Anthropic has agreed to pay at least $1.5 billion to settle a class action lawsuit brought by authors who accused the artificial intelligence startup of illegally using their books to train its systems.
Under the deal, the company will pay about $3,000 per book plus interest and destroy datasets containing the disputed material, according to a court filing Friday. If approved, the agreement would mark the largest publicly reported copyright recovery in history.
“This settlement sends a powerful message to AI companies and creators alike that taking copyrighted works from these pirate websites is wrong,” Justin Nelson, an attorney for the authors, told CNBC.
The case was filed in the U.S. District Court for the Northern District of California last year by authors Andrea Bartz, Charles Graeber and Kirk Wallace Johnson. They alleged Anthropic engaged in “largescale copyright infringement by downloading and commercially exploiting books that it obtained from allegedly pirated datasets,” including Library Genesis and Pirate Library Mirror.
In June, a judge found that Anthropic’s use of books for training could qualify as “fair use,” but allowed the trial to move forward to determine whether the company had infringed by sourcing the works from pirated sites. The trial was set for December before the settlement.
Anthropic, which recently raised $13 billion in new funding at a $183 billion valuation, did not immediately comment on the filing.
Related Readings:Â
