Ultima Online Playguide. “In Ultima Online, items that are placed on the ground (i.e. not locked down or secured in a house, or placed in your bank box) can decay. This means that after a period of time, the item will disappear from the game. There’s no way to retrieve items that have decayed. To keep your items safe from decaying, you need to either lock them down or place them in a secure container in a house that you own, co-own, or are a friend of, or store them away in your bank box.” See www.uoherald.com/node/216. 97   The level of complexity with regard to rules of access can become extremely high. For a good example of how such rules work, see www.uoherald.com /node/216 and also www.­camelotherald. com/housing/manual/chapter4.php.

BARFIELD PRINT.indd 649

13/11/2018 11:53

650  Research handbook on the law of virtual and augmented reality Maybe a mental breather is in order here so that one can briefly think about the situation. By this stage most of you will concur when I say that there is such a thing as virtual property, and that it does indeed exist. The problem is just that it is not generally recognized by real-world legal systems and courts. If we were to concede that one is able to have rights in virtual property (apart from the contractual rights that one acquires from the EULA), it must also follow that someone is able to own this property. Let’s leave aside the question of the owner’s identity for the moment and return to that later. Once we accept that one can have virtual property ownership,98 we need to look at the consequences of this. Barfield mentions some interesting issues.99 The first interesting aspect of virtual property ownership relates to the economic impact that it has on the real world. Keeping in mind the statistics discussed earlier on, the following bears some reflection. Economists have estimated that as early as 2002, the real-world trade in virtual items was responsible for creating a combined gross national product of virtual worlds which was equivalent to, or more than, that of some developing countries.100 The second interesting thing to take note of is the fact that the sales market for property in virtual worlds sometimes moves into the real world. The developers of the virtual world Entropia Universe have contemplated issuing players with a real-world ATM card with which they could draw real-world money out of their virtual-world bank accounts.101 This would be automatically converted to real-world currency by the virtual world’s treasury. Developing from this concept, one is now able to withdraw funds from the virtual currency Bitcoin (which in itself is a form of virtual property) via ATMs located across the globe.102 This immediately raises concerns about tax law, money laundering, and currency exchange regulations. However, these concerns are also valid for normal transactions that take place in the real world. For this reason, it is inevitable that governments have started to take an interest in the regulation of the sale of virtual property. In fact, even though there is still a large amount of academic debate about whether virtual-world assets should be taxed,103 it has already started to become a bit of a moot point.104 The US internal revenue service is continuing to investigate the taxation possibilities of virtual assets and Australian tax officials have stated that as soon as a virtual-world transaction has any real-world  98   For an explanation of how ownership of virtual property can be acquired (and transferred) both originally and derivatively, see Erlank, Acquisition of Ownership.  99  Barfield, Virtual Reality 37–9. 100  www.terranova.blogs.com/terra_nova/2004/08/virtual_world_e.html. 101  www.marketwired.com/press-release/Entropia-Universe-Debuts-First-Ever-Real-World-AT​ M-Cards-Easy-Access-Virtual-Funds-686615.htm. 102   Thieves have also seen the value of this and have been actively stealing Bitcoin ATMs and also “robbing” people of their Bitcoins by forcing persons to hand over their virtual property. www.nltimes.nl/2015/02/18/third-amsterdam-bitcoin-atms-stolen; http://www.businessinsider.com/ new-york-city-bitcoin-stagecoach-robberies-2015-2. See also the discussion of the Dutch case of virtual property theft below. 103  Barfield, Virtual Reality 38. 104   For a discussion about taxation of virtual property see Erlank, Property in Virtual Worlds 131–5. See also Leandra Lederman, “Stranger than Fiction”: Taxing Virtual Worlds, 82 New York University Law Review 1620–72 (2007) www.gamespot.com/pc/rpg/there/news.html?sid=6162654&om_act=con​ vert&om_clk=newsfeatures&tag=newsfeatures;title;1; South African Income Tax Implications of Income Earned in Virtual Worlds (Sarah J. Pienaar, LLM dissertation, University of Pretoria, Pretoria 2008) iii.

BARFIELD PRINT.indd 650

13/11/2018 11:53

Law and property in virtual worlds  651 consequences it will attract the attention of the Tax Office.105 Following on from this possible taxation of virtual property, the whole debate about the ownership of virtual property suddenly becomes much more interesting. If developers continue to argue that players only have licenses to use a game (as they do in most cases), the developers may themselves become liable for taxation of the transactions as a result of their desire not to share the joys of ownership with the players.

5. THE PROBLEM WITH VIRTUAL PROPERTY AND THE TANGIBILITY ISSUE 5.1 Introduction In the preceding sections it has already been shown that virtual property is not purely relegated to being a subset of intellectual property, and it has also been shown how it can form a discrete object of property law, but it should be noted that this was with specific reference to the common law (Anglo-American legal) tradition. However, since virtual worlds and virtual property are accessed, used, and traded internationally, we cannot only look at one legal system. Broadly speaking, the problem concerning property in civil law (Roman-Germanic) systems stems from their narrow definition of property. In most civil law systems, “property”—as we colloquially refer to objects of property law—is referred to as “things,” and a “thing” has a very specific set of characteristics, of which one of the most important is that it must be tangible. Of course, this becomes somewhat of a conundrum when we consider that for all intents and purposes, virtual property is “intangible” or at the very least “incorporeal” or “immaterial.” In most common law jurisdictions, this is not much of an issue, since they use a wider definition of property which tend to focus on rights to/in property rather than objects—and “property” can, broadly speaking, include almost any object that has patrimonial (or monetary) value.106 5.2  The Characteristics of Virtual Things In this section, I will apply the civil law characteristics of real-world things (corporeality, external to humans, independence, appropriability/susceptibility to human control, use and value) to virtual things, in order to determine whether and how virtual things can be compared to real-world things. In short, only the characteristics of corporeality and externality differ from their real-world counterparts. However, the fact that virtual objects are intangible should not by itself stand in the way of virtual property being recognized as property in the real world. This is already not much of an issue since intangibles like intellectual property are already recognized and protected as property in the real world. With

105   www.house.gov/jec/news/news2006/pr109-98.pdf; http://www.theage.com.au/news/biztech/ virtual-world-tax-man-cometh/2006/10/30/1162056925483.html. 106   For a more in-depth discussion of the differences in the understanding of the concept of property, see Erlank, Property in Virtual Worlds 212–31.

BARFIELD PRINT.indd 651

13/11/2018 11:53

652  Research handbook on the law of virtual and augmented reality regard to the requirement of externality, it will be shown that this is also not problematic for the recognition of virtual property. With that in mind, it is important to note that the virtual-world thing is usually designed to mimic a tangible thing found in the real world. This mimicry extends to almost all aspects of the characteristics of a thing, apart from the fact that a virtual thing cannot be tangible in the real world—yet. VR, AR, and mixed reality might necessitate a reconceptualization of this. However, players have no qualms about regarding virtual items as tangibles, since they cross the conceptual barrier (that is, choose to ignore that the virtual world is not real) when dealing with virtual property. Therefore, one could argue that because the item is tangible and corporeal inside the virtual world and is regarded as a tangible thing in the real world by the players, it should be regarded as property by real-world legal systems. While one can make this argument, one should also bear in mind that this is just one more justification for the protection of virtual property as property in the real world. 5.3 (In)corporeality107 In a virtual world, the requirement that a thing must be observable by any of the five senses as well as occupying three-dimensional space is fulfilled by the vast majority of legally recognized things.108 Because of the crossing of the conceptual barrier and due to technological advances, virtual items are represented in terms of the space that they occupy and they can be seen in the position where they are occupying that space. It could also be argued that a player can touch a virtual thing and exert physical control over it by means of controlling his or her avatar. As discussed earlier in the chapter, the increasing level of immersion into virtual worlds through the advent of widely available

107   When I apply the characteristic of corporeality to virtual world things I define them as (in) corporeal due to the fact that they are not regarded as corporeal in terms of real-world physics, but are regarded as such by virtual-world players. 108   This requirement just applies to a specific area of virtual property which deals with objects that theoretically fulfill the same form and function in both the real and virtual worlds. As such, these virtual property objects are virtual property in the “narrow” sense. In the “wider” sense, which includes both rights and objects in the understanding of “virtual property,” other more nonobvious things could be understood to form part of virtual property. As pointed out by Marc Blitz (correspondence on file with author), it could be possible to include almost any type of “in-app purchase” as a form of virtual property. For example, Blitz has the following question: “if I pay to endow my avatar with X-ray vision or additional strength, or a spaceship I have purchased with the power to enter different kinds of dimensions, is that personal virtual property? What about an additional life in a virtual game where one loses ‘lives’ . . . when one experiences certain kinds of harms (or perhaps ‘hit points’ in virtual D&D). It doesn’t seem right to say these kinds of powers are ‘observable by any of the five senses as well as occupying three-dimensional space is fulfilled by the vast majority of legally recognized things.’ . . . Nor does it seem right to say these powers are thus intellectual property. But they do seem to be some kind of property if I pay for them or they otherwise have value for me.” In short, the answer to this will be that these virtual property “rights” are understood to be included in the “wider” understanding of virtual property, where in most cases, the fact that they have value (economical or sentimental) will be indicative of a possible (personal) property right. For more on this, refer back to the introduction to this chapter and especially fn 9.

BARFIELD PRINT.indd 652

13/11/2018 11:53

Law and property in virtual worlds  653 VR, AR, and MR platforms has the effect of closing the divide between the real and virtual. Nowhere is this more relevant than when discussing the corporeality of virtual objects. After all, what defines corporeality? Is it the aspect of something being tangible? But how does one define tangibility, that is, the ability to touch or feel something? Should this tangibility be relegated to being able to touch something with one’s bare hands? Is the fact that one cannot (or rather should not) touch molten glass with one’s bare hands relevant to deciding whether molten glass is a thing? I would suggest that we approach the requirement of corporeality with some common sense. If one were to accept that something can be touched or manipulated by some sort of mechanical mechanism, such as remote-controlled pincers, or through an artificial barrier, such as a glove, then it would logically follow that an object that one can manipulate by means of a controlling interface such as a keyboard/mouse/joystick/game controller/or VR/AR-sensing interface should also be regarded as tangible. As such I like to think of things/objects inside virtual worlds as being in(corporeal), since inside the virtual world they are tangible and corporeal to the avatar, and by proxy therefore to the controlling user in the real world. This becomes even more relevant when one looks at the latest developments in AR or mixed reality systems, such as Microsoft’s HoloLens. Take, for example, the fact that one is able to view one’s actual surroundings while wearing the HoloLens headset. The HoloLens AR system then inserts virtual objects into the real environment that, at least from the wearer’s perspective, look as if they are physically present. For example, it gives one the ability to have a number of large display screens appear all around one’s office. In terms of form and function, while one is wearing the headset, these screens fulfil the same purpose as would real display screens—but without the need for them to be real objects. Thus, the need for a number of corporeal, expensive, electricity guzzling and dust collecting objects is obviated by their replacement with their virtual, noncorporeal, in(tangible) counterparts. It is also not inconceivable that we will soon all be happily working in very bland, nondecorated identical office and home areas with utilitarian pieces of “real” furniture, on and around which our mixed reality systems superimpose our choice of interior decoration and customization, much like we do with our computer desktops. Why, for example, would you bother with the maintenance and upkeep of a real six-foot fish tank in your office if you could have a virtual one which is indistinguishable from the real thing while wearing your HoloLens? Of course, this new technology, and the social, societal, psychological effects that the use of such systems will have, will only be understood with time. If someone cannot see or (eventually) feel (haptic feedback is also developing extremely fast) the difference between something that is corporeal and something incorporeal, the whole limiting factor of this characteristic falls away. As the levels of immersion into virtual worlds increase to the singularity point at which the real and virtual are indistinguishable from each other, this requirement of corporeality is becoming less of an issue. However, as long as the doctrinal requirement of tangibility of things remains present in a legal system, virtual property should be classified as an intangible thing that is recognized as property as an exception to the rule. This will follow the precedent set by the recognition of other intangible things in, for example, South African and German law. The easiest and possibly the best way to create this exception will be through new legislation that deals specifically with this issue, since other exceptions to the corporeality principle have been recognized quite easily, if the property interest in question was created, recognized, or protected in legislation.

BARFIELD PRINT.indd 653

13/11/2018 11:53

654  Research handbook on the law of virtual and augmented reality 5.4  External to Avatars In the real world, the externality requirement’s only function is to exclude humans and human body parts from being objects. In a virtual world, a player’s avatar represents him or herself in the virtual world and although the avatar is regarded as the virtual equivalent of a living human being in the virtual world, elsewhere it is regarded as an object of a property right. This fact that an avatar can be both a subject and an object of a property right seems to be in direct contrast to the application of the real-world characteristic. However, due to the code-based, fantastical, and flexible nature of the content and rules in virtual worlds, there is no reason why this should present a difficulty for the recognition of a virtual thing, since the moral reasons for requiring externality of things in the real world are mostly absent in virtual worlds. For example, while trade in body parts and avatars is quite acceptable in virtual worlds, trade in avatars that depict children being tortured or sexually exploited falls into an extremely gray moral and ethical area. However, externality still features to a lesser extent in the virtual world. The avatar is closely linked to the virtual-world account and personifies the player’s account inside the virtual world. The avatar can be sold, altered, or destroyed, and in certain cases (such as Second Life) body parts are freely available for sale on the in-world market. Slaves could also be encountered in virtual worlds and can be the objects of property rights.109 One area in which many virtual worlds differ from the real world is where items appropriated by an avatar become “attuned” to the avatar. This means that the item will only be usable by that one specific avatar and never again by any other. The item can never be resold or transferred to another avatar, although in certain cases it can be sold to a nonplaying character (NPC)110 or destroyed if an avatar wants to get rid of it.111 In other words, even though it is external to an avatar, it becomes a personal part of the character.112 In essence, because avatars are not humans, the characteristic of externality is not a big issue for the recognition of virtual property as property in the real world.

