Anthropic settles $1.5B copyright case with authors
A U.S. federal judge in San Francisco gave final approval on July 20, 2026, to a $1.5 billion copyright settlement between AI company Anthropic PBC and a class of authors and publishers who alleged the company used pirated books to train Claude, resolving what Reuters describes as the largest known U.S. copyright settlement in history.
U.S. District Judge Araceli Martínez-Olguín of the U.S. District Court for the Northern District of California approved the settlement after the class action was brought by authors including Andrea Bartz, Charles Graeber, and Kirk Wallace Johnson. The judge said the settlement provides "meaningful relief" to affected authors and publishers.
Settlement Structure and Payouts
The $1.5 billion total will be distributed to copyright holders at approximately $3,000 per book or work after administrative fees and costs are deducted, according to AP News and Reuters reporting. Anthropic Deputy General Counsel Aparna Sridhar confirmed that more than 91% of authors and publishers covered by the settlement have claimed their share of the payment.
The Legal Background
AnthropIc had reached the settlement in 2025, following what Sridhar described as "the court's landmark ruling that training AI on books is fair use under copyright law — which remains the law today." The civil class action centered on allegations that Anthropic sourced pirated copies of copyrighted books to train Claude, without compensating authors or publishers.
The case represents a watershed moment for copyright enforcement in the generative AI sector. While the fair use ruling remains in effect—meaning training on books does not inherently violate copyright—the settlement reflects a negotiated resolution that provides compensation to the creative community whose works were used.
Implications for the AI Industry
The magnitude of the settlement signals potential financial and reputational risks for other AI companies facing similar litigation. Other major model developers, including OpenAI and Meta, are defending against comparable copyright class actions in U.S. courts. The Anthropic settlement does not establish new copyright law but demonstrates how much a company may be willing to pay to resolve disputed training practices and move forward without prolonged litigation.
Sridhar's statement that the company is "looking forward to bringing this matter to a close" suggests Anthropic is prioritizing regulatory clarity and market confidence over extended court battles. The 91% claims rate also indicates that the settlement administration process has been effective in reaching the intended beneficiaries.