Overview
- Filings on Friday detail a proposed class deal that plaintiffs say would be the largest publicly reported copyright recovery in history.
- The agreement would pay about $3,000 per work for roughly 500,000 books, with total payments increasing if more works are added; an initial $300 million installment would be due shortly after approval.
- Anthropic would destroy downloaded copies obtained from piracy sites such as Library Genesis and Pirate Library Mirror, and the pact includes no admission of liability.
- The settlement follows a June ruling that found AI training on lawfully obtained books can be fair use but said Anthropic unlawfully retained more than 7 million pirated books in a central repository.
- Plaintiffs and legal observers say the outcome could influence parallel copyright suits and accelerate licensing of training data across the AI industry.