109   These slaves will usually not be avatars of other real-world players. The slaves would normally be NPCs, which are provided by the developer to enhance the gameplay and populate the virtual world. However, anything is possible in a virtual world and some players do in fact choose to become virtual slaves. An example of this is encountered in one of the subcultures in Second Life where some players own harems of sex slaves that are populated by other players’ avatars who fulfil their fantasies by roleplaying as sex slaves. See “Sex in Second Life,” www.informationweek.com/ news/software/hosted/showArticle.jhtml?articleID=199701944 (Nov. 3, 2010). 110   When the item is sold to the NPC, the player would get a nominal amount for the item, but at the same time the game code would remove the specific item from general circulation. In this situation, the NPC just acts as an agent of the game code. A visual illustration of how an NPC can act as an agent of the game code is found in the movie The Matrix (1999), where the game code creates NPCs (all of them clones of one another and called Mr Smith or Agent Smith) to fight the breach in the code created by NEO’s conscious immersion into the virtual world, from the outside. 111   Similar to a tattoo in real life. 112   See http://everquest2.com/manual /EQII_Manual.pdf 17.

BARFIELD PRINT.indd 654

13/11/2018 11:53

Law and property in virtual worlds  655 5.5 Independence In the virtual world, independence is given by the code that separates and rebuilds the individual bits of code into recognizable and manageable entities. As soon as the data is transferred from the storage server and recreated into identifiable things on the screen, the independence of the virtual item is attained. This independent object is then imported into the virtual world either as a feature of the virtual landscape, or as an item lying around as res derelictae (belonging to no one and subsequently “up for grabs”), or being carried around by wild beasts, monsters and other NPCs. The idea behind most virtual objects is that they should at one stage or another be acquired by a player’s avatar as a possession. As in the real world, things like running water, crops, and minerals are not initially considered to be independent, but are designed to be separated by an avatar into manageable and independent entities, as would be the case with mining, harvesting, or bottling of water in both the virtual and real worlds. 5.6  Appropriability/Susceptibility to Avatar Control Virtual things are by their nature controlled by the computer, although the software enables the player to manipulate her avatar and virtual items in real time in the same way as would be possible in the real world. Once again, most virtual-world items are explicitly designed (from a code level) to be appropriable and susceptible to control by avatars. The only exclusions would be so-called communal areas or buildings that are provided as background scenery or essential infrastructure to the game, and are as such not appropriable by an avatar. Examples of this include town commons, developercontrolled taverns, municipal buildings, bridges, and road infrastructure, to name but a few. 5.7  Use and Value As with real-world things, a virtual thing should also be a thing that could be subjected to human control. In this case it would be indirectly subjected to human control and directly subjected to avatar control. It follows that virtual objects that are not susceptible to avatar or player control would not be regarded as having the characteristic of use and value (for an avatar). Examples in the virtual world mirror those in the real world and include objects such as the virtual sky, sea, and planets. However, if any of these objects can be separated into manageable units and therefore subjected to avatar control, they could be regarded as virtual things.113 Another example of objects that are not regarded as things, unlike in the real world, would be most noninteractive objects. Generally speaking, if an object in a virtual world is noninteractive, that means that it is not subject to avatar control and therefore will not be classified as a virtual thing. An example of this would be a building

113   An example of this will be a virtual world where one of the functions of the game is to take control of a whole planet in a galactic war. In that case, the player and her avatar could indeed exert control over the planet and make use of it. For the purposes of that virtual world, a planet would qualify as a thing.

BARFIELD PRINT.indd 655

13/11/2018 11:53

656  Research handbook on the law of virtual and augmented reality in a virtual town that was created and coded into the virtual world with the sole purpose of being a nonfunctional piece of scenery. The characteristic of being of value has to date been one of the most important characteristics that a court will take into consideration when deciding to afford a virtual property interest with protection as property, or not.114 Although virtual property normally has considerable economic value, sentimental value is also recognized in virtual property.115 In a virtual world, most things are either of use or value to an avatar. This value may be financial inside the game—that is, a sword can be sold or traded in the game for a reciprocal amount of gold—or outside, where it can be traded or sold for real-world money. The sword or other items can also be of value due to their function and usability inside the game. This means that the sword can be used by an avatar to complete quests and advance in the game. Therefore, it will be of both use and value to the player. Sometimes an item might not have any financial or use value, but rather would enhance the status of the player or have sentimental value purely because it is a trophy or indication of how far a player has advanced. For example, a level 70 mage might not have much use for carrying around “the staff of Gandalf,” except for the fact that no other player who is not a level 70 mage will be able to pick it up. Thus, the staff symbolizes the player’s status in the game without having any real use or value apart from sentimental value. 5.8 Conclusion One of the arguments for the recognition of virtual property is that if an item is tangible and corporeal inside the virtual world and is regarded by players as a tangible thing in the real world, it should be treated as property in the real world. According to the argument made above about the (in)tangibility of virtual things, it can be said that in the virtual world, a virtual thing seems to have many similar characteristics to those of real things in the real world. These are the characteristics of corporeality, of being external to an avatar, of an independent nature that is appropriable, and of being of use and value to an avatar.116 While this comparison showed that virtual things have many of the same characteristics as their real-world counterparts, it was discovered that the characteristics of corporeality and externality have to be applied differently in a virtual world. It was also shown that these differences do not detract from the argument that virtual things should be recognized as property in the real world. The characteristic of corporeality is not a problem because real world property law already recognizes certain exceptions to the requirement and the requirement of externality is not problematical because, although similar in virtual worlds, avatars are not humans. It should be borne in mind that even though the characteristics cannot be applied perfectly to a virtual world, the similarity in form and function between a thing in the real world and a thing in the virtual world tends to be clear. This can be illustrated by the example of a chair in the virtual world. From a code-based perspective, it is designed to be, and functions as if, it were tangible inside the virtual world. The avatar can move it

  See Erlank, Property in Virtual Worlds at 3.5.  See id at 3.5.5. 116   Or to a player via his or her avatar. 114 115

BARFIELD PRINT.indd 656

13/11/2018 11:53

Law and property in virtual worlds  657 around, fall over it and sit on it. It follows the form and function of its real-world counterpart, since it is of use and value to the avatar. It is independent of the avatar in that it can be left behind and be separated from the control and possession of the avatar. If the avatar ceases to exist or leaves the virtual world, the chair still remains behind in the virtual world and does not disappear. It is appropriable and susceptible to human control on two levels. From the real-world perspective, the chair is controllable and appropriable by a real person through the direct manipulation of an avatar. From the perspective of being inside the virtual world, the chair can be sold, traded, stolen, or destroyed by an avatar or other avatars. Last, it is of use and value because it follows the form and function of a chair in the real world that is also of use and value. Apart from being useful and valuable inside the virtual world, the thing is usually also commercially valuable both inside and outside the virtual world. In conclusion, although the subjects of property differ in the real and virtual worlds, the focus here is rather on the objects of property than the subjects. This makes the discrepancies between the characteristics of virtual things and real-world things much less of a problem.

6.  CASE LAW AND COUNTRY-SPECIFIC EXAMPLES117 Since virtual property has value, both economic and sentimental, I will now discuss the issues surrounding the protection or, to state it alternatively, the theft of virtual property. Many authors argue that this is one of the major reasons why players should get recognized property rights in their virtual property.118 In other words, virtual property owners should (in certain cases) be able to protect their property. While there are numerous critics who sharply object to acknowledging property rights in virtual items, this is—­interestingly enough—the area in which some courts have found it intuitively logical to accept that property rights do exist. Even though the courts usually do not explicitly state this, it can be inferred from the judgments of the few cases that have been heard so far. These cases are briefly discussed in what follows. 6.1 China It would seem as if the Asian legal systems have been quick on the uptake with regard to the concept of protecting virtual property. In fact, to date, most cases that deal explicitly with virtual property have come from China. For example, Beijing’s Chaoyang District People’s Court in China recently ordered that a virtual-world developer had to return virtual property to a player whose account had been hacked.119 The hackers gained access to his account and essentially looted it of all valuable assets.120 The gamer in question 117   For the purposes of this chapter I will not fully discuss all the cases. They are dealt with in more detail in Erlank, Property in Virtual Worlds 90–98. 118  Barfield, Virtual Reality 38. 119   Li Hong Chen v Beijing Arctic Ice Technology Development Company (2003, Beijing Second Intermediate Court). 120  Barfield, Virtual Reality 38.

BARFIELD PRINT.indd 657

13/11/2018 11:53

658  Research handbook on the law of virtual and augmented reality had spent two years of his time and approximately 10,000 Yuan (which translates to roughly $1200) on pay as you go cards that enabled him to participate in the game. In the process of playing, he collected weapons and managed to level his character in the game Honyue,121 also known as Red Moon. The court found the developers liable for the loss suffered by the player,122 which was due to security loopholes in their software that enabled the hackers to steal the player’s property.123 6.2  South Korea In South Korea,124 more than 40 percent of teenagers spend time in online virtual worlds, and the country has the largest adoption of broadband internet per capita.125 There have been a huge number of South Korean cases dealing with virtual property. Most of these cases concern incidents in which one player forced another to transfer virtual property to his avatar in the virtual world. As a result, more than 22,000 cybercrime reports relating to virtual property were reported to police in 2003 and 10,187 teenagers were arrested for virtual property theft in a single year.126 6.3 US In Bragg v. Linden Research,127 the complainant was denied use of his virtual-world assets, his virtual property was removed, and his account was suspended due to making use of a glitch in the system to improperly benefit from auctions of real estate in Second Life.128 The parties settled out of court. In Evans et al v. Linden Research Inc et al,129 a class-action suit was launched against Linden Research involving virtual property in Second Life. This case was also settled outside of court, but the court did approve the settlement.130

121   Leveling refers to the increase in experience, status, wealth, and power that an avatar gets as it progresses through the virtual world. 122   This is an important aspect that developers of virtual worlds need to take note of in terms of corporate governance and control of virtual worlds. 123   See www.lexology.com/library/detail.aspx?g=0d26ced1-ebd8-4e32-bf83-061a831d64a6 where Ma discusses the five leading Chinese virtual property cases. 124   See Erlank, Property in Virtual Worlds 94. For a comprehensive discussion of South Korean cases and policy, see Ung-gi Yoon, Real Money Trading in MMORPG Items from a Legal and Policy Perspective, http://ssrn.com/abstract=1113327, 1–59 (2004). Yoon 2004 SSRN. 125  Fairfield, Virtual Property 1088. 126   Id. 127   Bragg v Linden Research Inc 487 FSupp 2d 593 (ED Pa 2007). See Erlank, Property in Virtual Worlds 84–5. 128  Barfield, Virtual Reality 38. 129   Evans et al v Linden Research Inc et al, Case No 4:11-cv-1078-DMR. 130   For comprehensive details about the class action suit and subsequent settlement see www. secondlifesettlement.com/hc/en-us.

BARFIELD PRINT.indd 658

13/11/2018 11:53

Law and property in virtual worlds  659 6.4 Australia The Australian example is not a case and the affected parties never appeared in court. It is, however, a good example of why virtual property should be protected.131 An Australian businessman, who was one of the most trusted and best known players in the virtual world of EvE Online, was the manager in charge of a bank in the virtual world. Because of his trusted status, other players were willing to deposit their earnings and profits measured in gold with his bank. The bank manager then succumbed to the effects of the credit crisis and found himself in dire financial straits in the real world. He stole all of the bank customers’ money and converted it to Australian dollars, then proceeded to use his ill-gotten gains to pay off his mortgage. The only aftereffect of this appears to have been a ban from the game. In China, he would have been convicted of theft. 6.5  The Netherlands A case dealing with the theft of virtual property was recently decided in the Netherlands.132 In a manner of speaking, the Dutch courts accepted that there must have been a protectable property right which was infringed upon, otherwise they would not have convicted the perpetrators of theft.133 I will deal only briefly with this case here. The case was heard in Leeuwarden, by a full bench criminal court of the first instance. In essence, the facts are as follows. Two Dutch teens were convicted of stealing virtual game items from another teen. All three of the boys played the game Runescape, which is a large virtual world and has millions of players. What is especially interesting about this case is that the boys not only stole the virtual items, but did it by using extreme force. They beat the other boy up and even threatened to kill him with a knife. The fact of the matter is that it would have been very easy for the court just to convict the two boys of assault and leave the matter at that. The court, however, chose to deal with the facts as argued by the prosecutor: that the boys stole virtual property. After analyzing the facts of the case and the applicable sections of the legislation (Artikel 310 of the Wetboek van Strafrecht), the court found that the stolen property is able to fulfill all the requirements of being classified as goed or property as it is defined in the legislation.134 Some of the criteria which the court found to be relevant were the following. According to the court, before something can be classified as goed under the Artikel, it has to have value for the possessor thereof. The value does not need to be expressible in monetary terms and as such I infer that it would include items with purely sentimental value. The court referred to the fact that virtual worlds have become a huge phenomenon and that players attach a lot of value to their virtuele goederen. The court even mentioned that among the places where these virtual items are sold and traded is the schoolyard. It then found that the virtual property had value for both the complainant and the accused. The court explicitly stated that the items did not need to be physical items (stoffelijke voorwerpen) and that it could be equated to electricity  www.reuters.com/article/2009/07/02/us-gaming-theft-tech-life-idUSTRE5613SL20090702.   LJN: BG0939, Rechtbank Leeuwarden, 17/676123-07 VEV. 133   The court even took judicial notice of the details of the virtual world! 134   Please note that the meaning of “goed” is defined in statute and includes both material as well as immaterial things under that definition. 131 132

BARFIELD PRINT.indd 659

13/11/2018 11:53

660  Research handbook on the law of virtual and augmented reality and money held in an account (giraal geld). Another important aspect is that there has to be transfer of factual (not physical) control from the accuser to the accused. The sentence was one of community service for 160 hours which had to be finished within 12 months. If this was not done properly, 80 days of confinement in a youth detention facility was mandated. The important thing to take from this case is that a Western court was willing to recognize, for the first time, first that there is such a thing as virtual property, and second that it is important enough to protect. It is indeed a victory for virtual property that such a judgment has at last emerged from a Western legal system. The case went to appeal,135 and the judgment of the court of first instance was confirmed by the Dutch Supreme Court (Hoge Raad) except in relation to the number of hours of community service, which was reduced. One last thing to note about the appeal is that the court chose to specifically highlight and address an argument raised by the appellant that since the game makes provision for theft inside the game, the court should accept that theft of the virtual property in the real world should be allowed and not be deemed a crime. The court dealt with this issue firmly, noting that even if the virtual world allows for theft as part of the gameplay, this does not carry over to infringements outside the virtual world. In other words, theft of virtual property in the real world is not part of the gameplay inside the virtual world, and therefore the argument posed cannot be accepted as a defense.136 I must note here that while South Africa sometimes lags in creating legislation that deals with new technological developments, it is interesting that provision is made in the recent Cybercrimes and Cybersecurity Bill to achieve the same result as in the Dutch case just discussed, but with discrete legislation.137 In s. 13, the bill states the following: Theft of incorporeal 13. The common law offence of theft must be interpreted so as not to exclude the theft of an incorporeal.

