Court of Appeals Affirms Registration Refusal for AI-Generated Output
In 2019, Stephen Thaler developed an AI system he called The Creativity Machine. He generated output he called A Recent Entrance to Paradise. When he applied to register a copyright claim in the output, he listed the machine as the author. He claimed ownership of the work as a work made for hire. In his application, he asserted that the work was autonomously created by a machine. The Copyright Office denied the claim on the basis that human authorship is a required element of a copyright claim.
On appeal, the United States district court affirmed the Copyright Office’s decision. Thaler attempted to argue, for the first time, that it was copyrightable because he provided instructions and directed the machine’s creation of the work. The district court found that he had waived that argument.
The Court of Appeals Affirms
Thaler sought review in the Court of Appeals for the Federal Circuit. On March 18, 2025, the Court of Appeals affirmed. The Court cited language in the Copyright Act that suggested Congress intended only human beings to be authors. The Court did not reach the question whether the Copyright Clause of the U.S. Constitution might protect machine-generated works if Congress should choose someday to extend copyright protection to these kinds of materials.
The Court held that the question whether Thaler could claim authorship on the basis of the fact that he made and directed the operation of the Creativity Machine has not been preserved for appeal.
Copyright registrations are being issued for works created with generative-AI tools, subject to some important qualifications. Also, Internet Archves revisited (briefly)
The long-awaited U.S. Copyright Office report on the copyrightability of works created using AI-generated output, is here. The legality of using copyrighted works to train generative-AI systems is a topic for another day.
Key takeaways:
Copyright protects the elements of a work that are created by a human, but does not protect elements that were AI-generated (probably the key take-away from the Report) The is the “human authorship” requirement that the Copyight Office invoked in denying registration of Stephen Thaler’s AI-generated output. I wrote about that a couple of years ago in “AI Can Create But Is It Art?” and also have commented on new AI copyright guidance from the Office before.
The Copyright Office believes existing law is adequate to deal with AI copyright issues; it does not believe any new legislation is needed
Using AI to assist in the creative process does not affect copyrightability
Prompts do not provide sufficient control over the output to be considered creative works.
Protection exists for the following, if they involve sufficient human creativity:
Selection, coordination, and arrangement of AI-generated output (compilation)
Modification of AI-generated content (Derivative works)
Human-created elements distinguishable from AI-generated elements.
Prompts
A key question for the Copyright Office was whether a highly detailed prompt could suffice as human creative expression. The Office says no; “[P]rompts alone do not provide sufficient human control to make users of an AI system the authors of the output. Prompts essentially function as instructions that convey unprotectable ideas. While highly detailed prompts could contain the user’s desired expressive elements, at present they do not control how the AI system processes them in generating the output.”
How much control does a human need over the output-generation process to be considered an author? The answer, apparently, is “So much control that the AI mechanism’s contribution was purely rote or mechanical. “The fact that identical prompts can generate multiple different outputs further indicates a lack of human control.”
Expressive prompts
If the prompt itself is sufficiently creative and original, the expression contained in the prompt may qualify for copyright protection. For example, if a user prompts an AI tool to change a story from first-person to third-person point of view, and includes the first-person version in the prompt, then copyright may be claimed in the story that was included in the prompt. The author could claim copyright in the story as a “human-generated element” distinguishable from anything AI thereafter did to it. The human-created work must be perceptible in the output.
Registration of hybrid works
The U.S. Copyright Office has now issued several registrations for works that contain a combination of both human creative expression and AI-generated output. Examples:
Irontic, LLC has a registered copyright in Senzia Opera a, a sound recording with “music and singing voices by [sic] generated by artificial intelligence,” according to the copyright registration. That material is excluded from the claim. The registration, however, does provide protection for the story, lyrics, spoken words, and the selection, coordination, and arrangement of the sound recording.
Computer programs can be protected by copyright, but if any source code was generated by AI, it must be excluded from the claim. Thus, the Adobe GenStudio for Performance Marketing computer program is protected by copyright, but any source code in it that was AI-generated is not.
A record company received a copyright registration for human additions and modifications to AI-generated art.
As an example of a “selection, coordination and arrangement” copyright, there is the registration of a work called “A Collection of Objects Which Do Not Exist,” consisting of a collage of AI-generated images. “A Single Piece of American Cheese,” is another example of a registered copyright claim based on the selection, coordination, or arrangement of AI-generated elements.
