In the relatively recent history of AI development, a specific niche called generative AI is now setting the stage for vital policy and legal discussions. These systems, while not entirely novel, are being implemented on a new scale that is quickly starting to impact legal contexts in a significant way. AI technology has been under development for decades, analogous to other technological advances such as early color cinema and the internet (ref. CNN’s 1999 assertion “the internet changes everything”).
Increasingly, generative AI’s potential legal implications are becoming tangible, with court dockets this year seeing numerous copyright infringement complaints associated with generative AI models. Some companies targeted by these lawsuits include Stability AI and Meta Platforms Inc.
Within these disputes are mixed opinions from district court judges regarding whether products of generative AI models can qualify as derivative works, although claims of direct infringement during AI model training are being allowed to proceed.
Meanwhile, larger reaching discussions are underway on the broader question of whether AI model outputs are eligible for intellectual property protection. Currently, the Copyright Office and Patent Office assert that AI models are not designated as authors or inventors. Yet, the debates endure on who or which entity should be credited for any expressive content generated by AI, whether that is the party requesting the output or the one developing and training the model.
Whilst AI cases dominate the legal landscape, the intellectual property (IP) law territory is also seeing activity outside the realm of technology. The US Supreme Court has taken on two IP cases this term, Vidal v. Elster and Warner Chappell Music v. Nealy. The Supreme Court examines the denial of registration for controversial trademark “Trump Too Small” in Vidal v. Elster, and it clarifies the damages period regarding copyright infringement cases in Warner Chappell Music v. Nealy.
Beyond these immediate cases, ongoing discussions on patent eligibility continue as the legal community struggles with the Section 101 precedent. We are witnessing continuous denials of cert petitions – recently in CareDX v. Natera – and ongoing patent reform underway. The Patent Eligibility Restoration Act of 2023, introduced by Sens. Thom Tillis (R-N.C.) and Chris Coons (D-Del.), may offer some clarification on these pressing patent issues if passed.