If this section survives to the legislative stage, then virtual property would by default be protected as property in South Africa, at the very least in terms of criminal law.

7. CONCLUSION In conclusion, it should be clear that the development of virtual worlds and virtual property go hand in hand. As the virtual extends into the real world and everyday life by means of the already ubiquitous computer, gaming, virtual, and augmented reality systems, and especially with the advent of high performing mixed reality systems, the law will no longer concern itself with questions such as whether virtual objects are property for real-world purposes, but rather will simply accept this as the case. As the line between the real and the virtual continues to blur and the level of immersion increases, we will in   LJN: BQ9251, Hoge Raad, CPG 10/00101 J.   For a discussion on how this relates to the acceptance and protection of virtual property by courts, and the rule that courts will have to adjudicate each and every case concerning virtual property on a case-by-case basis—see Erlank, Property in Virtual Worlds 401–7. 137   Cybercrimes and Cybersecurity Bill B6-2017 (South Africa). 135 136

BARFIELD PRINT.indd 660

13/11/2018 11:53

Law and property in virtual worlds  661 all probability see the law not concerning itself with whether an object is real or virtual, but just accepting that the object exists. There are, of course, a plethora of other aspects of virtual property and property in the virtual world that can be discussed, but I hope that this chapter provided the reader with some small understanding of the intricate interplay in the law concerning the real and virtual counterparts of property.

SOURCES Bragg v Linden Research, Inc 487 FSupp 2d 593 (E.D. Pa. 2007). Dorel v Arel 60 F. Supp. 2d 558 (E.D. Va. 1999).

BARFIELD PRINT.indd 661

13/11/2018 11:53

20.  Legal jurisdiction and the deterritorialization of social life Paul Schiff Berman*

Social lives are increasingly unmoored from physical location. This deterritorialization arises in part from successive waves of technological innovation that have repeatedly transformed human conceptions of space, place, and proximity. In the nineteenth and twentieth centuries, new developments in rail, automobile, and airplane travel shrank the sense of physical distance. Communications technologies such as the telegraph, the telephone, and the internet allowed data to move across territorial boundaries with increasing ease. And twenty-firstcentury developments in social media, virtual worlds, augmented reality, electronic financial transactions, drones, robotics, and artificial intelligence allow human beings to interact in more and more robust ways at a physical remove from their location. Meanwhile, the ubiquity of multinational corporations, global supply chains, and cloud-based data means that our lives are more likely to be affected by activity that is spatially distant. As a thought experiment, one can imagine an “effects map,” in which one identifies a territorial locality and plots on a map every action that has an effect on that locality.1 Five hundred years ago, such effects would almost surely have been clustered around the territory, with perhaps some additional effects located in a particular distant imperial location. A hundred and fifty years ago, those effects might have begun spreading out. But today, while locality is surely not irrelevant, the effects would likely be diffused over many corporate, governmental, technological, and migratory centers. The deterritorialization of effects is felt on an everyday, intuitive level. For example, Anthony Giddens noticed as far back as 1990 that the “local” shopping mall is not experienced as truly local at all; nearly “everyone who shops there is aware that most of the shops are chain stores,” identical to stores elsewhere, and that the mall itself closely resembles innumerable other malls around the globe.2 Thus, while experiencing a “local” place, we recognize the absent forces that structure our experience. Such forces include the steady decline in local ownership of public spaces, which can itself be linked to the globalization of capital. And now, almost thirty years after Giddens wrote, even the mall is too territorial; shopping is increasingly moving wholly online. We may also feel the growing significance of these “remote” forces on our lives, whether those forces are multinational corporations, world capital markets, or distant *  Some material in this chapter is derived from Paul Schiff Berman, Global Legal Pluralism: A Jurisprudence of Law Beyond Borders (2012); Paul Schiff Berman, The Globalization of Jurisdiction, 151 U. Pa. L. Rev. 311 (2002); and Paul Schiff Berman, Introduction in Law & Society Approaches to Cyberspace (Paul Schiff Berman ed., 2007). Special thanks to Jiyoon Moon for helpful research assistance. 1   This thought experiment is articulated in David G. Post, Against “Against Cyberanarchy,” 17 Berkeley Tech. L.J. 1365, 1371–3 (2002). 2   Anthony Giddens, The Consequences of Modernity 140–41 (1990).

662

BARFIELD PRINT.indd 662

13/11/2018 11:53

Legal jurisdiction and the deterritorialization of social life  663 bureaucracies. As John Tomlinson has observed: “People probably come to include distant events and processes more routinely in their perceptions of what is significant for their own personal lives. This is one aspect of what deterritorialization may involve: the ever-broadening horizon of relevance in people’s routine experience.”3 And, of course, those with less power to influence the processes of globalization—those forced to cross borders for work, those bankrupted through global competition, those affected by environmental degradation, and many others—experience this deterritorialization in even more insidious ways. Finally, though the effects may often seem to flow in only one direction (industrialized centers affecting nonindustrialized peripheries), one need only consider migratory labor patterns and immigration; the rapid spread of diseases such as Ebola, the Zika virus, or avian flu; and the bidirectional impact of labor outsourcing to realize that deterritorialization affects us all. These various trends have a significant impact on law for two reasons. First, law is a social product, and therefore changing social realities of any sort create the context for legal change. Second, and more specific to these deterritorialization trends, law has historically been tied to physical territory. Indeed, the core idea of legal jurisdiction is often premised on territorial location and the delineated borders that are thought to separate distinct governmental entities and legal regimes. Accordingly, any set of social changes that alters both the perception and the reality of space, distance, location, territory, travel, and effects is bound to put pressure on legal regimes, creating tensions, uncertainties, and transformations. This chapter provides a broad overview of, and introduction to, some of those tensions, uncertainties, and transformations, as they relate primarily to rules of legal jurisdiction and choice of law. First, I suggest that jurisdictional rules always inevitably both reflect and construct social conceptions of space, and I summarize some of the ways in which technological and social changes in the nineteenth and twentieth centuries prompted changes in jurisdiction and choice-of-law rules. Second, I look at new legal jurisdictional questions emerging in the twenty-first century, as social media, virtual worlds, cloudbased data, electronic currencies, and automation disrupt settled legal doctrines once again. Finally, I look at law’s possible response to these new social changes. Some scholars have suggested that the disruptions require us to invent entirely new legal paradigms. Others respond that existing legal rules are readily adaptable to the new environment. Nevertheless, it seems clear that even without completely new paradigms, legal conceptions of jurisdiction and choice of law will, over time, inevitably be dramatically altered. This is because in the end it is difficult to imagine any large-scale societal change that does not create large-scale legal change as well.

1.  LEGAL JURISDICTION AND SOCIAL CHANGE Most debates about legal jurisdiction and choice of law tend to revolve around either political theory questions about when a judicial or administrative exercise of authority is legitimate, or legal policy questions about the most efficient or effective system for solving   John Tomlinson, Globalization and Culture 115 (1999).

3

BARFIELD PRINT.indd 663

13/11/2018 11:53

664  Research handbook on the law of virtual and augmented reality specific legal dilemmas. There is more to the assertion of jurisdiction or the extraterritorial imposition of norms, however, than simply questions of political legitimacy or efficient dispute resolution. The assertion of jurisdiction, like all legal acts, can also be viewed as a meaning-producing cultural product. What does it mean, after all, to say that some person, corporation, or activity is subject to a community’s jurisdiction? And how does the idea of jurisdiction relate to conceptions of geographic space, community membership, citizenship, boundaries, and self-definition? Although largely ignored in debates over jurisdiction and new technologies, these foundational issues must be considered seriously if we are to develop a richer descriptive account of the role of legal jurisdiction in a deterritorialized world of interaction.4 It has become commonplace for cultural critics and others to identify the ways in which social structures shape and constrain conduct, yet the link between social structures and physical spaces has received less attention.5 Nevertheless, “[t]he production of space and place is both the medium and the outcome of human agency and social relations.”6 This cultural construction of space includes the boundaries drawn between “public” and “private” spaces; the decisions a community makes about land use and zoning; the appropriation and transformation of “nature” as both a concept and a physical description; the local autonomy of governmental units; the use of specialized locations for the conduct of economic, cultural, and social practices; the creation of patterns of movement within a community; and “the formation of symbolically laden, meaning-filled, ideologyprojecting sites and areas.”7 In addition, topological space, which consists of the formal boundary lines we have chosen, is distinctively different from social space, which includes the meanings given to space (both local and nonlocal), to the distances between delineated spaces, and to the time necessary to traverse those distances.8 For example, a hundred-mile automobile trip may seem like a greater journey to residents of the northeastern United States, who are accustomed to relatively short distances between destinations, than to residents of the west, where cities and towns are more dispersed. Similarly, a thousand-mile trip carries a very different social meaning today, in the age of relatively inexpensive air travel, than it did one hundred years ago, even if the topological space remains the same.9 And 4   A broader discussion of these issues can be found in Berman, The Globalization of Jurisdiction, supra note *. 5   For two notable exceptions within legal scholarship, see Terry S. Kogan, Geography and Due Process: The Social Meaning of Adjudicative Jurisdiction, 22 Rutgers L.J. 627 (1991); Richard T. Ford, Law’s Territory (A History of Jurisdiction), 97 Mich. L. Rev. 843 (1999). Kogan’s work specifically addressed the social significance of adjudicative jurisdiction and so is particularly relevant here. My discussion in this subsection is heavily indebted to Kogan’s argument. 6   Allan Pred, Making Histories and Constructing Human Geographies 10 (1990). 7   Id. 8  Kogan, supra note 5, at 634. 9   John Tomlinson describes this shift as follows:

In a globalized world, people in Spain really do continue to be 5,500 miles away from people in Mexico, separated, just as the Spanish conquistadors were in the sixteenth century, by a huge, inhospitable and perilous tract of ocean. What connectivity means is that we now experience this distance in different ways. We think of such distant places as routinely accessible, either representationally through communications technology or the mass media, or physically, through the

BARFIELD PRINT.indd 664

13/11/2018 11:53

Legal jurisdiction and the deterritorialization of social life  665 of course America’s well-documented postwar demographic shift from city to suburb is not merely a change of topology, but a politically and symbolically significant cultural transformation.10 Moreover, the construction of legal spaces and the delineation of boundaries is always embedded in broader social and political processes.11 “Legal categories are used to construct and differentiate material spaces which, in turn, acquire a legal potency that has a direct bearing on those using and traversing such spaces.”12 For example, in the history of European conquest of Australia, the naming of particular spaces—rivers, mountains, capes, bays, and so on—became a central point of political contest.13 The Europeans believed that the Aboriginals did not classify or name the landscape and transformed that purported “spatial deficiency” into a “legal deficiency”: If the Aboriginals did not name their places, so the thinking went, their “grasp of it [was] so tenuous . . . [that] it was hardly a crime to take possession of it.”14 To take another example, Jeremy Waldron has observed that increasing restrictions on the use of public spaces for activities such as sleeping or washing deny homeless people any opportunity to perform those acts because there is neither a public nor a private space to do so.15 The social meaning of geographical space also includes the way in which an individual or community perceives those who are outside the community’s topological or social boundaries. As people develop attitudes of familiarity toward the spaces in which they reside and conduct their daily activities, they may also come to view unfamiliar people and locations as frighteningly alien. Alternatively, the outside “other” can be seen as inviting, friendly, and hospitable, or as mysterious, exotic, and romantic.16 There are a seemingly infinite variety of attitudes one may hold toward unfamiliar social spaces. expenditure of a relatively small amount of time (and, of course, of money) on a transatlantic flight. So Mexico City is no longer meaningfully 5,500 miles from Madrid: it is eleven hours’ flying time away. Tomlinson, supra note 3, at 4. 10   For the sociopolitical history of American suburbanization, see Joel Garreau, Edge City: Life on the New Frontier (1992); Kenneth T. Jackson, Crabgrass Frontier: The Suburbanization of the United States (1985). 11   See Nicholas K. Blomley, Law, Space, and the Geographies of Power, at xi (1994) (“The legal representation of space must be seen as constituted by—and in turn constitutive of—complex, normatively charged and often competing visions of social and political life under law”). 12   Id at 54. 13   See Paul Carter, The Road to Botany Bay: An Exploration of Landscape and History (1988) (describing European exploration and subsequent naming of various Australian geographical features). 14   Id at 64; see also Robert D. Sack, Human Territoriality: Its Theory and History 6–8 (1986) (describing similarly loose conceptions of territoriality among members of the Chippewa tribe at the time Europeans settled in the United States). 15   Jeremy Waldron, Homelessness and the Issue of Freedom, 39 UCLA L. Rev. 295, 315 (1991) (“Since private places and public places between them exhaust all the places that there are, there is nowhere that these actions [such as sleeping] may be performed by the homeless person”). 16   As Stuart Hall has described: “To be English is to know yourself in relation to the French, and the hot-blooded Mediterraneans, and the passionate, traumatized Russian soul. You go round the entire globe: when you know what everybody else is, then you are what they are not. Identity is always, in that sense, a structured representation which only achieves its positive through the narrow eye of the negative.” Stuart Hall, The Local and the Global: Globalization and Ethnicity, in Culture,

