Overview
- The lawsuit, filed June 17, 2026 in the US District Court for the Northern District of California, was brought by more than 100 authors who opted out of a prior settlement with Anthropic.
- Plaintiffs allege Anthropic downloaded more than 500 books from shadow libraries and peer‑to‑peer networks such as LibGen, Pirate Library Mirror, and BitTorrent and then stored the material in a central repository used to train Claude.
- The complaint names at least 100 authors, including Laura Esquivel and Nolan Bushnell, and seeks statutory damages of $150,000 per work for a total claim that exceeds $75 million.
- The case shifts the legal focus from whether training a model is fair use to how training data was acquired, a strategy that, if allowed to proceed, could create a new template for copyright suits against AI firms.
- The filing raises fresh financial and reputational risk for Anthropic and its partners and follows a separate roughly $1.5 billion settlement that resolved other claims over the company’s use of books.