Federal Judge Rules Meta’s AI Training on Copyrighted Books Qualifies as Fair Use
A U.S. federal judge has ruled that Meta Platforms' use of copyrighted books to train its artificial intelligence models falls under fair use, dealing a significant setback to 13 authors including Sarah Silverman and Pulitzer winners Junot Díaz and Andrew Sean Greer. The decision echoes a similar ruling in favor of Anthropic, though courts caution that AI training practices remain legally ambiguous.
Judge Vince Chhabria dismissed the authors' claims, noting their failure to demonstrate how Meta's AI WOULD harm the market for their works. "The argument barely gives this issue lip service," the ruling stated, emphasizing the lack of substantive evidence. However, the judgment stops short of endorsing tech companies' broader AI training methodologies.
