Published on Jul 22, 2025
Anthropic is facing a certified class action lawsuit—and the prospect of massive statutory damages—over its use of pirated datasets to train its generative AI systems.
“It’s a huge deal,” Edward Lee, a professor at Santa Clara University Law School told The Capitol Forum, of U.S. District Judge William H. Alsup’s July 17 move to certify a class in the case, Bartz v. Anthropic.
The newly certified class could include millions of copyright holders, entitled to damages as high as $150,000 per violation. The size of that class and scope of alleged infringement could technically lead to a damages award exceeding $1 trillion, although, in practice, due process considerations would almost certainly limit that number.
Nonetheless, a damages award that’s even a fraction of the statutory limit could be ruinous to Anthropic, as the Amazon-backed company conceded in fighting the motion for class certification.
But Alsup, in certifying the class, found Anthropic’s arguments unconvincing. “This is the classic case for certification,” Alsup wrote.
The rest of this article is currently locked and only available to subscribers. Request a trial to receive unlimited access to all articles.