BARFIELD PRINT.indd 665

13/11/2018 11:53

666  Research handbook on the law of virtual and augmented reality Such attitudes are embedded in context and shaped and influenced by manifold factors including politics, socioeconomic relationships, and the extent of contact that one has with the “other.”17 Thus, jurisdictional rules have never simply emerged from a utilitarian calculus about the most efficient allocation of governing authority. Rather, the exercise of jurisdiction has always been part of the way in which societies demarcate space, delineate communities, and draw both physical and symbolic boundaries. Such boundaries do not exist as an intrinsic part of the physical world; they are a social construction. As a result, the choice of jurisdictional rules reflects the attitudes and perceptions members of a community hold toward their geography, the physical spaces in which they live, and the way in which they define the idea of community itself. In order to convey this basic idea, it might be useful to give an admittedly oversimplified, functionalist account of the change in American jurisdictional rules over time. In the nineteenth century, both jurisdictional and choice-of-law rules in the United States were grounded in the territorial power of the sovereign. Thus, each sovereign was deemed to have jurisdiction, exclusive of all other sovereigns, to bind persons and things present within its territorial boundaries. In 1877 the United States Supreme Court determined that this territorially based principle was an inherent part of constitutional due process itself. In Pennoyer v. Neff 18 the Court ruled that “[t]he authority of every tribunal is necessarily restricted by the territorial limits of the State in which it is established. Any attempt to exercise authority beyond those limits would be deemed in every other forum . . . an illegal assumption of power, and be resisted as mere abuse.” Thus, although “every State possesses exclusive jurisdiction and sovereignty over persons and property within its territory . . . no State can exercise direct jurisdiction and authority over persons and property without its territory.” The Pennoyer decision reflected a conception of jurisdiction based on territorial borders and pure power. The underlying message was: If a state can “tag” you, it can have power over you; if it can “tag” your property, it can have power over your property. Thus, the limits of judicial power under Pennoyer were rigidly defined by the boundaries of each state. Choice-of-law rules were similar. The prevailing “vested rights” theory of the era sought to locate all disputes spatially by identifying the physical location of the essential act that was deemed to constitute the cause of action. Applying the vested rights approach, judges looked to the place of the tort,19 or the place of contracting,20 or the location of the property at issue.21 Having “localized” the cause of action, only the state where the cause Globalization and the World-System: Contemporary Conditions for the Representation of Identity 19, 21 (Anthony D. King ed., 1997). 17  Kogan, supra note 5, at 637. 18   95 U.S. 714 (1877). 19   Restatement (First) of the Law of Conflict of Laws § 378 (“The law of the place of wrong determines whether a person has sustained a legal injury”). 20  2 Joseph Beale, A Treatise on the Conflict of Laws 1091 (1935) (“The question whether a contract is valid . . . can on general principles be determined by no other law than that which applies to the acts, that is, by the law of the place of contracting . . . If . . . the law of the place where the agreement is made annexes no legal obligation to it, there is no other law which has power to do so”). 21   See id at 938 (finding that immovable property, “being unable to be taken away from the state in which [it is located], must always in the last analysis be governed by the laws of that state”).

BARFIELD PRINT.indd 666

13/11/2018 11:53

Legal jurisdiction and the deterritorialization of social life  667 of action “vested” could apply its law to the dispute.22 Such localization effectively avoided any consideration of multiple norms or the relationship between the familiar and the strange because it placed all disputes within a familiar, territorially delineated framework. These territorially based jurisdictional and choice-of-law principles derive in part from a particular understanding of social space in the United States at that time. As historian Robert Wiebe has famously observed, “America during the nineteenth century was a society of island communities.”23 With weak communication and limited interaction, these “islands” felt widely dispersed, and it is not surprising that local autonomy became “[t]he heart of American democracy.”24 Even though France had long since developed a centralized public administration, Wiebe argues that Americans still could not even conceive of a distant managerial government. In such a climate, geographical loyalties tended to inhibit connections with a whole society. “Partisanship . . . grew out of lives narrowly circumscribed by a community or neighborhood. For those who considered the next town or the next city block alien territory, such refined, deeply felt loyalties served both as a defense against outsiders and as a means of identification within.”25 As the nineteenth century progressed, so this story goes, massive socioeconomic changes brought an onslaught of seemingly “alien” presences into these island communities. Immigrants were the most obvious group of outsiders, but perhaps just as frightening was the emergence of powerful distant forces such as insurance companies, major manufacturers, railroads, and the national government itself. Significantly, these threats appear to have been conceived largely in spatial terms. According to Wiebe, Americans responded by reaffirming community self-determination and preserving old ways and values from “outside” invasion.26 Given such a social context, it is not surprising that the jurisdictional rules of the period emphasized state territorial boundaries. Indeed, it is likely that the burdens of litigating in another state far exceeded simply the time and expense of travel, substantial as those burdens were. Just as important was the psychic burden of being forced to defend oneself in a foreign state, which may have felt little different from the idea of defending oneself in a foreign country. An 1874 Pennsylvania state court decision issued shortly before Pennoyer illustrates the extent of this psychic burden.27 In the case, a resident of New York had contested jurisdiction in Pennsylvania. The court acknowledged that the Pennsylvania courthouse was only “a few hours’ travel by railroad” from New York, but nevertheless ruled that the defendant could not be sued personally, in part because “nothing can be more unjust than to drag a man thousands of miles, perhaps from a distant state, and in effect compel him to appear.”28 The court disregarded the relatively slight literal burden in the case at hand, and instead focused on the specter of being “dragged” to a “distant

22  1 Beale, supra note 20, at 64 (“[T]he chief task of the Conflict of Laws [is] to determine the place where a right arose and the law that created it . . . ”). 23   Robert H. Wiebe, The Search for Order: 1877–1920, at xiii (1967). 24   Id. 25   Id at 27. 26   Id at 52–8. For a fictional account of this period that gives texture to this description, see Willa Cather, My Antonia (Houghton Mifflin Co 1926). 27   Coleman’s Appeal, 75 Pa. 441 (1874). 28   Id at 457 (1874).

BARFIELD PRINT.indd 667

13/11/2018 11:53

668  Research handbook on the law of virtual and augmented reality state” located “thousands of miles” away. Indeed, the decision seemed to equate other states with foreign countries, referring to a “defendant living in a remote state or foreign country . . . [who] becomes subject to the jurisdiction of this, to him, foreign tribunal.”29 These passages indicate that the psychic significance of defending oneself in another state was at least as important as the literal difficulties of travel. Both the literal and psychic burdens associated with out-of-state litigation changed as a result of the urban industrial revolution at the turn of the twentieth century, a revolution that profoundly altered American social space. Increasingly, economic and governmental activities were administered from afar by impersonal managers at centralized locations. In such a world, another state was likely to be viewed less as a foreign country and more as yet another distant power center, just one of many “anonymous, bureaucratic, regulatory bodies in an increasingly complex society.”30 In addition, advances in transportation and communications helped to weaken territoriality as the central category in which Americans understood their space. “As long as daily lives were focused to a large extent on the local, a state boundary symbolized the edge of the world and everything outside that boundary was alien and foreign.”31 With increased mobility, however, Americans regularly crossed state boundaries by train, by car, and by airplane, which inevitably diminished the sense that other places were alien. The rise of radio and television meant that events in other states could become a regular part of one’s daily consciousness. “Physical distance as a social barrier began to be bypassed through the shortening of communication ‘distance.’”32 These communication and transportation advances reinforced the functional interdependence that characterized the United States throughout the twentieth century. As a result, almost all of us are now regularly affected by people, institutions, and events located far away. In this altered social space, the call to defend a lawsuit in the courts of another state remained an imposition, but the burdens were no longer perceived in stark territorial terms. In other words, though many economic and practical burdens remained, the psychic burden was no longer as strong. Given such changes, it is not surprising that courts considering questions of legal jurisdiction began to chafe against the strict territorialist vision that Pennoyer imposed. In particular, the invention of the automobile and the development of the modern corporation meant that faraway entities could inflict damage within a state without actually being present there at the time of a lawsuit. In response, a number of states enacted statutes based on a theory of “consent” to jurisdiction, a theory that the Pennoyer Court had recognized as valid. For example, a Massachusetts statute decreed that an out-of-state motorist using the state’s highways would be deemed to have consented to Massachusetts jurisdiction in actions arising from accidents on those highways. The United States Supreme Court ultimately upheld this rather strained notion of consent in 1927. Similarly, most states 29   Id. Indeed, for juridical purposes, other states had, since the founding, been treated much like foreign countries, even for some time after the Civil War. 30  Kogan, supra note 5, at 651 (citations omitted). 31   Id at 652. 32   Joshua Meyrowitz, No Sense of Place: The Impact of Electronic Media on Social Behavior 116 (1985).

BARFIELD PRINT.indd 668

13/11/2018 11:53

Legal jurisdiction and the deterritorialization of social life  669 enacted statutes essentially requiring out-of-state corporations to agree to jurisdiction within the state as a condition of conducting business there. Moreover, even if the corporation did not explicitly agree, it was often viewed as having implicitly consented to state jurisdiction simply by transacting business in the state. Alternatively, courts sometimes ruled that corporations were physically “present” in any state in which that corporation was “doing business,” making the corporations subject to jurisdiction regardless of their consent. While allowing more flexibility in the jurisdictional calculus, the concepts of “consent” and “presence” were analytically unsatisfying in large part because corporations have neither individual wills nor physical reality. Thus, “consent” to jurisdiction and “presence” in a state were no more than legal fictions invented to cope with changing economic reality. In the 1945 case of International Shoe Co. v. Washington,33 the Supreme Court rearranged the landscape of personal jurisdiction. International Shoe, a Delaware corporation with its principal place of business in Missouri, manufactured and sold shoes and other footwear. The corporation maintained no offices, factories, or stores in the state of Washington, but did employ traveling salespeople there. These salespeople took orders and sent them to St Louis, whereupon International Shoe shipped the merchandise directly from St Louis to the customers in Washington. The state of Washington initiated an administrative hearing to assess International Shoe for contributions to the state unemployment compensation fund, but International Shoe argued that Washington had no jurisdiction over it because International Shoe was not “present” within the state. The Supreme Court rejected International Shoe’s arguments. Although the Court might simply have stated that the corporation’s contacts with Washington were sufficient to create corporate presence there, it chose instead to articulate an entirely new test for personal jurisdiction. This test replaced the strict territorial rules of Pennoyer with a more flexible due process inquiry based on whether the defendant had sufficient contact with the relevant state “such that jurisdiction is consistent with traditional notions of fair play and substantial justice.” This “minimum contacts” test would be satisfied as long as the “quality and nature of the activity” of the defendant within the state was sufficient “in relation to the orderly administration of the laws that it was the purpose of the due process clause to protect.”  Since the decision in International Shoe, the minimum contacts test has provided the framework for determining the outer limits of personal jurisdiction under the United States Constitution. Nevertheless, although the test’s flexibility is its greatest strength, such flexibility has meant that the minimum contacts analysis does not provide a clearly defined rule, relying instead on a highly particularized, fact-specific inquiry. Accordingly, one can never be sure in advance how many and what sort of contacts will be enough for a state to exercise jurisdiction under the federal Constitution. The Supreme Court has variously looked to whether defendants have “purposely availed” themselves of the state, whether it was “reasonably foreseeable” that they could be sued there, whether the defendant “targeted” the state, or whether the interests of the state in adjudicating a dispute outweighed the defendant’s concerns about increased cost, inconvenience, or potential bias. As transportation and interstate commerce have continued to grow in the decades   326 U.S. 310 (1945).