China
A Chinese court has taken a contrary position, holding that an AI-generated image produced by Stable Diffusion is copyrightable because the prompts he chose reflected his aesthetic choices.
Internet Archive Postscript
In January, the Second Circuit Court of Appeals affirmed the decision in Hachette Book Group, Inc. v. Internet Archive. This came as no surprise. A couple of important things that bear repeating came out of this decision, though.
First, the Court of Appeals reaffirmed that fair use is an affirmative defense. As such, the defendant bears the burden of establishing the level of market harm the use has caused or may cause. While a copyright owner may reasonably be required to identify relevant markets, he/she/it is not required to present empirical data to support a claim of market harm. The defendant bears the burden of proof of a fair use defense, including proof pertinent to each of the four factors comprising the defense.
Confusion seems to have crept into some attorneys’ and judges’ analysis of the issue. This is probably because it is well known that the plaintiff bears the burden of proof of damages, which can also involve evidence of market harm. The question of damages, however, is separate and distinct from the “market harm” element of a fair use defense.
The second important point the Second Circuit made in Hatchette is that the “public benefit” balancing that Justice Breyer performed in Google LLC v. Oracle America, Inc. needs to focus on something more than just the short-term benefits to the public in getting free access to infringing copies of works. Otherwise, the “public benefit” in getting free copies of copyright-protected stuff would outweigh the rights of copyright owners every time. The long-term benefits of protecting the rights of authors must also be considered.
True, libraries and consumers may reap some short-term benefits from access to free digital books, but what are the long-term consequences? [Those consequences, i.e.,] depriv[ing] publishers and authors of the revenues due to them as compensation for their unique creations [outweigh any public benefit in having free access to copyrighted works.]
A status update on 24 pending lawsuits against AI companies – what they’re about and what is happening in court – prepared by Minnesota copyright attorney Thomas James.
Advancements in artificial intelligence technology, including generative-AI, have introduced a wide range of new or exacerbated legal problems. Collectively, I call these AI legal issues. Although not all of them are unique to scenarios involving AI, they are certainly testing and stretching the capacity of legal institutions. Here is a very brief summary of how these issues are playing out in the courts, as of February 28, 2024.
Copyright
Thaler v. Perlmutter (D.D.C. 2022).
Complaint filed June 2, 2022. Thaler created an AI system called the Creativity Machine. He applied to register copyrights in the output he generated with it. The Copyright Office refused registration on the ground that AI output does not meet the “human authorship” requirement. (I explained that requirement in a previous blog post that explored the difference between human and AI creation of a work. He then sought judicial review. The district court granted summary judgment for the Copyright Office. (SeeA Recent Exit from Paradise.) In October, 2023, Thaler filed an appeal to the District of Columbia Circuit Court of Appeals (Case no. 23-5233).
Doe v. GitHub, Microsoft, and OpenAI (N.D. Cal. 2022)
I wrote about this case in Generative AI: The Top 12 Lawsuits. The complaint was filed November 3, 2022. Software developers claim the defendants trained Codex and Copilot on code derived from theirs, which they published on GitHub. Some claims have been dismissed, but claims that GitHub and OpenAI violated the DMCA and breached open source licenses remain. Discovery is ongoing.
Andersen v. Stability AI (N.D. Cal. 2023)
The Andersen v. Stability AI complaint was filed January 13, 1023. Visual artists sued Midjourney, Stability AI and DeviantArt for copyright infringement for allegedly training their generative-AI models on images scraped from the Internet without copyright holders’ permission. Other claims included DMCA violations, publicity rights violations, unfair competition, breach of contract, and a claim that output images are infringing derivative works. On October 30, 2023, the court largely granted motions to dismiss, but granted leave to amend the complaint. Plaintiffs filed an amended complaint on November 29, 2023. Defendants have filed motions to dismiss the amended complaint. Hearing on the motion is set for May 8, 2024.
Getty Images v. StabilityAI (U.K. 2023)
Complaint filed January, 2023. Getty Images claims StabilityAI scraped images without its consent. The Getty Images lawsuit has survived a motion to dismiss. The case appears to be heading to trial.
In re OpenAI ChatGPT Litigation (N.D. Cal. 2023)
Complaint filed June 28, 3023. Originally captioned Tremblay v. OpenAI. Book authors sued OpenAI for direct and vicarious copyright infringement, DMCA violations, unfair competition and negligence. Both input (training) and output (derivative works) claims are alleged, as well as state law claims of unfair competition, etc.
