Anthropic Wins Approval for Record $1.5 Billion Copyright Settlement With Authors

A US federal judge in San Francisco has approved a landmark $1.5 billion settlement between artificial intelligence company Anthropic and a group of authors who accused the company of using their books without permission to train its Claude chatbot.

US District Judge Araceli Martinez-Olguin granted final approval to the agreement on Monday, making it the largest known settlement involving a copyright case in the United States.

The case is among dozens filed by copyright holders, including authors and news organisations, against technology companies over the use of copyrighted material to train large language models. It is the first major US case of its kind to reach a settlement.

The authors sued Anthropic in 2024, alleging that the company used pirated copies of their books to train Claude, which generates responses to users’ questions and instructions. Anthropic is backed by major technology companies including Amazon and Alphabet.

Retired US District Judge William Alsup had previously approved the settlement in September. In an earlier ruling last June, he found that Anthropic’s use of the authors’ works to train Claude qualified as fair use.

However, Alsup also ruled that the company had violated the authors’ rights by storing more than seven million pirated books in a central library. The books were not necessarily all used to train the AI system.

A trial had been scheduled for last December to determine potential damages related to the alleged piracy. The financial consequences could have reached hundreds of billions of dollars, making the settlement a significant resolution for Anthropic.

An attorney representing the authors told the court that claims had been filed covering more than 92% of the more than 480,000 works included in the settlement.

The agreement has faced opposition from some authors, who argued that the payment was insufficient or that the settlement gave too much compensation to the lawyers representing the plaintiffs. Others questioned why certain copyright owners were excluded from the agreement.

Some authors and publishers chose not to participate in the settlement and have filed separate lawsuits against Anthropic. Those cases remain active.

The dispute comes as technology companies face growing legal challenges over how copyrighted books, news articles, images and other material are collected and used to train AI systems.

The outcome of the Anthropic case could influence negotiations and legal strategies in other copyright disputes involving AI developers. While the settlement resolves claims brought by the participating authors, it does not establish a broad legal ruling covering all uses of copyrighted material in AI training.