33

BARFIELD PRINT.indd 669

13/11/2018 11:53

670  Research handbook on the law of virtual and augmented reality since 1945, the Supreme Court has been called upon many times to determine how far to expand the reach of personal jurisdiction. Similarly, choice-of-law doctrine experienced a “revolution” in the twentieth century, with many states moving away from a strict territorialist conception based on where a lawsuit “arises.” And, not surprisingly, automobile, train travel, and cross-border contracts helped usher in the change, because it became increasingly difficult to justify why the territorial location of only one event in a complicated set of transactions should determine the law to be applied. For example, in the oft-discussed case of Alabama Great Southern Railroad v. Carroll,34 an employee was injured when a train coupling failed. The vested rights theory held that the relevant law was of the state where the injury occurred.35 However, given that the negligence took place elsewhere and neither the injured worker nor the corporation was based in the state where the injury occurred, the choice of the place of accident seems arbitrary. Similarly, it may be difficult to determine when and where an injury takes place (for example in a toxic tort case), and again the state where the person happens to be when the first symptoms appear seems relatively unimportant.36 Yet, under the vested rights approach such a determination is the only relevant factor in choice-of-law analysis. Thus, even if one agreed that territorialism should be the criterion for choice of law, it is not at all clear how best to choose the appropriate territorial nexus, and the mechanical vested rights rules seem substanceless. Moreover, the focus on territoriality in choice of law, as with jurisdiction, ignores important nonterritorial factors, such as community affiliation. For example, in a pure territorialist system, two parties who are domiciled in one state could avoid a local contract rule by crossing the border, entering into the contract, and returning, thereby taking advantage of the foreign law.37 This is not at all hypothetical in the modern world of offshore regulatory havens, where corporations frequently attempt to avoid various forms of governmental control by changing their territorial location. A purely territorial approach, therefore, will have difficulty coping with a modern world where transactions and transportation across borders are so frequent. Indeed, although vested rights analysis treats as transcendental truth the idea that the state has complete and unchallenged authority within its own sphere and that no state, therefore, may meddle in another state’s “local” affairs, we live today in a far more fluid world, where a purely territorial choice-of-law rule is unlikely to be satisfying. For example, when two New Yorkers drive an automobile from New York to Canada and get into an accident involving no one else, it seems to make little sense to apply Canadian insurance rules to any resulting claims just because the accident territorially occurred in Canada.38 Instead, we need to acknowledge that people may have multiple community affiliations beyond simply their territorial location at a particular moment in time, and therefore it might be appropriate for nonlocal law to apply to their transactions, at least under some circumstances. Thus, new conflicts approaches were adopted, focusing on a comparison of governmental interests or attempting to determine   11 So. 803 (1892).   See Restatement (First) of Conflicts of Law § 386. 36   See, e.g., Lea Brilmayer, Conflict of Laws 20–22 (1995) (discussing such difficulties). 37   This scenario is discussed in Brainerd Currie, Selected Essays on the Conflict of Laws 88 (1963). 38   See Babcock v. Jackson, 191 N.E.2d 279 (N.Y. 1963). 34 35

BARFIELD PRINT.indd 670

13/11/2018 11:53

Legal jurisdiction and the deterritorialization of social life  671 which government had the most significant relationship to the dispute,39 rather than using an approach relying on strict territorial location. As we can see from even this brief overview, in both jurisdiction and choice of law, a more flexible set of tests based on fairness and context has at least in part replaced rigid, though more predictable, tests based exclusively on territorial location. And while this is an oversimplified account of the shift in American jurisdictional and choice-of-law rules, for the purposes of this discussion it makes the essential point clearly enough: Changes in political and social conceptions of space form at least part of the context for changes in jurisdictional understandings. Thus, although some might ask why the rise of the internet, social media, virtual worlds, augmented reality, cloud-based data, e-currency, and autonomous agents requires us to rethink our ideas about legal jurisdiction and choice of law, the reality is that jurisdictional rules are always evolving, and this evolution has always responded to changing social constructions of space, distance, and community. As a result, the relevant question is whether the sense of social space has shifted once again. Arguably, people around the world now share economic space to a greater degree than ever before, in large part because of the increase in online interaction. Modern electronic communications, recordkeeping, and trading capacities have allowed the world financial markets to become so powerful that the actions of individual territorial governments often appear to be ineffectual by comparison. Essential services, such as computer programming, can easily be “shipped” across nation-state boundaries and can even be produced multinationally. The international production and distribution of merchandise means that communities around the country—and even around the world—increasingly purchase the same name-brand goods and shop at the same stores. Virtual worlds and social media can ignore territoriality altogether and instead are organized around shared interests. An individual’s data can be stored literally anywhere regardless of the location of that individual. Currencies such as bitcoin and other blockchain systems defy individual territorial governmental efforts to localize them. Autonomous weapons allow people far removed from the battlefield to make targeted military strikes. All of these changes radically reshape the relationship of people to their geography. As Joshua Meyrowitz observed decades ago, electronic media create “a nearly total dissociation of physical place and social ‘place.’ When we communicate through telephone, radio, television, or computer, where we are physically no longer determines where and who we are socially.”40 Meyrowitz pointed out that, historically, communication and travel were synonymous, and it was not until the invention of the telegraph that text messages could move more quickly than a messenger could carry them. Thus, “informational differences between different places began to erode.” 41 Moreover, many of the boundaries that define social settings by including and excluding participants—including walls, doors, barbed wire, and other physical and legal barriers—are less significant in a world where “the once consonant relationship between access to information and access to places has been greatly weakened.”42

 See Restatement (Second) of Conflicts of Law.   Meyrowitz, supra note 32, at 115 (1985). 41   Id at 116 (describing the impact of telegraphic technology). 42   Id at 117. 39 40

BARFIELD PRINT.indd 671

13/11/2018 11:53

672  Research handbook on the law of virtual and augmented reality Given such changes, it is possible that the psychic burden of foreign jurisdiction is less significant today because of our increased contact with foreign places. On the other hand, as suggested by both the 2016 presidential election in the United States and the election in the United Kingdom regarding disengaging from the European Union, we may feel the need to cling even more tenaciously to localism in the face of the encroaching global economic system.43 Moreover, in either scenario the “we” is problematic. After all, different social groups, and different individuals, have very different degrees of exposure to, and control over, global flows of information, capital, and human migration.44 Nevertheless, the important point is that if jurisdictional rules both reflect and construct social space, further investigation is needed in order to better comprehend the relationship between community affiliation, physical location, and personal identity in a world where the importance of territorial borders and of geographical distance is being challenged.

2.  SOME TWENTY-FIRST-CENTURY CHALLENGES 2.1  Virtual Worlds Where is a virtual world for jurisdictional purposes? The answer depends in part on your perspective.45 From within the world itself, it feels like one is somewhere else. Think of the spatial set of metaphors. This is a “world.” One “goes to it” or “enters it.” From this perspective, a virtual world is a new community of individuals unbound by their literal physical location, “meeting” and forming a set of community bonds that could be stronger than the bonds they feel with their physically oriented communities. On the other hand, for one external to the world watching the participants, there is no world at all, just a large number of people located somewhere in physical space, looking at computer screens. Regardless of what sort of interactions they are having online, from this perspective these people are always bound to hundreds of separate territorially based communities and therefore subject to the legal jurisdiction of those communities.

43   Cf Anthony Giddens, The Consequences of Modernity 65 (1990) (“The development of globalised social relations probably serves to diminish some aspects of nationalist feeling linked to nation-states (or some states) but may be causally involved with the intensifying of more localised nationalist sentiments”). 44   Doreen Massey refers to this as the “power geometry of time–space compression.” Doreen Massey, Space, Place, and Gender 149 (1994). She contrasts those who are “in charge” of timespace compression—“the jet-setters, the ones sending and receiving the faxes and the e-mail, holding the international conference calls  .  .  . distributing the films, controlling the news, organizing the investments”—with those who do a lot of physical moving, but are not “in charge” of the process in the same way. Id. These people include those such as undocumented migrant workers who cross borders illegally or those who lose their jobs to less expensive labor abroad, or those whose livelihood is affected by global currency fluctuations. Thus, social conceptions of space, distance, and community definition are, of course, themselves varied and contested. 45   See James Grimmelmann, Virtual Borders: The Interdependence of Real and Virtual Worlds, 11 First Monday 3 (2006), https://perma.cc/JC6X-P6KB; Orin Kerr, The Problem of Perspective in Internet Law, 91 Georgetown Law Journal 357 (2003).

BARFIELD PRINT.indd 672

13/11/2018 11:53

Legal jurisdiction and the deterritorialization of social life  673 Given the relationship between the social experience of community and the legal definition of community, this difference of perspective matters. If one sees the relevant community affiliation as the in-game community, then the law of that community is established through the software code built into the world by the designer, the rules of behavior dictated by the game itself, the end-user license agreement (or EULA), and the social norms of that virtual world. In contrast, if the relevant community affiliation is the territorial location of its users, then a slew of national laws governing speech, depictions of violence, gambling, sexualized imagery, and so on might apply. Or, one could try to divide up the jurisdictional pie. Virtual-world designers and distributors could be subject to the law of their corporate headquarters, while users could be subject to the jurisdiction of their home countries. However, experience shows that governments are unlikely to be satisfied going after only end-users. The larger corporate actor profiting from the activity makes for a much more palatable (and easier to find) target. For example, consider the well-known early cyberlaw case involving Yahoo! in France. French users were downloading Holocaust denial material and Nazi memorabilia that were available on Yahoo!’s auction sites. Undeniably, the French end-users were violating French hate speech law and could be prosecuted. But those individuals were difficult to find, and prosecuting a few might well be futile because more would likely take their places. Thus, it is not surprising that the French plaintiffs focused on Yahoo! itself. On May 22, 2000, the Tribunal de Grande Instance de Paris issued a preliminary injunction against Yahoo.com, ordering the site to take all possible measures to dissuade and prevent access in France of Yahoo! auction sites that sell Nazi memorabilia or other items that are sympathetic to Nazism or constitute holocaust denial.46 Because selling such merchandise in France would violate French law,47 Yahoo.fr, Yahoo!’s French subsidiary, complied with requests that access to such sites be blocked.48 What made this action noteworthy, however, was the fact that the suit was brought not only against Yahoo.fr but against Yahoo.com, an American corporation, and the fact that the court sought to enjoin access to non-French websites stored on Yahoo!’s non-French servers. Of course, one can easily see why the court and the complainants in this action would have taken this additional step. Shutting down access to web pages on Yahoo.fr does no good at all if French citizens can, with the click of a mouse, simply go to Yahoo.com and access those same pages. On the other hand, Yahoo! argued that the French assertion of jurisdiction was impermissibly extraterritorial in scope.49 According to Yahoo!, in order to comply with the injunction it would need to remove the pages from its servers altogether (not just block access for French people), thereby denying such material to non-French citizens, many of whom have the right to access the materials under the laws of their countries. Most   T.G.I. Paris, May 22, 2000, https://perma.cc/738B-V9BM.   See Code pénal [C. pén.] art. R. 645-1 (Fr.) (prohibiting the public display of Nazi memorabilia except for the purposes of a historical film, show, or exhibit). 48   See T.G.I. Paris, Nov. 20, 2000, supra note 46 (noting that Yahoo! France had posted warnings on its site that through Yahoo! US the user could access revisionist sites, the visiting of which is prohibited and punishable by French law). 49   Id. 46 47

BARFIELD PRINT.indd 673

13/11/2018 11:53

674  Research handbook on the law of virtual and augmented reality important, Yahoo! argued that such extraterritorial censoring of American web content would run afoul of the First Amendment of the US Constitution. Thus, Yahoo!—and others50—contended that the French assertion of jurisdiction was an impermissible attempt by France to impose global rules for internet expression.51 Of course, the extraterritoriality charge ran in the opposite direction as well. If France were not able to regulate this content, then it could plausibly be said that Yahoo! was essentially exporting First Amendment norms to France. Today’s technology makes it far easier for websites to track the physical location of users than was the case in 2000, but the essential problem for virtual-world designers remains: Will they potentially face liability for content of their virtual spaces based on the laws in the jurisdiction of each end-user? Even if they can track where those end-users are, altering the elements of the virtual world to accord with each jurisdiction’s laws may be difficult. The EULA could, in theory, solve such jurisdictional and choice-of-law problems by imposing a uniform set of rules on all end-users, regardless of where they are in the physical world. Yet, EULAs have three possible shortcomings. First, because the EULA is a contract, it is potentially subject to local contract law. And even if the EULA contains a choice-of-law clause, it is possible that a court would override that clause by invalidating the contract itself as unconscionable or a violation of local public policy. Second, because the EULA is a contract between the game designer/distributor and each user, it might not cover torts one user suffers in a virtual world due to the acts of another user. And third, the EULA similarly would not apply if local regulatory authority claimed that criminal behavior was taking place in the virtual world, which of course could include acts deemed criminal in one jurisdiction but not in others. Virtual worlds and territorially based sovereignties are also potentially forced to interact because objects and attributes in virtual worlds may have market value outside of the virtual world, setting up the possibility of transactions that territorially based authorities (and game designers themselves) may wish to regulate.52 For example, in many virtual worlds, avatars that possess certain objects or territory or that have acquired certain experience can gain access to parts of the world others cannot, or they can do tasks or perform feats that others cannot. As such, these objects or territories or attributes have value to those immersed in the world. And of course, markets will almost inevitably arise to allow the buying and selling of nearly anything that human beings value. As long as that market operates only within the virtual world itself, exchanges are likely to be governed by in-world code, rules, and the EULA. But as soon as two people

50   See, e.g., Carl S. Kaplan, Experts See Online Speech Case as Bellwether, N.Y. Times, Jan. 5, 2001, www.nytimes.com/2001/01/05/technology/experts-see-online-speech-case-as-bellwether.html (quoting the warning of Barry Steinhardt, associate director of the American Civil Liberties Union, that if “litigants and governments in other countries . . . go after American service ­providers . . . we could easily wind up with a lowest common denominator standard for protected speech on the Net”). 51   As Greg Wrenn, associate general counsel for Yahoo!’s international division, put it: “We are not going to acquiesce in the notion that foreign countries have unlimited jurisdiction to regulate the content of U.S.-based sites.” Id. 52   See Grimmelmann, supra note 45.

