Anthropic told a San Francisco federal judge it has agreed to pay $1.5 billion to settle a class-action lawsuit brought by authors who said the company used their books to train its AI chatbot Claude without permission, according to CNN. The proposed deal, which follows an August announcement that did not disclose terms, now awaits approval from U.S. District Judge William Alsup.
A “landmark” deal and the road to approval
In a joint court filing, Anthropic and the plaintiffs asked Judge Alsup to approve the settlement. The authors called it a “landmark” agreement and said that, if approved, it would be the largest publicly reported copyright recovery in history, exceeding any prior copyright class action settlement or individual case that went to final judgment.
The class action was filed last year by writers Andrea Bartz, Charles Graeber and Kirk Wallace Johnson. They alleged that Anthropic, backed by Amazon and Alphabet, unlawfully used millions of pirated books to train Claude to respond to human prompts. Their claims echoed numerous other lawsuits from authors, news organizations and visual artists who contend tech companies took their work for AI training without permission.
Fair use rulings and December trial timeline
Judge Alsup ruled in June that Anthropic made fair use of the authors’ work to train Claude. However, he also found the company violated their rights by saving more than 7 million pirated books to a “central library” that would not necessarily be used for training. A trial had been scheduled for December to determine damages for the alleged piracy, with potential exposure described in the filing as reaching into the hundreds of billions of dollars.
Wider implications for generative AI lawsuits
The settlement is the first resolution among a series of cases targeting major tech companies over the use of copyrighted material in training generative AI systems. Other companies, including OpenAI, Microsoft and Meta Platforms, have argued that their systems make fair use of copyrighted content to produce new, transformative outputs.
The fair-use debate remains unsettled in other courts. Another San Francisco judge overseeing a separate lawsuit against Meta ruled shortly after Alsup’s decision that using copyrighted work without permission to train AI would be unlawful in “many circumstances.” The outcome of Anthropic’s proposed settlement could influence how similar disputes proceed as courts continue to assess how copyright law applies to AI training.
