Anthropic trained Claude with at least 10 Stanford professors’ pirated books

A federal judge ruled in a May 2026 lawsuit that AI company Anthropic illegally copied books by more than 10 Stanford professors to train its Claude AI models.

Anthropic had already trained Claude using texts obtained illegally from professors Tobias Wolff, Abraham Verghese, Chang-Rae Lee, Daniel Mason, Hayagreeva Rao and Myra Strober. In Bartz v. Anthropic, which the Authors Alliance calls the largest copyright settlement involving an AI company, Judge William Alsup ruled that Anthropic could no longer illegally copy books to build a permanent central library.

Although Alsup stopped Anthropic from illegally copying books to train AI, he said buying books legally and using them to train Claude's large language models (LLMs) could count as fair use. He compared AI training to the way people read books, learn from them and remember their themes and writing styles. “Like any reader aspiring to be a writer, Anthropic’s LLMs trained upon works not to race ahead and replicate or supplant them — but to turn a hard corner and create something different,” he wrote.

The $1.5 billion settlement paid authors and publishers for nearly 500,000 registered works. Wolff and 27 other authors say the settlement does not provide enough compensation.

“To put me to work that way, without remittance, is a form of slave labor,” said Wolff, a retired English professor and winner of the National Medal of Arts. “They’re using my work without paying for it. As time goes on, they’re going to make the lives of writers very difficult if there’s no fruit to their labor.”

Ten of Wolff's works had been used to train AI without his permission.

“[The settlement] still gives Anthropic pretty free rein to roam through our literature and basically take what they think will be useful and profitable for them,” Wolff said.

Wolff and Chang-Rae Lee, an English professor, chose not to take part in the Bartz settlement. In May 2026, they filed a separate lawsuit, Cruz v. Anthropic, and asked for a jury trial. They are seeking the full legal damages of up to $150,000 for each work whose copyright was violated. This is much more than the roughly $3,000 per work offered in the Bartz settlement.

As in Bartz, the plaintiffs say Anthropic illegally downloaded, copied and shared their copyrighted books from online collections of pirated books, such as LibGen and PiLiMi. They also say Anthropic broke the law by downloading and keeping the pirated files in a central library.

“Our agents didn’t think the blanket lawsuit covered quite enough of the liability to us and are seeking redress in more particular ways,” Wolff said.

If Wolff and the other authors win, they could get a court order permanently stopping Claude's training from using their works.

Paul Goldstein, a leading Stanford Law scholar on intellectual property and one of the plaintiffs represented in the class-action lawsuit, said it is difficult to deal with copyright problems and prevent them at AI companies.

“The essential problem … is what economists call ‘transaction costs,’” Goldstein wrote in an email to The Daily. “Millions of individual copyrighted works may be copied in the course of any training incident, and it’s simply impracticable for the platform to obtain licenses from the copyright owner of every — or, indeed, any — single work.”

Beyond Bartz and Cruz, authors, visual artists and news organizations, including The New York Times and Getty Images, have filed more than 70 lawsuits claiming that AI companies violated their copyrights. Bartz set a current legal rule about illegally copying books, but it is still unclear how the law will treat the large-scale copying of material to train LLMs.

The Authors Guild is also bringing other lawsuits, including one against OpenAI and Microsoft. It is using what its chief executive, Mary Rasenberger, calls a “multi-pronged approach to ensure that the law is clear that AI companies must license the books they use.”

“Licensing allows authors and publishers to be compensated and to require the use of filters that ensure their works do not appear in outputs unless separately licensed,” Rasenberger wrote in an email to The Daily.

The Authors Guild is working with Congress to create a national law that would require companies to get licenses.

“The Guild is up for the challenge,” Rasenberger wrote. “This is something we have been doing for 114 years, and the moment we are in today, while challenging, also presents opportunities for us to firmly realize our mission.”

Wolff said this support is important for the next generation of writers.

“I’m lucky that I’m able to have a life as a writer where I enjoy the profits of my work,” Wolff said. “I worry much more about the young writers I’ve worked with at Stanford and elsewhere being able to do that in the face of these technological challenges, because they’re very hard to grasp and to combat.”

Some authors are fighting these difficult legal battles not just to get money from settlements, but also to receive credit for their work.

“Attribution … is very important to the vast majority of authors, and not just academic authors,” Goldstein wrote. “The great majority of countries have something called the ‘moral right of attribution’ on the books to ensure that authors are credited for their work. The U.S., alone among industrialized nations, has no such moral right, which is a terribly unfortunate situation for authors generally.”

As Stanford authors and AI companies continue to deal with copyright questions about AI training, both sides face an unfamiliar situation, Wolff said. They must find a way to balance the rights of creative people with the growth of technology.

April Yu ('30) writes columns for Arts & Life (vol. 270) and contributes to The Daily's News section. She plans to study Human Biology and English and is from New Jersey.