BARFIELD PRINT.indd 674

13/11/2018 11:53

Legal jurisdiction and the deterritorialization of social life  675 exchange virtual items in-world while paying each other using PayPal or Bitcoin or a credit card or a check, then territorially based authorities may well see the transaction as similar to any other and therefore within their regulatory jurisdiction. If the parties are from different countries, we might have a jurisdictional or choice-of-law problem, just as we would with any other cross-border commercial transaction. But perhaps the law of the virtual world could be deployed by a court as a touchstone for adjudicating the dispute. For example, we might see courts using some idea of comity to defer to the norms of the virtual world, just as courts sometimes use comity to defer to norms of other jurisdictions even when not literally obligated to do so. Of course, the dispute may not only be between two end-users. What if the designer of the virtual world objects to the creation of this market, or demands to have some power over the transaction, or a financial cut?53 We might then see claims based on property rights in the servers on which virtual worlds run.54 Or perhaps the designer could argue that the players are accessing the designer’s computer systems without authorization for a nonpermitted use.55 The trade could also violate the EULA, spurring a contract claim. Or the market in virtual goods could be violating the designer’s trademark or copyright. Each of these claims would likely be pursued in a territorially based court, invoking territorially based law. Finally, an end-user might bring suit against a game designer/operator. In Bragg v. Linden Research,56 for example, a user sued the operator of Second Life because the operator nullified a transaction involving virtual property. The federal district court determined that Pennsylvania could assert jurisdiction even though the defendants were not from the state, because a nationwide advertising campaign had explicitly encouraged people to join Second Life in part by promising them the ability to take ownership of virtual property. In addition, Second Life had accepted payment from the plaintiff knowing the plaintiff was in Pennsylvania. If such contacts are sufficient than it is not at all farfetched to think that suits against operators of virtual worlds could well be brought anywhere a user is located. Moreover, it is worth noting that the court also did not honor the terms of service governing Second Life—which dictated mandatory arbitration and other provisions—because the contract was ruled to be unconscionable and therefore invalid. This case therefore illustrates both the potential for broad assertions of jurisdiction against virtual worlds and the possibility that local law could be deployed to displace EULAs. In contrast, in Mason v. Mach. Zone, Inc.,57 a court refused to find a cognizable claim regarding in-game property. In Mason, the plaintiff tried to bring a class action lawsuit against a game operator who sold in-game “gold” for real dollars and then allowed players to gamble the gold at an in-game casino. The plaintiff alleged that the algorithms

  See id.   See, e.g., Intel Corp. v. Hamdi, 30 Cal. 4th 1342 (2003) (recognizing a trespass theory for unauthorized use of computer system). 55   See Computer Fraud and Abuse Act, 18 U.S.C. § 1030, (establishing criminal and civil liability for “knowingly access[ing] a protected computer without authorization, or exceed[ing] authorized access”); see also, e.g., Orin Kerr, Cybercrime’s Scope: Interpreting “Access” and “Authorization” in Computer Misuse Statutes, 78 New York University Law Review 1596 (2003). 56   487 F. Supp. 2d 593 (E.D. Pa. 2007). 57   140 F. Supp. 3d 457 (D. Md. 2015), aff’d, 851 F.3d 315 (4th Cir. 2017). 53 54

BARFIELD PRINT.indd 675

13/11/2018 11:53

676  Research handbook on the law of virtual and augmented reality controlling the games of chance in the online casino were deliberately rigged against users, making it more likely that they would squander the gold that they had purchased (and presumably make them more likely to purchase yet more in-game gold to replace what they had lost). Interestingly, the court did not take seriously the idea that being deprived of virtual currency could be a valid legal claim even if that currency were purchased with real dollars. According to the court, “Perceived unfairness in the operation and outcome of a game, where there are no real-world losses, harms, or injuries, does not and cannot give rise to the award of a private monetary remedy by a real-world court.” As formulated by the court, this statement seems potentially plausible, but the whole point of the claim in Mason rested on the fact that there actually were “real-world losses, harms, or injuries,” namely the dollars the plaintiffs spent on the in-game gold that was then allegedly stolen from them through a rigged algorithm. To take it out of a game context, if an individual purchases Bitcoins using dollars and then is swindled out of the Bitcoins by a Bitcoin exchange vendor, it seems unlikely that a court would dismiss such a claim as not alleging any real-world harm. Indeed, as we will see, in a case presenting just such an issue, a court did in fact find a potentially valid cause of action. Thus, the “where” question in disputes involving virtual worlds remains contested. Users could be deemed in the world or outside of it, and if outside then we need to answer the further question of what community affiliation is then the most salient for jurisdictional and choice-of-law purposes. Finally, as the Mason decision illustrates, the community affiliation question can also inform how courts evaluate the content of the claim. 2.2  Cloud-Based Data In a sense, of course, many people live at least part of their lives in virtual worlds whether they choose to or not. This is because more and more of our social identity is stored remotely, far from our physical location. Whether it be our emails, our social media posts, our musical preferences, our virtual-world activities, our online search histories, or our banking and health information, very little of our data identity actually remains tied to our person anymore. This increasing deterritorialization of data and identity creates conundrums for jurisdiction and choice of law. For example, in 2013 US officers conducting a criminal drug investigation sought a search warrant under federal law to seize the emails of a Microsoft customer. This is usually a relatively routine process, and as long as the search warrant is valid, then data storage companies such as Microsoft generally comply. And in fact Microsoft did turn over all account information it had that was being stored in the United States. However, the actual emails and their contents were stored overseas in Dublin, Ireland. Microsoft refused to turn over this content, arguing that the federal law pursuant to which the search warrant was issued—the Stored Communications Act—could not be applied extraterritorially. The US Court of Appeals for the Second Circuit agreed, albeit reluctantly.58 Applying the presumption against extraterritoriality, the court determined that because Congress

58   Microsoft v. United States, 829 F. 3d 197 (2d Cir. 2016), vacated as moot, 138 S. Ct. 1186 (2018).

BARFIELD PRINT.indd 676

13/11/2018 11:53

Legal jurisdiction and the deterritorialization of social life  677 had not, in the Stored Communications Act,59 contemplated cloud-based data storage, it had made no provision for warrants to apply beyond US borders. Significantly, the physical location or the nationality of the underlying subject of the investigation was irrelevant. Thus, as interpreted by the Second Circuit, the Stored Communications Act might not permit authorities to obtain a search warrant and seize email records of a US citizen located in the United States who sent emails to other US citizens from a computer located in the United States. The only relevant territorial nexus appears to be where the email data happens to be stored. And this storage decision is entirely within the control of the storage provider, leaving open the possibility of manipulation in order to avoid the law of a particular sovereign. In contrast, a district court in Pennsylvania subsequently ruled the opposite way on a similar warrant involving Google.60 Here, the data question was in some ways even more difficult because Google does not store customers’ data in one location, such as Ireland. Instead, Google uses an algorithm that divides an individual’s user data across data centers and even splinters the data such that an email is not stored as a “cohesive digital file” but in “multiple data ‘shards,’” each in a separate location around the world. Accordingly, even if US law enforcement sought the data through government-to-government treaty, there would be no one government to whom to address the request. Unlike the Second Circuit, the court in the Google case reasoned that, if the Stored Communications Act is meant to protect Fourth Amendment privacy interests, then the relevant question is where the potential invasion of privacy takes place, not where the data is located. And given that Google can move customers’ data at will around the globe, the court reasoned that forcing Google to reterritorialize the data in the United States does not violate any privacy interest. Then, once the data is repatriated, the warrant can issue just as it would in any other domestic situation. In order to resolve this ambiguity, the US Congress in 2018 enacted the CLOUD Act.61 Under this statute US data and communication companies must provide stored data for US citizens on any server they own and operate when requested by warrant, regardless of where in the world that data happens to be stored. Thus, the statute sensibly looks at the underlying community affiliation of the user rather than the arbitrary territorial location of the data. On the other hand, the Act does provide mechanisms for the communications companies or the courts to challenge or reject warrants if they believe the request violates the privacy rights of the foreign country where the data is stored. This caveat still seems to unduly reify the physical location of data even though that physical location may be arbitrary and completely unrelated to the social reality of the person whose data is at issue. 2.3  Online Search As with data, online searches are part of our social identity. And, as with data, searches are fundamentally deterritorialized, linking searchers anywhere in the world with websites

  18 U.S.C. §§ 2701–12 (2012).   In re Search Warrant No. 16–960–M–01 to Google, No. 16-1061-M, 2017 WL 471564 (E.D. Pa. 2017). 61   CLOUD Act, HR 4943 (2018). 59 60

BARFIELD PRINT.indd 677

13/11/2018 11:53

678  Research handbook on the law of virtual and augmented reality located anywhere in the world. But what if a territorially based sovereign wants to block certain search results because the sovereign objects in some way to the underlying website that would otherwise be retrieved in the search? In such circumstances, as in the Yahoo! case, regulators may focus on the intermediary—here the search company—rather than the offending website, because it is far easier to find the search company and deputize it to leverage the regulation. In 2014 the European Court of Justice (ECJ) took this approach in a case involving Google.62 A 1995 European Council data privacy directive had recognized that individuals possess a privacy right in data.63 As interpreted by the ECJ, such a right allows individuals to object to old reputation-damaging online information about them that is no longer relevant and not of sufficient public concern to continue to be searchable. Significantly, rather than apply the directive against the website operator, the court ruled that it was Google, as the search operator, who bore responsibility for ensuring that websites containing this sort of obsolete private information be blocked from search results. Moreover, again as with the Yahoo! case, the Court deemed it insufficient only to apply its decree to google.es, the Spanish subsidiary, instead ruling that google.com must also block offending websites from its search results. By making Google responsible for enforcing this so-called “right to be forgotten”, the ECJ effectively deputized Google as a sort of administrative agency. Henceforth, any individual seeking to have a website blocked from Google search must first file a notice with Google. Then, it will be Google’s legal team that will apply the ECJ’s balancing test to see if the elements of the right to be forgotten are satisfied and there is insufficient countervailing public interest in the information remaining accessible. If the individual disagrees with Google’s decision, then that decision can be challenged in court. Given that Google is constantly altering its search algorithms anyway, one can understand why the court would view this as an effective division of labor. Moreover, Google is not jurisdictionally constrained regarding the websites it blocks from its search algorithms, as a government regulator would be if it sought to have a website taken down. Nevertheless, the fact is that a European court has required Google, a US corporation, to perform a quasi-governmental adjudicatory function on a worldwide basis (albeit only at the request of European Union citizens). This case, therefore, illustrates not only issues of jurisdiction, but also the increased importance of intermediaries such as virtual world operators, online service providers, social media companies, and search engines. Given that our social lives are conducted through, or stored with, these intermediary companies, those companies are likely to become the brokers that territorially based governments use to pursue regulation. And, of course, these companies’ increasing power over data means that they will often be the target of complaints by individuals.

62   Case C-131/12, Google Spain SL v. Agencia Española de Protección de Datos (May 13, 2014), http://perma.cc/ED5L-DZRK. 63   Council Directive 95/46, 1995 O.J. (L 281) 31 (EC).

BARFIELD PRINT.indd 678

13/11/2018 11:53

Legal jurisdiction and the deterritorialization of social life  679 2.4  Global Electronic Currencies and Transactions Although the global alternative currency Bitcoin has generated the most attention, it is only one instantiation of a more general ledger technology known as blockchain. Blockchains store data in distributed computers and chain them together to form an unbroken record of that information. The information stored could be currency transactions, but it could also be any automated executable set of instructions, such as an insurance contract that pays out automatically if a given event occurs. Two features make blockchain technology valuable. First, identical copies of the particular blockchain (or ledger) are stored on, and accessed from, potentially thousands of computers around the world. Any changes to information on one are immediately and automatically authenticated by the others, and any authenticated change immediately updates on all computers in the chain. Second, the information is encrypted so that, in combination with its decentralization, it is difficult to hack. On the other hand, those same features make blockchains potentially difficult to regulate. Not only do blockchain transactions cross borders, but it will often be difficult to identify a particular computer or entity that is responsible if there is a dispute or problem. As one commenter has described it, “the infrastructure does not fall under any traditional jurisdiction, but the users of the infrastructure also naturally evade any sense of traditional jurisdiction. All parties may transact entirely anonymously on a public blockchain.”64 So far, blockchain technology has not been deployed sufficiently for us to know precisely how legal challenges are likely to be resolved. And the assumption that blockchain transactions completely lack connection to a territorially based entity may be overstated. That is because the parties to a blockchain transaction are still physically located somewhere on Earth, just as the participants in a virtual world are physically somewhere sitting in front of a screen. And, to the extent that money changes hands, that money is in some form sent from one physical place to another, creating a territorial nexus. For example, in Greene v. Mizuho Bank, Ltd.65 the plaintiff wired money from a Wells Fargo branch account in California to a bank in Japan, which then held the funds in an account used by a Bitcoin exchange. So, although the Bitcoin transactions themselves might not have a location, the exchange interacted with a bank chartered in Japan, and that bank in turn interacted with a bank in California, making all of the defendants potentially subject to jurisdiction in California. Of course, when we say that money was “sent from California to Japan,” we are really talking about metaphysical electronic signals crossing borders, and so we are in a sense using what is already a fictional connection to the physical world to territorially ground a Bitcoin transaction, which is an even more fictional connection to the physical world. It remains to be seen at what point all that is solid will melt into air and legal jurisdiction based on territory will cease to be meaningful as a way to describe and regulate electronic transactions at all.

64   Wulf A. Kaal & Craig Calcaterra, Blockchain Technology’s Distributed Jurisdiction (June 20, 2017), https://perma.cc/3PMT-T98E. 65   169 F. Supp. 3d 855 (N.D. Ill. 2016).

