Image: Thaler

Amid the rising use of artificial intelligence (“AI”) to create works of art and the legal gray areas – including nuances over whether AI art is eligible for copyright registration – that are coming with it, there is a current case that is worth keeping an eye on. In Stephen Thaler v. Shira Perlmutter and the U.S. Copyright Office (“USCO”), the plaintiff is looking to get a U.S. federal district court to set aside a refusal from the USCO Review Board to register his two-dimensional AI-generated artwork – entitled, “A Recent Entrance to Paradise” – on the basis that the author of the work was identified as Thaler’s AI system and Thaler did “not assert that the work was created with contribution from a human author.” 

In the complaint that he lodged with a federal court in Washington, D.C. in June 2022 against the USCO and Perlmutter in her role as Register of Copyrights and Director of the USCO, Thaler argues that based on the plain language of the Copyright Act, works created entirely by AI should be eligible for copyright registration. “At no point does the Act limit authorship to natural persons,” Thaler asserts in the complaint, stating that “corporations and other non-human entities have been considered ‘authors’ for purposes of the Act for over a century.”

As for the visual artwork at issue, Thaler claims that it “meets all the requirements for copyright protection.” Additionally, he contends that if he “had submitted the same AI-generated work with his company listed as the author, the USCO would have granted his company a registration, and no one would have known the work was AI-generated.” Aside from copyright-centric claims, Dorsey & Whitney’s Evan Everist notes that Thaler also advances “a number of property-based theories, including based on accession, first possession, and the work-made-for-hire doctrine.”

Characterizing the USCO’s refusal to register the work as “arbitrary, capricious, an abuse of discretion and not in accordance with the law, unsupported by substantial evidence, and in excess of [its] statutory authority,” Thaler is looking to get the court to issue an order compelling the USCO to set aside the refusal. 

Fast forward to this week and Thaler is angling for an expedited win, lodging a motion for summary judgment with the court on Monday on the basis that “no genuine issue as to any material fact exists,” and thus. He is entitled to judgment as a matter of law. Setting the stage in the Jan. 9 motion, counsel for Thaler asserts that “with the facts not in dispute” – the parties agree that artwork was entirely created using AI – “this case boils down to one novel legal question: Can someone register a copyright in a creative work made by an artificial intelligence (‘AI-Generated Work’)?” 

This should be allowed, Thaler maintains, as “the plain language of the Copyright Act does not restrict copyright to human-made works, nor does any case law.” Against this background, Thaler contends that the USCO erred in denying him a registration, “mistakenly rely[ing] on dicta, predominantly from cases predating even the existence of modern computers, together with inappropriate reliance on a technical report that pre-dates autonomously creative AI.” The USCO also got it wrong from a policy perspective, per Thaler, who claims that by failing to register the artwork, the Copyright Office is “frustrat[ing] the purpose of the Act, which is to promote the dissemination and creation of works.” On the other hand, he asserts that “allowing copyright on AI-Generated Works encourages the development and use of creative AI which results in the generation of more works and provides incentives for those works to be disseminated.”  

Reflecting on the significance of the case, Mesh IP Law’s Michael Eshaghian says that the impacts – including policy-centric ones – “will be felt far beyond the specific facts at issue” and could very well introduce “a sea-change in what is copyrightable in the first place.” If Thaler prevails, he says, “Any AI-generated work could be copyrighted, which introduces another, perhaps bigger, question: Who owns that copyright?” Thaler’s motion offers an answer to that question: Whoever owns the AI should own the copyright. However, that “seems like an unsatisfying answer,” Eshaghian says, particularly in light of the fact that absent any contractual terms stating otherwise – like DALL-E creator OpenAI’s terms, which actually assign ownership of all outputs to its users – the creator of the underlying AI would remain the owner. 

Such an approach appears as though it would “favor large entities that have hundreds of millions of dollars to spend developing advanced AI at the expense of independent creators,” according to Eshaghian, who says that this is probably a question that Congress should answer, as it is “better suited to weighing the large policy implications that this case implicates.” 

As for what we can expect in terms of an outcome for Thaler, Everist stated that “given the current state of the case law, Thaler likely faces an uphill battle, [as] courts, including the Supreme Court, have uniformly interpreted the Copyright Act to require human authorship.” More broadly, he called the case an important one in light of the fact that “AI is becoming increasingly present in our everyday lives, extending far beyond the scientific realm.” 

“As shown here,” he asserted that “AI machines can now produce creative works, including music, inventions, and artwork, and the human-authorship requirement could have negative implications” for the growth and advancement of this relatively novel technology. 

The case is Stephen Thaler v. Shira Perlmutter, et al., 1:22-cv-01564 (DDC).