Anthropic has settled a civil suit and agreed to pay authors $1.5B in a landmark copyright case involving the use of training artificial intelligence models on pirated copies of books.
Anthropic, which is backed by Amazon (NASDAQ:AMZN) and other tech giants, was accused by several authors of downloading pirated copies of books to train its Claude chatbot. According to details of the settlement, Anthropic will pay $3,000 for each of the approximately 500,000 pirated books used to train Claude. It will also destroy two of its datasets.
The $1.5B settlement deal must still be approved by U.S. District Judge William Alsup for the Northern District of California in San Francisco. A hearing for the settlement has been set for Monday, Sept. 8.
“The proposed settlement well surpasses other copyright recoveries, will provide meaningful compensation for each Class Work, and will set a precedent of AI companies paying for their use of pirated websites like Library Genesis and Pirate Library Mirror,” reads the motion for the settlement, which was filed today.
The case, which is filed under Bartz et al v. Anthropic PBC, 24-cv-05417, began last year when authors Andrea Bartz, Charles Graeber and Kirk Wallace Johnson sued the startup. A larger group of writers and publishers eventually joined the suit.
Alsup issued a split decision in his summary judgment order earlier this summer. He determined that training AI models on purchased, copyrighted books that were digitized was fair use. However, training models on copyrighted books obtained through pirate websites is not.
“The Court reached different conclusions regarding different sources of training data,” according to court documents. “It found that reproducing purchased and scanned books to train AI constituted fair use. However, the Court denied summary judgment on the copyright infringement claims related to the works Anthropic obtained from Library Genesis and Pirate Library Mirror.”
“The proposed settlement well surpasses other copyright recoveries, will provide meaningful compensation for each Class Work, and will set a precedent of AI companies paying for their use of pirated websites like Library Genesis and Pirate Library Mirror,” according to the settlement.
The settlement sends a “strong message to the AI industry that there are serious consequences when they pirate authors’ works to train their AI, robbing those least able to afford it,” said Authors Guild CEO Mary Rasenberger.
If Anthropic had not agreed to a settlement, the case was expected to go to trial this December, which could have eventually cost Anthropic even more. Anthropic faces another civil suit by Reddit (RDDT) for allegedly using the social media company’s data without a licensing agreement.
Other AI companies are facing similar suits. OpenAI and its primary backer, Microsoft (NASDAQ:MSFT), have been sued by the New York Times (NYT) for alleged copyright infringement for using the newspaper’s content to train AI models.
In June, a federal judge sided with Meta Platforms (NASDAQ:META) in dismissing a copyright infringement lawsuit of authors who alleged the company stole their works to train its AI models.