Anthropic to Pay Authors $1.5 Billion Over Pirated Books for AI Training

Artificial intelligence company Anthropic has agreed to pay $1.5 billion to settle a class-action lawsuit brought by authors who say the firm used pirated books to train its chatbot, Matt O’Brien reported for the Associated Press (AP).

Under the terms, Anthropic will pay authors and publishers roughly $3,000 for each of an estimated 500,000 books used to train its Claude chatbot. (Image: Anthropic)

The landmark deal could reshape legal battles between AI companies and creators alleging copyright infringement.

Under the terms, Anthropic will pay authors and publishers roughly $3,000 for each of an estimated 500,000 books used to train its Claude chatbot. Thriller writer Andrea Bartz and nonfiction authors Charles Graeber and Kirk Wallace Johnson filed the lawsuit last year.

They now represent a broader group of writers and publishers in the settlement, which still requires approval from US District Judge William Alsup, Blake Brittain and Mike Scarcella also reported for Reuters.

“If approved, this landmark settlement will be the largest publicly reported copyright recovery in history, larger than any other copyright class action settlement or any individual copyright case litigated to final judgment,” the plaintiffs said in their filing, CNN also reported.