After utilizing unauthorized copies of writers’ works to train its Claude chatbot, Anthropic, an artificial intelligence corporation, will pay thousands of authors around $3,000 per book as part of a $1.5 billion copyright settlement approved by a federal judge.
In a decision on Monday, District Judge Araceli Martínez-Olguín stated that the class-action settlement offers impacted writers and publishers “meaningful relief.”
Approximately 91% of the roughly 482,000 books affected by the verdict have been claimed by publishers or authors who are now owed money.
The payment was “the largest known copyright recovery in history,” according to a statement from plaintiff lawyer Justin Nelson. We are eager to provide the class with distributions as soon as we can.
The preliminary approval was granted in September of last year in a federal court in San Francisco by U.S. District Judge William Alsup, who has since retired.
Last summer, Alsup rendered a divided decision in the case, concluding that while it was not illegal to train AI chatbots on copyrighted literature, Anthropic had improperly obtained millions of books through pirate websites.
Aparna Sridhar, the deputy general counsel for Anthropic, described the decision on Friday as a historic demonstration “that training AI on books is fair use under copyright law.”
In a written statement, Sridhar stated, “We are happy that more than 91% of authors and publishers covered by the settlement have claimed their share of the payment, and we’re looking forward to bringing this matter to a close.”
In 2024, Andrea Bartz, a best-selling thriller novelist, filed the lawsuit alongside two other writers. In the dozens of AI copyright cases that are still pending in court, this is the first significant settlement.
