skip to main content
Close Button
Last Name
First Name
Practice Area

The Newsroom

Me, Myself or AI? Who is the author under U.S. copyright law?

Me, Myself or AI? Who is the author under U.S. copyright law?

With the rise of new technology such as blockchain, the digital assets supported by it, like NFTs, and the concomitant increase in NFT litigation, we are witnessing how courts are grappling with how to apply existing laws to these new forms of technology. We saw the first ever NFT trial unfold recently where the jury considered the applicability of the First Amendment to a Hermès trademark used on the MetaBirkins NFTs, and just last week, a court found that the NBA Top Shot NFTs could constitute securities under U.S. law. But blockchain and NFTs aren’t the only game in town when it comes to new technology causing ripples in the legal world.

Artificial intelligence (“AI”) has been incrementally permeating our society, and takes center stage at the moment with the release of Chat GPT, an AI chatbot that answers questions, writes stories, makes music, and even writes code. The bot remembers previous chats with the user, which is used to contextualize its next responses, and it derives its answers from vast amounts of information found on the Internet. I could have asked Chat GPT to write this article instead of writing it myself, and no one would know the difference.

But if Chat GPT can write articles and songs, who owns the copyright in the output? The US Copyright Act protects “original works of authorship fixed in any tangible medium of expression.” 17 U.S.C. § 102. And copyright “vests initially in the authors of the work.” 17 U.S.C. § 201(a). Can AI technology be deemed the author within the meaning of this statute?

This precise question is the one being challenged in Stephen Thaler v. Shira Perlmutter et al., a recently-filed federal action pending in the US District Court for Washington D.C. against the U.S. Copyright Office and Shira Perlmutter, the Register of Copyrights. The plaintiff, Dr. Thaler, alleges that he created an AI program, which created a two-dimensional artwork, and he attempted to copyright this artwork by naming his AI as the author. He claimed ownership of the copyright as the owner of the AI under the work-for-hire doctrine.

The U.S. Copyright Office rejected the application on the basis that it lacked human authorship, which it found to be a necessary requirement to support a copyright claim. Dr. Thaler appealed the denial to the Copyright Office Review Board (“Board”), arguing that the human authorship requirement was unconstitutional and unsupported by statute or case law. The Board, in refusing Dr. Thaler’s second request for reconsideration, observed that the work at issue contained no human contribution (according to Dr. Thaler’s submission, the work was created autonomously by a computer algorithm running on a machine), and thus was distinguishable from a situation where a work was one of human authorship with the use of computer assistance, where registration could theoretically be possible.

In explaining its decision, the Board cited the Compendium of U.S. Copyright Office Practices (Third Edition), a practice manual used by the Copyright Office, which states that an original work of authorship requires that it be “created by a human being.” The manual provides that the law only protects “‘the fruits of intellectual labor . . . founded in the creative powers of the mind,’” and it “will not register works produced by a machine or mere mechanical process that operates randomly or automatically without any creative input or intervention from a human author.” The Board also reviewed the Copyright Act’s text, legislative history, as well as case law to decide that copyright protection was limited to human authors.

Dissatisfied with the outcome, Dr. Thaler filed his court action seeking to set aside the Copyright Office’s refusal to register, alleging that its decision was arbitrary, capricious, and contrary to law, and requests the court to grant the registration of his work. The parties have cross-moved for summary judgment, with each side raising competing arguments on a variety of legal issues, with the main dispute over why AI can or cannot be considered to be an author. On Dr. Thaler’s side, he argues that nothing in the text of the Copyright Act actually states that an author must be a human being. The Copyright Office argues that the 1909 Copyright Act (i.e., the predecessor to the current 1976 Copyright Act) expressly provided that only a “person” could “secure copyright for his work,” and that Congress did not intend to change this requirement by enacting the 1976 Act. It also argues that the 1976 Copyright Act was structured based on the assumption that an author was human.

The parties are also diametrically opposed on their reading of the case law. Dr. Thaler asserts that all of the cases cited by the Copyright Office suggesting a human authorship requirement are dicta, some of which were rendered in an era before computers were invented, and that no case actually prohibits copyright protection of AI-generated work. The Copyright Office disagrees, starting with the venerable Burrow-Giles Lithographic Co. v. Sarony, 111 U.S. 53 (1884), where it was first held that photographs constituted “writings” of an author under the Constitution’s copyright clause. The Copyright Office argues that Burrow-Giles buttresses its position since there the court only found photographs protectable because they reflected creative human choices, such as the posing, lighting, and other directions by the photographer that created the image. Dr. Thaler argues that Burrow-Giles actually helps his cause by expanding the scope of copyright to include photographs, and argues that old statutes must be interpreted in the light of technological advancements and the Copyright Act’s purpose of encouraging the creation and dissemination of works. To this, the Copyright Office responds, Dr. Thaler makes nothing more than policy arguments in favor of copyright protection for AI works which is inadequate to demonstrate that the Copyright Office was arbitrary or capricious, and even if these policy interests exist, the policy is to encourage humans to create and disseminate works, not AI.

Needless to say, the court’s determination of these issues could have potentially far-reaching consequences if AI can be considered to be an author under U.S. copyright laws. Moreover, with respect to AI technology such as Chat GPT, to the extent its output is based upon pre-existing copyrighted content of others, query what copyright rights could attach to that output and whether such output can be said to be “original.” And even if AI could be considered to be an author, how that authorship ties in with the human or entity seeking to claim ownership of the copyright is another hurdle, which is also an issue where the parties in Thaler diverge considerably.

An argument in reverse to Dr. Thaler’s position is currently being advanced by Ryder Ripps in connection with one of his counterclaims in his action against Yuga Labs. Ripps and his co-defendant seek a declaratory judgment that Yuga holds no valid copyright because “the BAYC Images hav[e] been generated by an automated computer algorithm where no humans were involved in determining which of the 10,000 BAYC Images were selected from the more than 1.3 billion possible permutations (except perhaps with respect to a few custom BAYC Images that Yuga may have produced with human involvement).” Yuga Labs has moved to strike the counterclaim and argues that there is no need to reach this question because it never sued defendants for copyright infringement. Yuga admits that it has no copyright registrations on the Bored Ape images, but asserts it does not need one to own the copyright in the images. Without a registration, Yuga cannot sue for infringement, and accordingly, Yuga argues that there can be no case or controversy, which is required to warrant declaratory relief.

With the proliferation of AI and other automated technology, we can expect to see more cases that will consider the impact of AI and other technology-driven works on copyright laws. Indeed, the Copyright Office indicates in its papers in Thaler that it will be addressing these issues in the coming year, including “preparing registration guidance for works generated by using AI, planning public events to discuss emerging issues, and taking steps to issue a notice of inquiry on complex questions involving copyright and AI.”

Stay tuned for more from Ingram's NFT Newsroom, and follow us on Twitter and LinkedIn.

By: Mioko C. Tajika