Meta Beats Authors’ Copyright Suit Over AI Training on Books (2)

Meta Platforms Inc. escaped a first-of-its-kind copyright lawsuit from a group of authors who alleged the tech giant hoovered up millions of copyrighted books without permission to train its generative AI model called Llama.

San Francisco federal Judge Vince Chhabria ruled Wednesday that Meta’s decision to use the books for training is protected under copyright law’s fair use defense, but he cautioned that his opinion is more a reflection on the authors’ failure to litigate the case effectively.

“This ruling does not stand for the proposition that Meta’s use of copyrighted materials to train its language models is lawful,” Chhabria said. “It stands only for the proposition that these plaintiffs made the wrong arguments and failed to develop a record in support of the right one.”

Boies Schiller Flexner LLP, the authors’ firm, said in a statement that “despite the undisputed record of Meta’s historically unprecedented pirating of copyrighted works, the court ruled in Meta’s favor. We respectfully disagree with that conclusion.”

A Meta spokesperson said in a statement that “fair use of copyright material is a vital legal framework for building this transformative technology.”

The ruling comes two days after a judge in the same court ruled largely in favor of AI competitor Anthropic PBC, which was similarly accused of illegally using books to train its models.

Chhabria’s opinion, the second-ever federal court ruling to weigh in on the use of copyrighted works to train today’s most powerful generative AI models, appeared to reject the idea that the industry can use billions of human-made works without seeking permission.

“It’s hard to imagine that it can be fair use to use copyrighted books to develop a tool to make billions or trillions of dollars while enabling the creation of a potentially endless stream of competing works that could significantly harm the market for those books,” he said.

The judge also rejected Meta’s argument that an adverse ruling would harm AI technology growth by requiring companies to pay for copyrighted training data, saying that theory “doesn’t pass the straight face test.”

“These products are expected to generate billions, even trillions, of dollars for the companies that are developing them,” Chhabria said. “If using copyrighted works to train the models is as necessary as the companies say, they will figure out a way to compensate copyright holders for it.”

Sweeping Implications

The back-to-back rulings in favor of Meta and Anthropic will have major ramifications for dozens of similar lawsuits across the country against top AI companies, including OpenAI Inc. and Stability AI Ltd., as well as the growing data licensing market for AI models.

The proposed class of authors including comedian Sarah Silverman and Pulitzer Prize-winning novelist Andrew Sean Greer first sued Meta in the US District Court for the Northern District of California in 2023, alleging the tech giant downloaded their books from notorious digital piracy websites in an effort to catch up with competitors in AI development.

The case was followed by similar infringement suits brought by media organizations, other authors, music publishers, and visual artists.

The AI industry has long argued that training models falls under the fair use exception to the law, which allows unauthorized copying in certain cases where the copying “transforms” the original work.

Although both Meta and Anthropic won their fair use arguments in their respective cases, their separate rulings reveal a divide over how courts could interpret the legal test.

Judge William Alsup, who was overseeing the Anthropic case, ruled on Monday that the company’s use of works to train its AI model called Claude is “among the most transformative many of us will see in our lifetimes.” Chhabria’s ruling rebuffed Alsup’s analysis, saying that the earlier opinion brushed aside the harm that generative AI models will have on the market for the original books.

But Chhabria said his decision to find in favor of Meta came down to the authors’ lack of any evidence that AI chatbots were undermining their book sales. “All the plaintiffs presented is speculation, and speculation is insufficient,” he said.

Joseph Saveri Law Firm LLP, DiCello Levitt LLP, Lieff Cabraser Heimann & Bernstein LLP, and Cafferty Clobes Meriwether & Sprengel LLP also represent the plaintiffs.

Cooley LLP, Cleary Gottlieb Steen & Hamilton LLP, and Paul, Weiss, Rifkind, Wharton & Garrison LLP represent Meta.

The case is Kadrey v. Meta Platforms Inc., N.D. Cal., No. 23-cv-03417, 6/25/25.