BARFIELD PRINT.indd 679

13/11/2018 11:53

680  Research handbook on the law of virtual and augmented reality 2.5  Autonomous Agents The development of unmanned aerial vehicles and increasingly autonomous weapons systems, and the potential use of drones, autonomous automobiles, and other robots using artificial intelligence radically challenge legal conceptions of jurisdiction and accountability. In particular, autonomous, machine-learning vehicles tend to spread responsibility for decisionmaking across a larger number of actors. Thus, it becomes difficult to determine who is responsible for a machine-based action when it causes harm, or where the relevant action took place in order to determine legal responsibility and jurisdiction. So far, this issue has been most explored in the context of automated weaponry and warfare. For example, consider the Israeli Harpy which, once launched by a human operator, can detect an enemy radar system and then autonomously dive bomb and strike that target. In the words of Laura Dickinson: If the Harpy killed a large number of civilians in a manner that could be said to violate international humanitarian law, who could or should be held responsible? The human being who has the responsibility to override the weapons system? The commander of the territory where the weapons system was deployed? The individuals who set policy for using the technology? The individuals who drafted the targeting criteria? The engineers who designed the weapons system to apply the targeting criteria? Anyone who supplied intelligence that fed into the weapons system and that formed the basis for target selection? In the case of even partially autonomous systems, it is difficult to locate a responsible human agent.66

Similar questions can, and almost certainly will, be raised by automation in vehicles and by machine-learning based robotics and drones for commercial and consumer use. These new deployments will likely reshape legal rules regarding product liability, insurance, contract, jurisdiction, criminal law, and other areas.

3.  LAW’S RESPONSE Or possibly not. Ever since the rise of the commercial internet in 1995, legal scholarship regarding online innovation has often divided into two broad camps. On one side were the cyberspace “unexceptionalists” who argued in various contexts that the online medium did not significantly alter the underlying legal framework and that well-settled principles of law therefore can simply be applied to online interaction.67 On the other 66   Laura A. Dickinson, Drones, Automated Weapons, Private Military Contractors, and Their Challenge to Domestic and International Legal Regimes Related to Armed Conflict, in The Promise and Peril of Human Rights Technology 93, 122 (Molly K. Land & Jay D. Aronson eds, 2018); see also Marcus Wagner, The Dehumanization of International Humanitarian Law: Legal, Political, and Ethical Implications of Autonomous Weapons Systems 47 Vanderbilt Journal of Transnational Law 1371 (2014); H. Roff, Killing in War: Responsibility, Liability and Lethal Autonomous Robots, in Routledge Handbook of Ethics and War: Just War Theory in the 21st Century (A. Henschke et al eds, 2013). 67   Jack L. Goldsmith, The Internet and the Abiding Significance of Territorial Sovereignty, 5 Ind. J. Global Legal Studies 475 (1998); Allan R. Stein, The Unexceptional Problem of Jurisdiction

BARFIELD PRINT.indd 680

13/11/2018 11:53

Legal jurisdiction and the deterritorialization of social life  681 side, ­cyberspace “exceptionalists” argued that the medium itself created radically new problems that required new analytical work to be done.68 This was a false dichotomy from the start, of course. First, one’s perspective on whether the online medium creates a new problem depends in large part on what the legal question is. For example, defamation is generally defined as the communication of a false statement about someone with the requisite degree of intent. This communication can be written or oral, and if it is written, there does not seem to be much of a legal difference whether the written defamation is communicated by postal mail, fax, email, or through social media. Thus, unexceptionalism seems to apply to defamation law. But when we turn to the question of intermediary liability and ask whether (and under what circumstances) various types of internet service provider should be held liable for defamatory content accessible through their portals, then we have a question that is not easily answered without some conceptual understanding of what an internet service provider is or should be. So, online defamation issues require either an exceptionalist or an unexceptionalist approach, depending on whose liability one is discussing. The second reason that the dichotomy was unhelpful from the beginning is that the unexceptionalist position assumes that there actually are well-settled “general” principles of law that can simply be applied to new legal settings without alteration. And yet, it is the nature of law (and especially the common law) that it changes over time. Thus, what is well settled for one generation (or in one century) is apt to be very different from what is well settled for the next. Even more importantly, new technologies that alter the culture are precisely the sorts of changes that tend to result in shifts to well-settled legal principles. For example, as previously discussed, in the nineteenth century “well-settled” principles of legal jurisdiction and choice of law saw jurisdiction as rooted almost exclusively in the territorial power of the sovereign. Each sovereign was deemed to have jurisdiction, exclusive of all other sovereigns, to bind persons and things present within its territorial boundaries. By the early twentieth century growth of interstate commerce and transportation put pressure on the idea that a state’s judicial power extended only to its territorial boundary. In particular, the invention of the automobile and the development of the modern corporation meant that faraway entities could inflict harm within a state without actually being present there at the time of a lawsuit. Not surprisingly, by the end of the twentieth century it had become “well settled” that a state may indeed assert jurisdiction over a defendant if the effects of the defendant’s activities are felt within the state’s borders, even if the defendant has not literally set foot there. Likewise, it had become “well settled” that choice-of-law rules could be based on governmental interests or relationships as well as territorial connections. And, of course, these new “well-settled” rules felt as

in Cyberspace, 32 International Lawyer 1167 (1998). Many of these arguments are reprised in Andrew K. Woods, Against Data Exceptionalism, 68 Stan. L. Rev. 729 (2016). 68   David R. Johnson and David G. Post, Law and Borders: The Rise of Law in Cyberspace, 48 Stan. L. Rev. 1367 (1996); David G. Post, Against “Against Cyberanarchy”, 17 Berkeley Tech. L.J. 1365 (2002). Many of these arguments are echoed in Jennifer Daskal, Borders and Bits, 71 Vanderbilt L. Rev. 179 (2018); see Paul Schiff Berman, Legal Jurisdiction and the Deterritorialization of Data, 71 Vanderbilt L. Rev. En Banc 11 (2018) (responding to Daskal’s article and pointing to continuities with earlier legal scholarship concerning online interaction).

BARFIELD PRINT.indd 681

13/11/2018 11:53

682  Research handbook on the law of virtual and augmented reality commonsensical and obvious to most judges, lawyers, and observers as the territorialist view felt in the nineteenth century. It seems safe to say that jurisdictional and choice-of-law rules are now in flux again, at least in part because of online interaction. Indeed, as many of the cases described in this chapter suggest, the idea of basing jurisdiction on where effects are felt is difficult to apply to online interaction because our social lives may cause effects or be effected in many different locations, anywhere data is stored, used, or viewed. The answers that law will ultimately evolve to address the sorts of problems raised in this chapter are difficult to predict, and scholars and judges will no doubt have differing approaches to specific questions of jurisdiction and choice of law regarding online interaction, virtual worlds, data storage, digital currencies, autonomous entities, and the like. Suffice to say that however one resolves the issues, “well-settled” principles of law are unlikely to be very helpful because such principles are themselves always in flux, often precisely because of the pressures placed on such principles by new communications technologies such as the internet and new ways in which social lives become deterritorialized. Thus, in some sense a pure unexceptionalist position is difficult to maintain. But if unexceptionalists have relied too much on the application of mythical well-settled principles, the exceptionalists have at times tended to the opposite extreme, assuming that the rise of online interaction, data storage, and the like upends nearly all extant ideas about law and the role of the state. Moreover, as the discussion of digital data makes clear, governments need not only act through the traditional apparatus of lumbering, territorially limited law enforcers or regulators. Instead, they can commandeer online entities such as search engines or online service providers or virtual world operators to regulate on the government’s behalf, either by turning over data or by building certain legal requirements into the code that dictates online activity itself. This point about building law into code requires a final comment. Many of us tend to think of regulation solely as the coercive commands of a sovereign entity. When the legislature passes a law proscribing some behavior, or an administrative agency establishes rules, or a court issues an order, it is easy to see how such activity regulates behavior. In contrast, it is more difficult to recognize the powerful way in which our environment and the architecture of our space applies a regulatory force. If one wanted to prohibit automobiles from driving into a public park, for example, one could pass a law forbidding such conduct and then hire police officers to patrol for violations. But one could also simply build walls that make it much more difficult (if not impossible) as a practical matter for the automobile to enter the park in the first place. It is important to realize that both the law and the wall function as regulatory tools. Online, this second type of regulation is likely to be more powerful than the first because the “architecture” of online activity—what is permitted and what is not—is determined by software code, which, by its very nature, is infinitely malleable and operates through the technology itself.69 This code-based law can be mandated by governments, as when the code underlying a Google search algorithm simply does not show a website that is definitely in existence because that website violates European Union law. But it can also

69

  See Lawrence Lessig, Code: And Other Laws of Cyberspace (1999).

BARFIELD PRINT.indd 682

13/11/2018 11:53

Legal jurisdiction and the deterritorialization of social life  683 be wielded by private entities, as when virtual-world designers decide what will be possible and what will be impossible to do in the world. Thus, the movement of our social activity into various forms of the virtual world is a real trend, and it is one that is bound to unsettle previously settled legal principles. Yet, a time of flux is also a time of opportunity. As judges, legislators, and scholars struggle to apply old legal principles to new contexts, they are—in a far more self-conscious way than usual—questioning whether those old legal principles really work in the brave new world they are encountering. Such a time of self-conscious inquiry opens the conceptual space to allow one to go back to first principles and ask important jurisprudential and sociologically charged questions that run throughout all of law. And what we find is law and society turning like a Möbius strip, one into the other.

SOURCES Ala. G. S. R.R. v. Carroll, 97 Ala. 126, 11 So. 803 (1893) Babcock v. Jackson, 191 N.E.2d 279 (1963) Bragg v. Linden Research, Inc., 487 F. Supp. 2d 593 (E.D. Pa. 2007) Case C-131/12, Google Spain SL v. Agencia Española de Protección de Datos (May 13, 2014) Coleman’s Appeal, 75 Pa. 441 (1874) Council Directive 95/46, 1995 O.J. (L 281) 31 (EC) Greene v. Mizuho Bank, Ltd., 206 F. Supp. 3d 1362 (N.D. Ill. 2016) In re Search Warrant No. 16–960–M–01 to Google, No. 16-1061-M, 2017 WL 471564 (E.D. Pa. 2017) Intel Corp. v. Hamidi, 30 Cal. 4th 1342 (2003) Int’l Shoe Co. v. Washington, 326 U.S. 310, 66 S. Ct. 154 (1945) Mason v. Mach. Zone, Inc., 140 F. Supp. 3d 457 (D. Md. 2015), aff’d, 851 F.3d 315 (4th Cir. 2017) Microsoft v. United States, 829 F. 3d 197 (2d Cir. 2016), dismissed as moot 138 S. Ct. 1186 (2018) Pennoyer v. Neff, 95 U.S. 714 (1877) T.G.I. Paris, May 22, 2000, https://perma.cc/738B-V9BM

BARFIELD PRINT.indd 683

13/11/2018 11:53

BARFIELD PRINT.indd 684

13/11/2018 11:53

Index advertising xvi, 5–6, 19, 36, 53, 104, 197, 199–204, 211, 307–8, 324, 331–2, 421–512, 530–31, 540, 543, 546, 552–4, 559, 561, 563, 576, 589, 612, 631 billboards 16, 201–2, 204, 213, 304–6, 338, 432, 434, 440, 442, 460, 465–7, 514, 531, 545, 552–3, 558–9 children’s 429, 463, 473, 482, 490–97, 510 false advertising claims 6, 36, 458–62, 468, 545 keyword advertising 200–204, 211, 213–14, 439 location-based 199–204 regulatory regimes 462–7 subliminal 489 video games 552–4 artificial intelligence (AI) xv, 16–17, 20, 22, 26–7, 37–9, 41, 158, 162, 365, 382, 389–419, 483, 662, 680 autonomous virtual agents 680 entities 396 expert system 18 genetic algorithm 17–18, 22, 30 neural networks 17–18, 22, 26, 32, 38, 40 punishing AI 412–17 augmented reality (or mixed reality) architecture 123 art 306, 315, 324, 337 commerce/shopping 119–20 computer games (including location-based games) xv, 6–7, 55, 119, 156, 312, 313–33, 398, 424, 425–9, 435–8, 447–8, 493–4, 591–602 consumer products 103–4, 613–14 consumer reviews 109, 616, 624, 630–32 driving 120, 307, 309, 326, 331, 335, 338, 426–8, 611 education 116–17, 611, 612 historical development 311–12, 608–9 medical 119, 500, 614–15 overview 108–9, 115–17 tourism xv, 109, 309, 312, 318, 320–21, 331, 609, 612 avatars xv–xvi, 2–43, 65–7, 79–83, 85, 87, 91, 100, 114–15, 119, 212–13, 278, 308, 335, 353, 363, 365, 384, 391, 393–4, 397, 400,

451, 456–8, 475, 479, 486, 514, 516–18, 536, 548–63, 569, 571–2, 578, 581–2, 589, 639, 640–42, 644, 647–8, 652, 654, 656–7 avatar self-awareness 26 contracts, entering into 41 death or deletion 86–9, 578, 589 harm to (and harm to persons through avatars) 98, 363–88 liability of 400, 402, 406, 408, 413 neural networks (for creating smart avatars) 17, 26, 40–41 rights of 7, 77–8 smart avatars 4, 13, 16 biometrics (including facial recognition) 51–2, 62, 110, 132, 135, 136, 139, 210–14, 450–51, 473, 479, 480, 482, 490, 495, 504–10, 513, 615 children 40, 463 advertising for 490–97 and virtual worlds 69, 103, 116–17, 246, 333, 340, 429, 453, 463, 473, 490–97, 499, 510, 522–3, 654 choice of law 663, 666–7, 670–71, 674–5, 676, 681–2 cloud 113, 122, 124, 676–7 cognitive enhancement 16, 19 constitutional rights and civil liberties 74–8 due process (under the Fifth and Fourteenth Amendments) 239, 247, 270–73, 466, 666, 669 Equal Protection 16, 323, 523 First Amendment 20, 29, 59, 189, 205–7, 234, 242–74 commercial speech 19, 237, 245, 308, 324–5, 331–2, 422, 443, 461–2, 531, 546 content-neutral restriction 307–8, 317, 325–30 copyright law, fair use and the First Amendment 186–9 coverage 247–55, 258, 260, 263, 266, 272–3, 312–24 freedom of thought 243, 265–74 functionality doctrine 295–9

