Stand-Alone AI-Generated Content Is Not Copyrightable

On August 18, 2023, the U.S. District Court for the District of Columbia denied Dr. Stephen Thaler’s motion and granted the U.S. Copyright Office’s cross motion to dismiss Thaler’s complaint. The facts of Thaler’s struggle to overcome the Copyright Office’s Human Authorship Requirement and register copyright in an AI-generated work are recounted here.

As anticipated, the Court sided with the Copyright Office, concluding that the plain text of the Copyright Act supports the Copyright Office’s Human Authorship Requirement. In doing so, the Court looked to the Supreme Court’s nearly 150-year-old decision in Burrows v. Giles, 111 U.S. 53, 56 (1884) and various decisions denying copyright protection to non-human authors, including “monkey selfies,” books allegedly authored by spirits, and nature-grown gardens. The Court was unpersuaded by Thaler’s fallback arguments under the work made for hire doctrine and common law property principles, noting that such arguments “concern to whom a valid copyright should have been registered, and in so doing put the cart before the horse.” The Court also found support for the Human Authorship Requirement in the Copyright Act’s discussion of the work made for hire doctrine, which is explicitly limited to works created by employees (who are necessarily human) and its various provisions that use personal pronouns.

Regarding Thaler’s argument that he authored the work insofar as he “provided instructions and directed his AI to create the Work,” that “the AI is entirely controlled by [him],” and that “the AI only operates at [his] direction” the court found that Thaler was simply too late. Faced with the narrow question of whether the Office acted “arbitrary and capriciously,” the Court declined to credit these arguments, which were absent from the administrative record. The Court found that the administrative record demonstrated that Thaler “played no role in using the AI to generate the work.” While the Court acknowledged that, depending on Thaler’s level of involvement, Thaler could have been the author of the work, that was ultimately inconsistent with his application to register a work he claimed was authored by the Creativity Machine.

Notably absent from the Court’s decision (and the parties’ arguments) was the Office’s 1984 issuance of a registration for The Policeman’s Beard is Half-Constructed, an AI-generated work authored by the AI system known as “Racter,” which seemingly contradicts the Human Authorship Requirement. Although not answering that important question, the Court raised other questions around AI-generated content, noting that we are undoubtedly approaching “new frontiers in copyright as artists put AI in their toolbox to be in the generation of new visual and other artistic works.” In particular, the Court asks:

  1. How much human input is necessary to qualify the user of an AI system as an “author” of a generated work, the scope of the protection obtained over the resultant image?
  2. How to assess the originality of AI-generated works where the systems may have been trained on unknown pre-existing works?
  3. How copyright might best be used to incentivize creative works involving AI?

While Thaler could not have these questions answered, it’s only a matter of time before another trailblazer swoops in, promising even more guidance ahead.


Who (If Anyone) Owns AI-Generated Content?