A US federal judge in San Francisco has approved artificial intelligence company Anthropic's landmark 1.5 billion dollar settlement. This resolves a class action lawsuit brought by authors who accused the company of misusing their books. The authors claimed Anthropic used their copyrighted material to train its AI chatbot, Claude, without permission.
U.S. District Judge Araceli Martinez-Olguin granted final approval for the settlement on Monday. This decision makes it the largest known settlement of a US copyright case. The judge rejected arguments that the settlement amount was too small, emphasizing the risks and rewards of a trial.
This case is one of many lawsuits filed by copyright owners, including authors and news outlets. These lawsuits target technology companies over the use of their content to train large language models (LLMs). The Anthropic settlement is the first major US case of its kind to reach a resolution.
Anthropic's deputy general counsel, Aparna Sridhar, stated the settlement was reached in 2025. This followed a court ruling that training AI on books constitutes fair use under copyright law. Sridhar expressed satisfaction that over 91% of affected authors and publishers have claimed their share of the payment.
The authors' lead attorney, Justin Nelson, hailed the agreement as a "historic settlement." He noted it represents the largest known copyright recovery in history. Nelson confirmed that distributions to the class of authors covered by the settlement would be made promptly.
Writers initially sued Anthropic in 2024, alleging the company used pirated versions of their books. Anthropic, backed by major companies like Amazon and Alphabet, reportedly used these books to teach Claude to respond to human prompts. This practice raised significant concerns about intellectual property rights in the rapidly evolving AI landscape.
Last June, now-retired Judge William Alsup ruled that Anthropic's use of the authors' work for training Claude was fair use. However, he also found that the company violated their rights by saving over 7 million pirated books to a "central library." This library was not necessarily used for AI training, indicating a broader issue of unauthorized storage.
A trial was initially scheduled for last December to determine the damages Anthropic owed for the alleged piracy. Potential damages could have run into hundreds of billions of dollars, highlighting the financial stakes involved. The settlement averted this potentially lengthy and costly legal battle.
Some authors objected to the settlement, arguing it was insufficient or unfairly compensated the plaintiffs' attorneys. They also claimed it wrongly excluded some copyright owners. Judge Martinez-Olguin overruled these objections in her recent ruling, solidifying the settlement's approval.
The judge awarded the attorneys more than 101 million dollars from the 187.5 million dollars they requested in fees. This allocation reflects the significant legal work involved in securing the settlement. Some authors and publishers opted out of this settlement and have filed separate, ongoing lawsuits against Anthropic.
This landmark decision provides a significant precedent for future copyright disputes in the AI industry. It underscores the ongoing tension between technological innovation and intellectual property protection. Companies developing AI models must now carefully consider the legal implications of their data sourcing practices.