Overview
- A federal judge gave final approval to the $1.5 billion class‑action settlement on July 20, 2026, clearing the way for distributions to authors and publishers and for court‑ordered compliance steps to begin.
- The deal covers roughly 482,000–500,000 works and, with more than 91% of eligible claimants having filed claims, is expected to yield about $3,000 per qualifying title to rights holders.
- A prior ruling by retired Judge William Alsup found that using books to train AI models can be fair use but also found liability in Anthropic’s downloading and storing of millions of pirated files in a central library.
- The settlement requires Anthropic to delete or destroy the unlawful copies, submit to compliance measures, and funds plaintiffs’ counsel to the tune of roughly $101 million from the settlement pool.
- The agreement resolves past acquisition claims for Anthropic but leaves open separate opt‑out lawsuits and broader litigation over AI outputs, and it is likely to push the industry to prioritize data provenance and licensing controls.