Judge says training is fair use, but Anthropic pays $1.5B for piracy
anthropic's $1.5b copyright settlement for illegally downloaded books was approved by a federal judge, though the judge ruled that ai training on copyrighted texts is fair use. this legal split leaves the industry without a binding precedent on the core training issue.
anthropic reached final approval to pay $1.5b to settle a class action lawsuit regarding its use of pirated books, distributing funds across an estimated 500,000 works. while judge william alsup found that anthropic's acquisition of books from sites like library genesis was illegal, he ruled that the actual training of ai models on copyrighted text is fair use. anthropic settled to avoid a jury trial specifically on the piracy claims, meaning the industry-defining question of training legality remains unresolved at the appellate level. this leaves other ai labs like google, meta, and openai facing ongoing lawsuits without a clear legal safeguard.