Anthropic wins ruling on AI training in copyright lawsuit but must face trial on pirated books

A federal judge has ruled that AI company Anthropic did not violate U.S. copyright law by training its Claude chatbot on millions of copyrighted books, deeming the use 'quintessentially transformative' and therefore fair use. However, Anthropic must face trial in December over allegations it illegally downloaded those books from pirated online libraries. The lawsuit, brought by authors Andrea Bartz, Charles Graeber, and Kirk Wallace Johnson, accuses Anthropic of large-scale theft to profit from the intellectual labor in those books. Internal documents showed employee concerns over the legality of using pirated sources. Although Anthropic later began purchasing and digitizing books more lawfully, the judge stated that prior piracy could still incur liability. The outcome of this case may influence similar lawsuits against other AI companies like OpenAI and Meta. Anthropic expressed satisfaction with the fair use ruling but did not comment on the piracy allegations. Founded in 2021 by former OpenAI members, Anthropic promotes itself as a responsible AI developer, but the lawsuit challenges this image.