US District Judge Araceli Martinez-Olguin in San Francisco approved Anthropic’s $1.5 billion copyright settlement on Friday, ending a class action lawsuit brought by authors who accused the company of training Claude on pirated copies of their books. According to Reuters via LiveMint, the deal is the largest copyright settlement in US history.
The case centered on Anthropic’s training data practices. The authors alleged that the company maintained a repository containing more than 7 million pirated books, many stored even if they were never used for training.
The Fair Use Ruling That Changed Everything
The settlement follows a landmark ruling in 2025 by now-retired Judge William Alsup, who held that using copyrighted books to train AI models constitutes fair use under US copyright law. Alsup initially approved the deal in September 2025 before retiring. Judge Martinez-Olguin handled the final approval.
The fair use ruling created a legal foundation that extends well beyond Anthropic. OpenAI, Google, and Meta all face similar lawsuits from authors and publishers. A federal court holding that AI training qualifies as fair use provides strong precedent for those defendants.
However, Judge Alsup’s ruling was not a blanket clearance. He found that Anthropic could still face liability for downloading and retaining millions of pirated books in what the court described as its “central library,” regardless of whether those books were ultimately used for training. The distinction matters: training on copyrighted material may be fair use, but building and maintaining a pirated book repository carries separate legal risk.
Settlement Details
More than 91% of authors and publishers covered by the settlement have claimed their share of the $1.5 billion payment, according to Anthropic Deputy General Counsel Aparna Sridhar. “We reached this settlement in 2025, after the court’s landmark ruling that training AI on books is fair use under copyright law, which remains the law today,” Sridhar said in a statement reported by Reuters.
Justin Nelson, lead attorney for the authors, called it “the largest known copyright recovery in history.”
The settlement faced challenges from a group of authors and publishers who argued the payout was too low, attorneys were overcompensated, and some authors were wrongfully excluded. Judge Martinez-Olguin dismissed those objections, stating the complaints about the settlement’s size were “not grounded in a realistic assessment of the overall risks and rewards of a trial.” The court awarded the attorneys over $101 million of the $187.5 million they requested in fees.
A trial had been scheduled for December 2025 to determine damages, which could have reached hundreds of billions of dollars. Both sides chose to settle before that deadline.
Who Opted Out
Several authors and publishers opted out of the agreement and continue to pursue separate copyright lawsuits against Anthropic. Their cases will proceed independently, and the fair use ruling from Judge Alsup does not bind other courts or preclude different outcomes in those proceedings.
The Precedent for Agent Development
The fair use determination reduces legal uncertainty for any company training foundation models on public text. For agent builders, this is infrastructure-level clarity: the models powering autonomous agents have a stronger legal foundation for their training data. Before this ruling, the copyright question hung over every commercial deployment of a large language model. A federal court affirming fair use, combined with the largest settlement in copyright history, signals that the legal framework is stabilizing around permissive training rights with financial compensation for rights holders.
The question now shifts to whether other AI companies settle on similar terms or push their own cases to trial. OpenAI faces multiple active lawsuits from The New York Times, authors’ groups, and other publishers. Google and Meta face parallel claims. The Anthropic settlement creates a benchmark: $1.5 billion for a company valued at roughly $60 billion, paid to resolve claims that could have reached hundreds of billions at trial.