Generative AI Meets Copyright Law

On Wednesday, April 26, Pamela Samuelson, Richard M. Sherman Distinguished Professor of Law at UC Berkeley, delivered the final of four Distinguished Lectures on the Status and Future of AI, co-hosted by CITRIS Research Exchange and the Berkeley Artificial Intelligence Research Group (BAIR).

Samuelson’s talk explores a particularly controversial topic in the legal community: whether the texts and images generated by artificial intelligence (AI) should be protected under copyright law. 

Copyright, Samuelson noted, automatically protects works of authorship, such as written content, illustrations, photographs, videos and musical compositions, from the moment they are fixed in a tangible medium, and it vests in those authors certain exclusive rights, including the ability to control reproductions and displays. However, copyright has limitations, including the fair use doctrine, which allows the unlicensed use of copyrighted works in criticism, news reporting, research and other specific circumstances.

Samuelson focuses on three principal questions at the intersection of copyright and AI: Is the act of ingesting in-copyright works as training data for AI programs copyright infringement? Are the AI-generated outputs derivative works that infringe on the copyright of the ingested images? And who, if anyone, owns the copyright of those AI-generated images? 

“Copyright law is the only law that’s already in existence that could bring generative AI systems to their knees. If the court says ingesting [is] infringement, the whole thing can be destroyed. Copyright law is an existential threat to progress in this field,” Samuelson said.  

“If you think [generative AI] should continue to thrive,” she said, “help to build arguments that it will actually advance the purposes of copyright, rather than destroy it.”