Anthropic agrees to pay $1.5 billion to settle creator class motion towards coaching AI chatbot Claude with pirated books

Anthropic instructed a San Francisco federal choose on Friday (September 5, 2025) that it has agreed to pay $1.5 billion to settle a class-action lawsuit from a bunch of authors who accused the unreal intelligence firm of utilizing their books to coach its AI chatbot Claude with out permission. The plaintiffs in a court docket submitting requested U.S. District Judge William Alsup to approve the settlement, after saying the settlement in August with out disclosing the phrases or quantity.

“This settlement sends a strong message to AI firms and creators alike that taking copyrighted works from these pirate web sites is unsuitable,” the authors’ legal professionals stated in a press release. They known as it the biggest copyright restoration in historical past and the primary of its sort within the synthetic intelligence period.

Lawsuits towards tech giants

The proposed deal marks the primary settlement in a string of lawsuits towards tech firms together with OpenAI, Microsoft and Meta Platforms over their use of copyrighted materials to coach generative AI methods.

Anthropic as a part of the settlement stated it’ll destroy downloaded copies of books the authors accused it of pirating, and beneath the deal it may nonetheless face infringement claims associated to materials produced by the corporate’s AI fashions.

The $1.5 billion settlement fund quantities to $3,000 for 500,000 downloaded books, and it may develop if extra works are recognized.

In a press release, Anthropic stated the corporate is “committed to developing safe AI systems that help people and organizations extend their capabilities, advance scientific discovery, and solve complex problems.”

The allegations

The settlement doesn’t embrace an admission of legal responsibility. Writers Andrea Bartz, Charles Graeber and Kirk Wallace Johnson filed the category motion towards Anthropic final yr. They argued that the corporate, which is backed by Amazon and Alphabet, unlawfully used hundreds of thousands of pirated books to show its AI assistant Claude to reply to human prompts.

The writers’ allegations echoed dozens of different lawsuits introduced by authors, information shops, visible artists and others who say that tech firms stole their work to make use of in AI coaching. The firms have argued their methods make honest use of copyrighted materials to create new, transformative content material. Alsup dominated in June that Anthropic made honest use of the authors’ work to coach Claude, however discovered that the corporate violated their rights by saving greater than 7 million pirated books to a “central library” that might not essentially be used for that objective.

A trial was scheduled to start in December to find out how a lot Anthropic owed for the alleged piracy, with potential damages ranging into the a whole lot of billions of {dollars}.

Mary Rasenberger, CEO of writers advocacy group the Authors Guild, in a press release on Friday known as the settlement “a significant step in acknowledging that AI firms can not merely steal authors’ artistic work to construct their AI.”

The pivotal fair-use query continues to be being debated in different AI copyright circumstances. Another San Francisco choose listening to the same ongoing lawsuit towards Meta dominated shortly after Alsup’s choice that utilizing copyrighted work with out permission to coach AI can be illegal in “many circumstances.”