Anthropic to Pay $1.5 Billion to Authors Over Piracy Claims

Anthropic agrees to $1.5 billion settlement over AI model training piracy claims.

Anthropic has agreed to pay $1.5 billion to settle a class-action lawsuit from authors who claim the company stole their works to train AI models, according to a report by BBC.

According to the plaintiffs’ lawyers, the deal, which requires approval from U.S. District Judge William Alsup, would be the largest publicly announced intellectual property settlement in history.

In June, Alsup ruled that Anthropic did not break the law by using books to train AI but must face trial for using pirated content.

The plaintiffs accused the defendant of creating a “central library of all the books in the world,” with works illegally downloaded from pirate sites.

“The fact that the company later bought a copy of the book it previously stole from the internet does not absolve it of liability for theft, but it may affect the amount of statutory damages,” the judge wrote in his decision.

Hearings in the case were expected to continue in December.

Representatives of the authors described the settlement as “the first of its kind in the age of artificial intelligence,” setting a legal precedent.

“This agreement sends a powerful message to companies and creators of AI projects that using copyrighted works from pirate sites is wrong,” said attorney Justin Nelson.

Apple Under Scrutiny

On Friday, authors Grady Hendrix and Jennifer Roberson filed a class-action lawsuit against Apple. They accused the tech corporation of illegally using their books to train AI systems.

The plaintiffs claim that Apple copied copyrighted works without their consent, attribution, or appropriate compensation. The company did not even “attempt to pay” the authors for their contribution to a “potentially lucrative venture,” the statement said.

Back in July, Disney and Universal accused the image-generating AI startup Midjourney of stealing and copying their characters.