By Stephanie Drawdy
How successful might an AI/IP treaty be? How does existing law stand up in the face of efforts to mold it to emerging technologies? A robust debate was had on these and other AI/IP queries during a recent panel discussion on Warfare of Art and Law podcast. The six-member panel was comprised of professionals from across the globe with backgrounds in the arts, law, and technology, who highlighted a range of concerns and insights. The recording of that discussion is now live on the podcast, and below are just a few of the compelling points raised.
Dr. Andres Guadamuz, a Reader in IP Law at University of Sussex and Editor-In-Chief of the Journal of World Intellectual Property, favorably viewed the creation of a global AI-related treaty, citing the need to harmonize certain copyright concepts like originality. Yet, mindful of “the current stage of international IP politics”, he opined that this would be a “very slow” process while “a few months in AI is a decade”.
Diverse opinions emerged on whether current law was sufficient to handle the input and output issues raised by Generative AI (GAI). India-based IP and technology laws attorney Ankit Sahni shared his experience of co-creating the GAI work "Suryast" with an AI style transfer tool (RAGHAV) and his thoughts on being denied U.S. copyright registration while receiving registration in India.
Sahni suggested a sui generis system of registration that might offer a limited scope of protection to GAI works. RightsClick co-founder David Newhoff remained unconvinced that a sui generis right was necessary. Likewise, Institute of Art & Law Assistant Director Emily Gould suggested that it would be “quite a tough one to get right”, citing the likely demise of Section 9(3) of the UK Copyright, Designs and Patents Act 1988 that addresses protection for GAI works.
Artist & Attorney Stefania Salles Bruins questioned whether the purpose of IP rights - to encourage creativity and innovation by allowing creators and inventors to profit from their work – was still being upheld. This concern is certainly a valid one in light of opinions expressed by developers like Meta that model training as “a quintessentially non-exploitative use of training material” that “does not implicate any of the legitimate rightsholder interests that copyright law exists to protect” (10 October 2023 Comments of Meta Platforms, Inc. to USCO’s Notice of Inquiry on AI and Copyright at 11).
Looking to the future, Kritika Sahni, an IP Attorney and Partner at India-based Ajay Sahni & Associates, advocated for “the right balance” between encouragement of tech and reward for human creativity to allow both to successfully co-exist. Guadamuz believed “potential licensing agreements” facilitated through “collective negotiation bodies” would emerge. Gould agreed and foresaw dialogue between developer and creator communities. Not Newhoff who harkened back to the Napster days of “mass piracy” to note that the creative community is still waiting for meaningful dialogue.
Watch for similar panels in the future on Warfare of Art and Law as AI/IP events unfold. From the USCO’s recent registration of “A Single Piece of American Cheese” to the rejection of the fair use defense in a non-GAI case (Thomson Reuters, et al., v. Ross Intelligence Inc.) and the many developments in pending GAI suits, there remains much to parse out.