Overview

• A court filing says the parties have negotiated a proposed class settlement, with terms undisclosed and finalization expected on September 3.
• The case centered on allegations that Anthropic downloaded up to 7 million pirated books to build a training library for its AI models.
• In June, Judge William Alsup ruled that training on lawfully purchased books can be fair use but found liability for long‑term retention of pirated copies, noting later purchases might reduce but not erase damages.
• The agreement spares Anthropic a jury trial that could have exposed it to statutory damages of up to $150,000 per willful infringement, potentially totaling billions.
• Plaintiffs’ attorney Justin Nelson called it a “historic settlement” with details to follow in the coming weeks, while Anthropic declined comment as broader AI copyright suits continue.