Most state law and DMCA claims have been dismissed, but claims based on unauthorized copying during the AI training process remain. An amended complaint is likely to come in March. The court has directed the amended complaint to consolidate Tremblay v. OpenAI, Chabon v. OpenAI, and Silverman v. OpenAI.
Kadrey v. Meta (N.D. Cal. 2023)
Complaint filed July 7, 2023. Sarah Silverman and other authors allege Meta infringed copyrights in their works by making copies of them while training Meta’s AI model; that the AI model is itself an infringing derivative work; and that outputs are infringing copies of their works. Plaintiffs also allege DMCA violations, unfair competition, unjust enrichment, and negligence. The court granted Meta’s motion to dismiss all claims except the claim that unauthorized copies were made during the AI training process. An amended complaint and answer have been filed.
In 2025, Judge Chhabria ruled in Meta’s favor on fair use with respect to AI training; reserved the motion for summary judgment on the DMCA claims for decision in a separate order, and held that the claim of infringing distribution via leeching or seeding “will remain a live issue in the case.” Kadrey v. Meta Platforms.
J.L. v. Google (N.D. Cal. 2023)
Complaint filed July 11, 2023. In another case I mentioned in Generative AI: The Top 12 Lawsuits, an author filed a complaint against Google alleging misuse of content posted on social media and Google platforms to train Google’s AI Bard. (Gemini is the successor to Google’s Bard.) Claims include copyright infringement, DMCA violations, and others. J.L. filed an amended complaint and Google has filed a motion to dismiss it. A hearing is scheduled for May 16, 2024.
Chabon v. OpenAI (N.D. Cal. 2023)
Complaint filed September 9, 2023. Authors allege that OpenAI infringed copyrights while training ChatGPT, and that ChatGPT is itself an unauthorized derivative work. They also assert claims of DMCA violations, unfair competition, negligence and unjust enrichment. Chabon v. OpenAI has been consolidated with Tremblay v. OpenAI, and the cases are now captioned In re OpenAI ChatGPT Litigation.
Chabon v. Meta Platforms (N.D. Cal. 2023)
Complaint filed September 12, 2023. Authors assert copyright infringement claims against Meta, alleging that Meta trained its AI using their works and that the AI model itself is an unauthorized derivative work. The authors also assert claims for DMCA violations, unfair competition, negligence, and unjust enrichment. In November, 2023, the court issued an Order dismissing all claims except the claim of unauthorized copying in the course of training the AI. The court described the claim that an AI model trained on a work is a derivative of that work as “nonsensical.” Chabon v. Mea Platforms.
Authors Guild v. OpenAI, Microsoft, et al. (S.D.N.Y. 2023)
Complaint filed September 19, 1023. Book and fiction writers filed a complaint for copyright infringement in connection with defendants’ training AI on copies of their works without permission. A motion to dismiss has been filed. Authors Guild v. Open AI et al.
Huckabee v. Bloomberg, Meta Platforms, Microsoft, and EleutherAI Institute (S.D.N.Y. 2023)
Complaint filed October 17, 2023. Political figure Mike Huckabee and others allege that the defendants trained AI tools on their works without permission when they used Books3, a text dataset compiled by developers; that their tools are themselves unauthorized derivative works; and that every output of their tools is an infringing derivative work. Claims against EleutherAI have been voluntarily dismissed. Claims against Meta and Microsoft have been transferred to the Northern District of California. Bloomberg is expected to file a motion to dismiss soon. Huckabee v. Bloomberg et al.
Huckabee v. Meta Platforms and Microsoft (N.D. Cal. 2023)
Complaint filed October 17, 2023. Political figure Mike Huckabee and others allege that the defendants trained AI tools on their works without permission when they used Books3, a text dataset compiled by developers; that their tools are themselves unauthorized derivative works; and that every output of their tools is an infringing derivative work. Plaintiffs have filed an amended complaint. Plaintiffs have stipulated to dismissal of claims against Microsoft without prejudice. Huckabee v. Meta Platforms and Microsoft.
Concord Music Group v. Anthropic (M.D. Tenn. 2023)
Complaint filed October 18, 2023. Music publishers claim that Anthropic infringed publisher-owned copyrights in song lyrics when they allegedly were copied as part of an AI training process (Claude) and when lyrics were reproduced and distributed in response to prompts. They have also made claims of contributory and vicarious infringement. Motions to dismiss and for a preliminary injunction are pending. Concord Music Group v. Anthropic.
