Federal Judge Certifies Class in Anthropic AI Book Piracy Lawsuit

Authors have filed a class-action lawsuit against AI company Anthropic, accusing it of using pirated books from pirate libraries like LibGen and Pirate Library Mirror to train its Claude AI model without permission or payment. U.S. District Judge William Alsup ruled that while the actual training of the AI on copyrighted books was transformative and constituted fair use, the large-scale downloading and creation of a digital library of pirated books was not protected under fair use, allowing the copyright infringement claims to proceed. The judge granted class certification limited to owners of copyrights in books downloaded from the two specified pirate libraries, enabling affected authors and copyright holders to potentially claim damages. Anthropic disputes the class action status and is exploring further legal options. The case is part of broader legal challenges facing AI companies over copyright infringement, with potential implications for licensing practices and the future of AI training data sourcing. These rulings mark significant developments in the intersection of AI development and copyright law, balancing the interests of creators and technology firms.