Judge approves Anthropic’s $1.5B settlement in AI book copyright case

In short: A federal judge approved Anthropic’s $1.5 billion settlement with authors and publishers over copyrighted books.

What happened

A US federal judge has given final approval to a $1.5 billion settlement between Anthropic and a group of authors and book publishers who sued the company for copyright infringement. The lawsuit was a class action, meaning it was filed on behalf of a large group.

The case focused on how Anthropic built a library of books used to train its AI model, which is like teaching software patterns by having it read a huge number of pages. A judge previously found Anthropic illegally downloaded and stored millions of copyrighted books. The judge also said Anthropic got books in two ways, by buying and scanning them and by downloading them from pirate sites like Library Genesis and Pirate Library Mirror, and the piracy part was illegal.

Under the settlement, the payout is expected to be $3,000 per work across an estimated 500,000 works. The money will be shared among the authors and publishers who own the rights. The original judge, William Alsup, has retired, and Judge Araceli Martinez-Olguin signed off on the settlement.

Why it matters

This settlement closes one major case, but it does not settle the bigger question across the whole industry. Alsup ruled that training an AI model on copyrighted text can count as “fair use” (a legal rule that sometimes allows limited use without permission), but that decision was from one district court. Other lawsuits against companies like Google, Meta, Midjourney, and OpenAI are still moving forward, so rules could still change depending on what other judges decide.

Source: TechCrunch AI