Anthropic wins key ruling on AI in authors' copyright lawsuit

A U.S. federal judge ruled that Anthropic’s use of copyrighted books to train its artificial intelligence (AI) system constitutes fair use under U.S. copyright law. Judge William Alsup concluded that the San Francisco-based company 'exceedingly transformed' the source material by using it to develop its Claude large language model. This decision marks a pivotal legal victory for the AI industry, as it is the first ruling addressing fair use specifically in the context of generative AI. However, the judge also found that Anthropic violated copyright law by storing authors' books in a centralized library, which was not considered fair use. The authors—Andrea Bartz, Charles Graeber, and Kirk Wallace Johnson—filed a class action lawsuit in 2024, accusing Anthropic, backed by Amazon and Alphabet, of using pirated versions of their books without permission or compensation. The ruling underscores the legal tension between traditional copyright protections and emerging AI technologies, with implications for similar lawsuits against other tech firms like OpenAI, Microsoft, and Meta. Anthropic defended its actions as legally permissible and beneficial to creativity and innovation, asserting that its system used the books to generate new, transformative content rather than replicate the originals.