Anthropic Settles AI Book-Training Lawsuit with Authors

Anthropic has come to a settlement in a class action lawsuit involving a collective of fiction and nonfiction authors, as disclosed in a filing with the Ninth Circuit Court of Appeals on Tuesday. Previously, Anthropic achieved a partial victory in a lower court ruling and was gearing up to appeal that verdict. Specifics about the settlement remain confidential, and the company has not yet responded to requests for comments.

The lawsuit, titled Bartz v. Anthropic, focuses on Anthropic’s use of books as training materials for its large language models. Although the court ruled that Anthropic’s usage of these books was within fair use, the presence of many pirated books resulted in significant financial consequences for the company.

Nonetheless, Anthropic celebrated the previous court ruling, framing it as a victory for generative AI models. “We believe it’s evident that we acquired books solely for the purpose of developing large language models — and the court clearly stated that use was fair,” the company told NPR after the June ruling.