Content as Data and Data as Content: Copyright Issues in Machine Learning When: November 23, 2020 at 12:00pm - 1:00pm EST - This event has passed ABOUT THE PROGRAM: In order for machine learning technologies to work, they require access to and processing of massive amounts of data. Sometimes that data can exist in the form of copyrightable content, such as photographs used to train an algorithm to recognize images. Other times it can take the form of unprotectable information that is compiled into copyright protectable databases or other compilations, such as analyzing tax databases to determine home values. Licensing markets for data used in training machine learning have already developed, but so have uses that data scientists believe are either not subject to copyright protection or qualify as a fair use. Disputes with content owners are likely to increase, as demonstrated in the ongoing litigation between UAB “Planner5D,” which provides a design tool that allows users to create virtual home designs, and Princeton University and Facebook, who allegedly scraped Planner 5D’s virtual objects and scenes for purposes of developing artificial intelligence applications to recognize real-world scenes. Our panel will provide a live demonstration of how machine learning converts pre-existing content into new expression through the GPT-2 and GPT-3 applications, discuss the data markets that enable development of machine learning tools, address the fair use and authorship issues raised by these technologies, and discuss the implications of the format changes of programs like GPT from an open innovation tool to a proprietary subscription service. SPEAKERS: Nadia Baksh is Legal Counsel at Oracle. She started her career at one of the world’s largest retailers, providing outstanding customer service and quality products. While building brands, she became fascinated with monetization strategies. She transitioned to an AI-driven real-time advertising company and learned the ropes. When the opportunity presented itself to attend law school and work with numerous technology companies, from traditional giants to innovative newcomers, she had to take it. Today, Nadia has a unique perspective on guiding customers through innovative, cloud based, legal solutions. She enjoys counseling business team members on minimizing risk while advancing the business. Hector Postigo holds degrees in Science & Technology Studies and Physiology and Neurobiology. Currently he is Associate Professor of Media Studies and Production at Temple University. He is the author of The Digital Rights Movement from the MIT Press and the co-editor of Managing Privacy through Accountability from Palgrave Press. His research anticipates emerging shifts in information technologies, video games, AI,and user behavior and how they will impact markets, privacy and intellectual property. Currently he’s focusing on Open AI’s GTP-2/3 as well as Google’s Transformer, using them to generate novel works in music and writing, and predicting how they may change the creative industries markets by facilitating AI assisted/novel creative production. Along with his work at Temple, over the last 5 years, Hector has also served as Research Associate for Rutgers Law School’s IIPL, Annenberg Fellow at Stanford University’s CASBS, visiting fellow at Yale Law School’s ISP, Visiting Researcher at Microsoft Research New England and co-founder of CultureDigitally.org. His past research has been funded by the NSF and the European Union. Additional information can be found at hectorpostigo.com Paul Reinitz is Director of Legal Advocacy and Operations Counsel for Getty Images. Paul leads the strategic oversight of Getty Images’ global advocacy strategy as well as managing the company’s legal operations. He advises Getty Images on a variety of legal issues including Copyright, rights of publicity and privacy. His work involves advocating for responsible and ethical innovation in the areas of artificial intelligence and machine learning including studying the interaction between computers, the creative process and Copyright. Paul and Getty Images are interested in exploring how visual data can be used in AI/ML and in developing standards for protecting IP and other third-party rights. Paul is active is various industry organizations and serves as the chair of the Digital Medial Licensing Association’s legal committee. Chad Rutkowski (Moderator) – Chad Rutkowski’s practice focuses on the intersection of copyright and technology. As co-lead of BakerHostetler’s Digital Transformation and Data Economy team, he assists clients in the midst of the “digital transformation,” helping them identify and capture the intellectual property (IP) in their valuable data, algorithms, and enterprise software. Clients rely on him to help identify their IP, to build internal processes for its management, navigate open source and other open innovation strategies, and to create licensing programs that enable full value realization. He further protects that value in disputes and litigation, whether through enforcing his clients’ IP rights or fending off attacks by competitors. Chad is the principal author of the firm’s Beyond Source Code web tool, which digests all court decisions addressing the scope of copyright protection in the non-literal elements of source code. Chad writes and presents regularly on the robust protections afforded by copyright law to software, data, artificial intelligence, and related innovations. He has helped his clients protect innovations in healthcare technology, water management, smart cities, population health management, and online consumer lending. Chad’s interest in how IP law has adapted to digital transformation began in his former role as a founder and business manager of a digital media publisher. Drawing from his professional and legal experiences, he co-leads the firm’s Technology & IP Transactions and Outsourcing practice team and contributes regularly to the firm’s IP blogs. Additionally, he is an active member of The Copyright Society and the ABA IP Law Section. Chad is AV Preeminent-rated by Martindale-Hubbell and has been voted by his peers as a Pennsylvania Super Lawyer and Rising Star. COST: Members: $35 Non-Members: $75 Student Members: Free Not a member? Consider joining today to receive a discount to this event and more! Membership Info Cancellation Policy Refunds must be requested in writing at least three business days before the event. Refunds will not be issued after that point. Unfortunately we will not be able to credit your registration payment toward a future event, but you may allow another person to attend in your place. Please email rogerio@csusa.org and let us know that person’s name, affiliation, city and state, and email address. Financial Aid Policy The Copyright Society provides a limited number of scholarships each year to students, law clerks, unemployed attorneys, and nonprofit and government employees. For information on how to apply, please click here. CLE CREDIT: The Copyright Society is a Certified CLE Provider in New York and California. Program will satisfy 1.0 NY CLE credit and 1.0 CA participatory credit. 1.0 PA CLE credit pending approval by the PA Bar. The intermediate program is transitional and appropriate for both newly admitted and experienced attorneys. Instructions for verifying attendance will be emailed to registrants prior to the program. CLE Credit Details The Copyright Society is a Certified CLE Provider in New York and California. Program will satisfy 1.0 NY CLE credit and 1.0 CA participatory credit. 1.0 PA CLE credit pending approval by the PA Bar. The intermediate program is transitional and appropriate for both newly admitted and experienced attorneys. Instructions for verifying attendance will be emailed to registrants prior to the program.
