A federal judge in San Francisco has granted final approval to a $1.5 billion class-action settlement between Anthropic and thousands of authors. The deal resolves claims that the AI company used pirated books to train its Claude chatbot, marking the largest copyright recovery in U.S. history.
Court Approval and the Scope of the Settlement
U.S. District Judge Araceli Martínez-Olguín issued the final order on Monday, concluding a high-profile legal battle that has served as a bellwether for the artificial intelligence industry. The settlement provides what the court described as meaningful relief
to authors and publishers whose works were ingested by Anthropic’s large language models. The agreement covers more than 482,000 individual books.
While the dollar figure is substantial, the payout equates to roughly $3,000 per book, a figure that has drawn criticism from some rights holders who contend the amount fails to reflect the true value of their intellectual property. Despite these objections, the court rejected arguments that the settlement was insufficient, noting that such complaints were not grounded in a realistic assessment of the overall risks and rewards of a trial,
according to reporting by Reuters.
Legal Distinctions: Fair Use vs. Piracy
The resolution of this case hinges on a critical legal nuance established by the court. While the litigation was sparked by the training of the Claude chatbot, the court’s rulings distinguished between the AI training process and the storage of source data. As previously determined by now-retired Judge William Alsup, the act of training an AI model on copyrighted material is considered fair use. However, the court found that Anthropic crossed a line by maintaining a central library
of more than 7 million pirated books that were not necessarily required for model training.
This distinction is vital for the broader AI sector. By settling, Anthropic avoids a trial that could have resulted in damages reaching into the hundreds of billions of dollars, while simultaneously securing a judicial nod that the core mechanism of its technology—AI training—remains legally protected under the umbrella of fair use.
Claims Process and Remaining Litigation
According to Anthropic, the administrative phase of the settlement has seen significant participation. The company reported that more than 91% of eligible authors and publishers have already submitted claims to receive their share of the payout. The legal team representing the authors has expressed satisfaction with the outcome, viewing it as a benchmark for future intellectual property disputes in the tech age.
“It is the largest known copyright recovery in history. We look forward to making distributions to the Class as promptly as possible.”
Justin Nelson, lead attorney for the plaintiffs
Despite the court’s approval, the matter is not entirely closed for Anthropic. Some authors and publishers, including Chicken Soup for the Soul, chose to opt out of the class action. These entities are continuing to pursue separate, ongoing litigation against the company, arguing that the settlement terms are inadequate and were effectively forced down the throat of authors,
as noted in earlier proceedings by Judge Alsup.
Attorney Fees and Financial Distribution
The financial mechanics of the settlement also involved a significant allocation for legal counsel. While the plaintiffs’ attorneys requested $187.5 million in fees, Judge Martínez-Olguín ultimately awarded them more than $101 million. This reduction reflects the court’s oversight of the class action structure, ensuring that the bulk of the $1.5 billion remains designated for the authors themselves rather than the legal teams.
The Future of AI Copyright Enforcement
This case stands as the first major settlement in a wave of copyright litigation targeting artificial intelligence companies. Anthropic’s deputy general counsel, Aparna Sridhar, stated that the company remains committed to the legal standards established throughout the process. 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.
As Anthropic moves to destroy the disputed copies of the books held in its library, the industry remains focused on the ongoing cases that opted out of this settlement. Whether these separate lawsuits will challenge the fair use precedent or simply seek higher compensation remains the next major hurdle for the company.