Alter v. OpenAI and Microsoft (S.D.N.Y. 2023)
Complaint filed November 21, 2023. Nonfiction author alleges claims of copyright infringement and contributory copyright infringement against OpenAI and Microsoft, alleging that reproducing copies of their works in datasets used to train AI infringed copyrights. The court has ordered consolidation of Author’s Guild (23-cv-8292) and Alter (23-cv-10211). On February 12,2024, plaintiffs in other cases filed a motion to intervene and dismiss. Alter v. OpenAI and Microsoft.
New York Times v. Microsoft and OpenAI (S.D.N.Y. 2023)
Complaint filed December 27, 2023. The New York Times alleges that their news stories were used to train AI without a license or permission, in violation of their exclusive rights of reproduction and public display, as copyright owners. The complaint also alleges vicarious and contributory copyright infringement, DMCA violations, unfair competition, and trademark dilution. The Times seeks damages, an injunction against further infringing conduct, and a Section 503(b) order for the destruction of “all GPT or other LLM models and training sets that incorporate Times Works.” On February 23, 2024, plaintiffs in other cases filed a motion to intervene and dismiss this case. New York Times v. Microsoft and OpenAI.
Basbanes and Ngagoyeanes v. Microsoft and OpenAI (S.D.N.Y. 2024)
Getty Images v. Stability AI Complaint filed on February 3, 2023. Getty Images alleges claims of copyright infringement, DMCA violation and trademark violations against Stability AI. The judge has dismissed without prejudice a motion to dismiss or transfer on jurisdictional grounds. The motion may be re-filed after the conclusion of jurisdictional discovery, which is ongoing.
Privacy and Publicity Rights
Flora v. Prisma Labs (N.D. Cal.)
Complaint filed February 15, 2023. Plaintiffs allege violations of the Illinois Biometric Privacy Act in connection with Prisma Labs’ collection and retention of users’ selfies in AI training. The court has granted Prisma’s motion to compel arbitration. Flora v. Prisma Labs.
Kyland Young v. NeoCortext (C.D. Cal. 2023)
Complaint filed April 3, 2023. This complaint alleges that AI tool Reface used a person’s image without consent, in violation of the person’s publicity rights under California law. The court has denied a motion to dismiss, ruling that publicity rights claims are not preempted by federal copyright law. The case has been stayed pending appeal. Kyland Young v. NeoCortext.
Complaint filed June 28, 2023. Users claim OpenAI violated the federal Electronic Communications Privacy Act and California wiretapping laws by collecting their data when they input content into ChatGPT. They also claim violations of the Computer Fraud and Abuse Act. Plaintiffs voluntarily dismissed the case on September 15, 2023. See now A.T. v. OpenAI (N.D. Cal. 2023) (below). P.M. v. OpenAI.
A.T. v. OpenAI (N.D. Cal. 2023)
Complaint filed September 5, 2023. ChatGPT users claim the company violated the federal Electronic Communications Privacy Act, the Computer Fraud and Abuse Act, and California Penal Code section 631 (wiretapping). The gravamen of the complaint is that ChatGPT allegedly accessed users’ platform access and intercepted their private information without their knowledge or consent. Motions to dismiss and to compel arbitration are pending. A.T. v. OpenAI.
Defamation
Walters v. OpenAI (Gwinnett County Super. Ct. 2023), and Walters v. OpenAI (N.D. Ga. 2023)
Gwinnett County complaint filed June 5, 2023.
Federal district court complaint filed July 14, 2023.
Radio Radio talk show host sued OpenAI for defamation. A reporter had used ChatGPT to get information about him. ChatGPT wrongly described him as a person who had been accused of fraud. In October, 2023, the federal court remanded the case to the Superior Court of Gwinnett County, Georgia. On January 11, 2024, the Gwinnett County Superior Court denied OpenAI’s motion to dismiss. Walters v. OpenAI.
Battle v. Microsoft (D. Md. 2023)
Complaint filed July 7, 2023. Pro se defamation complaint against Microsoft alleging that Bing falsely described him as a member of the “Portland Seven,” a group of Americans who tried to join the Taliban after 9/11. Battle v. Microsoft.
Caveat
This list is not exhaustive. There may be other cases involving AI that are not included here. For a discussion of bias issues in Google’s Gemini, have a look at Scraping Bias.
