Anthropic trained Claude on illegally obtained Stanford professors' books
Points and comments are a snapshot, not live.
Judge ruled Anthropic's AI training on legally acquired books is fair use, but pirating them is illegal.
A federal judge ruled in May 2026 that Anthropic illegally scraped and pirated over 10 Stanford professors' books from shadow libraries like LibGen to train Claude. The $1.5 billion Bartz v. Anthropic settlement covered nearly 500,000 works, but Judge Alsup distinguished the illegal downloading from the training itself, calling training on legally acquired books fair use. Professors Tobias Wolff and Chang-rae Lee opted out of the settlement to seek higher damages in Cruz v. Anthropic. The legal future of mass scraping for AI training remains uncertain as over 70 infringement lawsuits proceed against AI companies.
What commenters are saying
Commenters largely corrected the article's title, noting the judge explicitly ruled that training on legally acquired books is fair use; the illegal act was obtaining books from LibGen rather than buying them. A few argued this distinction is insignificant. Some compared Anthropic's civil settlement to Aaron Swartz's criminal prosecution for similar acts, criticizing unequal justice. Others defended information freedom, claiming knowledge should not be gatekept, while counterpoints stressed the need to compensate authors. The thread split between those seeing the ruling as reasonable and those calling it a failure of copyright law.