Anthropic has agreed to a $1.5 billion settlement following a class-action lawsuit accusing the company of using pirated e-books in its AI training. The lawsuit, filed in 2024, alleged that Anthropic downloaded millions of copyrighted works from shadow libraries like LibGen, incorporating around 482,000 of these works into its Claude training database. The court ruled that while AI reading legally obtained books might be fair use, acquiring and storing pirated content constitutes infringement. As part of the settlement, Anthropic will compensate authors approximately $3,000 per work and destroy the pirated files.