Judge: AI Book Training Fair Use, Piracy Trial Next

U.S. District Judge William Alsup ruled that Anthropic’s use of legally acquired copyrighted books to train its AI model Claude is transformative fair use, establishing a significant precedent for the AI industry. The judge likened AI training to a person reading and learning, noting that AI-generated outputs are not substitutes for the originals and do not harm authors' markets. However, Alsup found Anthropic’s downloading and storage of over seven million pirated books violated copyright law, setting a separate trial to determine potential damages. The decision is considered a major win for AI companies like Anthropic, OpenAI, and Google, but unresolved questions remain around pirated content. Legal experts expect the fair use ruling to be appealed, and the case is seen as a pivotal moment in ongoing debates over AI and copyright. The outcome is expected to shape future legal frameworks and industry norms in the U.S. and possibly worldwide.