685

BARFIELD PRINT.indd 685

13/11/2018 11:53

686  Research handbook on the law of virtual and augmented reality public forum doctrine 333–8 publicity rights 530, 532–48 purposes 285–8 Second Amendment 269 software code as speech ­29 Spence test 250–51, 253, 274, 279–84, 287–90, 316–17, 322 symbolic or expressive conduct 248, 276–77, 281–2, 290–91, 301–2, 322, 327 trademark and 59, 205–7, 234, 236–9, 542–51 contracts xvi, 4, 9–10, 14, 16, 21, 38, 41, 50–51, 53, 65, 78–88, 92, 93, 96–102, 104, 118, 125–32, 144–50, 579, 604–35, 646, 666, 670 adhesion 9–10, 84, 132, 579, 618–19 collective bargaining 515, 556–63 EULAs (end-user license agreements) 9, 12–13, 84, 90, 95, 104, 124, 126–32, 135–6, 143, 146, 148, 558, 568–70, 574–8, 646, 650, 673–5 limitation of liability 7, 12, 61, 98 non-compete contracts 50, 674­–6 copyright xv–xvi, 2–3, 5, 8, 12–13, 15, 19–30, 34, 37–41, 50–51, 60–62, 80, 125, 127–30, 135, 145–8, 150, 154–90, 218–20, 224–7, 229, 232–3, 238, 239, 267, 323, 345, 347, 391, 393, 422, 425, 437, 443–6, 519, 523–7, 533–51, 555, 558–9, 561–3, 570, 573, 575, 675 authorship (virtual) 20–26 Berne Convention 33–5, 168 characters 30–33, 156 copying of real objects 171–2 derivative works 28, 150, 155, 174, 180, 184, 188, 443–6, 551 fair use 128, 164, 174, 186–9, 218–21, 223, 523–7, 533–4, 536, 542–3, 548, 551 first sale doctrine 12, 145 fixation in tangible medium 21, 445 functionality (or scènes a faire) 166, 172–83, 220 idea-expression dichotomy 172–83, 184–6, 189, 220, 523–6 licensing 123, 125, 127–30, 135, 145 “look and feel” 165–8, 172–90, 216, 219, 221, 229 moral rights doctrine 34 originality 20–23, 32, 166, 169–71, 179, 182, 189, 225 programmer’s rights to program output 27 protection of visual images 30 software 165–8

BARFIELD PRINT.indd 686

visual displays (generated by software) 166–8 Copyright Office principles on 166–8 writing style 27 crime (virtual) assault 358–9 burglary 343–4, 356–8 cybercrime 10 sexual assault 363–88 criminal responsibility for 378, 381–6 terrorism 10, 268–9 crowdfunding 48 cybersecurity 360–61, 474, 660 hacking 248, 317, 340–62, 384 data privacy see privacy e-commerce 10, 201, 620–23, 625 European Union 10, 216, 473, 475, 479–82, 486–90, 497, 501–4, 508–11, 672, 678–9 Data Protection Directive (DPD) 480–82, 511, 678 e-Privacy Directive 481, 487–8, 502, 504 e-Privacy Regulation 480, 487 EU General Data Protection Regulation (GDPR) 473, 479–82, 486–90, 503–4, 508–10, 512 extended reality (xR) 217–18 facial recognition see biometrics Federal Trade Commission (FTC) 48, 52–3, 141, 148, 463–5, 476, 482–5, 493–5, 499–500, 502, 506, 530–32, 552 flight simulation 15, 158, 244–5, 255–7, 271, 273, 311, 643 funding VR startups 44–64 geolocation 109, 111, 116, 137, 204, 422, 424, 443, 472–3, 482, 487–90, 497–504, 591, 609, 612 Global Positioning System (GPS) 109–10, 130 Identification Services (IDS) 110–11 haptics xiv, xvi, 193–9, 217, 256, 275–303, 311, 333, 367–8, 375, 377, 386, 504, 653 current uses 278–9 full body suit 279, 299, 375, 386 glove 278–9 harm 300–301 historical development of 278–9 relation to emotion 294–9 relation to sexual assault 377 teledildonics 275–81, 292, 295, 298–302, 367–68, 377–8, 386

13/11/2018 11:53

Index  687 trademarking of textures and haptic sensations 193–9 harm xvi, 64, 90–91, 94, 98, 101, 107, 120, 133–4, 245–6, 267–70, 300–302, 307, 331–2, 338, 340–88, 432–5, 530, 538–9, 581, 676, 680 cognitive or psychological 53­–6, 243, 245, 300–301, 332 dignitary or reputation 107, 133, 308, 331, 530, 538–9 economic 308, 310 hacking 340–62 intentional 242–3, 268–9, 324 fact-finding, with respect to 90–91 physical 53–6, 107, 120–21, 302, 307, 310, 327–8 property 6, 342–3, 356–8, 430–34, 584, 593–4 within virtual worlds 581 Internet of Things (IoT) 488, 504 jurisdiction 9, 10, 82–4, 96, 99–100, 662–83 minimum contacts 10, 669–71 land use 338, 578, 582–4, 664 mapping, virtual 115–17, 120–22, 130–31, 137 massively multiplayer online role-playing games (MMORPGs) 13, 515–19, 570–90, 642 nuisance see torts olfactory feedback (sense of smell) xiv, 3, 45, 156–7, 193, 267, 471, 637 patent rights 41, 56–8, 157–61, 185, 193, 202, 218–20, 222–4, 227, 230, 232, 238–9, 345, 396, 422, 437, 446–51, 490–92 design patent 57 doctrine of equivalents 57 inducement infringement 57 utility patent 57 payment systems 67, 121, 197, 245, 426, 451, 495, 576, 675 pervasive computing (PerC) (or ubiquitous computing) 106, 118, 121–3, 608–9, 613–14 Pokémon Go 5, 6, 204–5, 311–12, 317, 326–7, 328, 330, 428–9, 490, 514, 591–603 privacy 51, 138–44, 471–512, 615, 676–7 Data Protection Privacy Directive 480 “do not track” 142–4, 149, 477 e-Privacy Directive 481, 487

BARFIELD PRINT.indd 687

EU General Data Protection Regulation (GDPR) 473, 479–82, 486–90, 508–10 notice and choice (“opt-in” and “opt-out”) 143–4, 148–9, 482–90 personally identifiable information 474–5 privacy by design 141–2 right to be forgotten 678 product liability 6–7, 54–5, 258, 425, 680 property, ownership of virtual goods 9, 10, 15, 66, 78, 84, 136–8, 573–91, 636–61 items 573–4 land ownership 577–91 publicity, right of 19, 207, 451–8, 527–52 commercial value 208 indicia of identity 209 infringement in AR 210–14 overview 527–32 quick response (QR) codes 107, 111, 439, 447, 605 radio frequency identification (RFID) 106–7, 119, 121, 123–4 Reality-Virtuality (RV) continuum 111–13, 116–17 relation of virtual to physical world 125–6, 155, 164, 299–300, 366, 371–4, 379, 380, 651–3 the “Magic Circle” 95, 98, 125–6 psychological experience of VR 349–54 rule of law 64, 71–2 Second Life xv, 4–5, 7–11, 15, 66–7, 69, 79–80, 82–7, 93, 115, 246, 327, 363, 367, 369, 391, 422, 475, 515–17, 548, 555, 571–91, 601–2, 643, 647, 654, 658, 675 sense of presence 3, 262, 349–52, 355, 359, 371–2, 424 spatial metaphors 351, 353–4 statutes Anticybersquatting Consumer Protection Act 136 California Consumer Legal Remedies Act 458 California Consumer Privacy Act of 2018 478 California Online Privacy Protection Act (CalOPPA) 477–8 California “Shine the Light” Law 477 California’s Privacy Rights for California Minors in the Digital World Act 496 Children’s Online Privacy Protection Act (COPPA) 474, 476, 491–6 Communications Decency Act (CDA) §230 5, 133–5, 427, 429

13/11/2018 11:53

688  Research handbook on the law of virtual and augmented reality Computer Fraud and Abuse Act (CFAA) 87, 317, 343–8, 354, 361, 434, 675 Controlling the Assault of Non-Solicited Pornography and Marketing (CANSPAM) Act of 2003 476, 483–6 Copyright Act 12, 20–24, 28–9, 60, 145, 167, 185–6, 220, 229, 232, 444–5, 524 Digital Millennium Copyright Act (DMCA) 5, 8, 12, 61–2, 106, 128, 147, 323, 345 Federal Trade Commission (FTC) Act, Section 5 48, 52, 476, 486, 530 Federal Trademark Anti-Dilution Act of 1995 5, 441 Health Insurance Portability and Accountability Act (HIPAA) 474, 507 Illinois Biometric Information Privacy Act (BIPA) 506–7 Lanham Act 6, 35, 36, 58–9, 182, 193–7, 202, 205–7, 213, 216, 219, 221–3, 226–39, 439, 442, 458, 460–62, 465, 542, 547 Sexual Offences Act of 2003 (England and Wales) 366, 370, 373–5 Telephone Consumer Protection Act (TCPA 476, 486 Visual Artists Rights Act of 1990 (VARA) 33–4 topological vs. social space 664–6 torts 6–8, 41, 50, 54, 88–90, 96–8, 101, 126, 132–6, 307–8, 331–2, 343, 392, 430, 468, 475, 529, 533, 590, 593–4, 596, 666, 670, 674 defamation 9, 307–8, 324–5, 331–2 intentional infliction of emotional distress 308, 331–2 invasion of privacy see privacy negligence 6–8, 54, 670 nuisance xv–xvi, 307, 422, 428–30, 434–7, 468, 592, 595–601 product liability 6–7, 54–5, 258, 425, 680 publicity see publicity, right of tortious interference 50, 440 trespass 6, 342–3, 356–8, 430–34, 593–4 trespass to chattels 343, 430–34 trade secrets 41, 48–51, 166, 219–20, 345, 422, 425, 437, 446–51 trademark law xvi, 2–16, 12–13, 15, 30–33, 36, 40, 58–60, 136–8, 166, 192–240, 308, 391, 422, 425, 437–43, 447, 452, 458, 460, 462, 523–4, 528, 542–9, 551–3, 562, 645, 675 digital alteration of images 442 dilution 195­–6, 201, 239, 438, 440–41, 443 blurring 196, 441 tarnishment 196, 438

BARFIELD PRINT.indd 688

distinguishing function of a mark 193–4, 438 fair use 58–60, 205–7, 218, 223, 234–9, 532, 542–9, 551 infringement 194–5 initial interest confusion 440–41 likelihood of confusion 195, 438, 440, 543, 545, 547 overview 192–6, 438–9 trade dress 193, 215–40 overview and history 220–34 use in commerce 59–60 trespass see torts ubiquitous computing see pervasive computing (PerC) unfair competition law 30–33, 50­–51, 203, 207, 218, 442, 459, 529, 537, 540, 542–3, 547–8, 551 video games 6–7, 15, 46, 54, 58–9, 62, 66, 83, 112, 160, 168, 177, 183–4, 205, 209, 213, 238, 243–7, 255–65, 272–3, 278, 280–81, 298–300, 309, 311–12, 317–23, 332, 337, 377, 379, 421–2, 428, 438, 455–6, 513–15, 519–23, 533, 535–6, 540–42, 545–53, 557, 598 advertisements in 461 copyright and 160, 168–9, 444–5, 416 First Amendment protection for 243–7, 319–23, 332–3, 337, 461, 519–23, 533, 535–6 publicity rights and 209, 213, 455–6, 533, 535–6, 548–52 simulating sports 177, 256–7, 513–14, 541–2, 549–50 technology 160, 445 trademark and trade dress 238, 438, 545–51 virtual currency 9, 11, 59, 118, 316, 372, 417, 571–3, 592, 601, 644, 650, 671, 676, 679 virtual reality current applications art 307, 309 commerce 16 computer games xiv, xv, 2–5, 8–11, 13–17, 45–6, 53, 62, 65, 66–7, 70, 77, 81–3, 93–7, 114–15, 163, 168–9, 184, 244–5, 253–9, 263–4, 272, 278, 299, 308, 311, 364, 377–9, 394, 398, 421–2, 447, 505, 514–63, 568–91 education 16, 44, 50, 162–3, 246, 261, 518, 643 entertainment 162,163, 257–8, 298, 340, 515, 518

13/11/2018 11:53

Index  689 legal evidence 163–4 medical 14, 50, 56, 162, 163–4, 393, 242, 247, 253–4, 258, 261, 273, 340, 393, 415, 451, 643 military 44, 53, 161–2, 242, 257, 273, 398, 643 overview 113–14 psychotherapy xv, 16, 56, 242, 264, 273 technology Cave Automatic Virtual Environment (CAVE) 255, 273 haptics see haptics head-mounted display xiv, 3, 44–6, 49, 52, 161–2, 242, 255, 266, 273, 278, 310–11, 321, 363–4, 386, 422, 424, 462, 466, 495, 608–9 physical elements 120, 256, 271–2 training 242–74

BARFIELD PRINT.indd 689

historical development 45–6, 156–62, 307–11, 243 virtual worlds 64–102, 515–19, 560–63, 570–91, 638–43, 672–6 community norms in 88–90 company towns (analogous to) 336–7 currency in 571–3, 679 economy of 66, 466 governance of 74–8 code as law 91–4, 650, 682 legislation within 96–9 penalties 87 lawlessness in 70 legal literature on 118–21 overview 114–15 purchase of real world products in 4 rule of law in 64, 71–7 virtual money 66, 571–3

13/11/2018 11:53