A U.S. federal court has ruled in favor of Meta in a major case concerning the use of copyrighted books to train AI systems. On Wednesday, Judge Vince Chhabria stated that Meta’s use of books by 13 authors to train its large language models qualifies as “fair use.” The judge said the plaintiffs failed to show that Meta’s actions caused them financial harm.
The lawsuit, filed in 2023 by authors including Sarah Silverman and Ta-Nehisi Coates, accused Meta of using their copyrighted books without permission. Judge Chhabria emphasized that for a fair use defense, the key issue is whether the copying harms the market for the original work. He found that the plaintiffs didn’t prove that Meta’s use of their books affected sales or income.
This ruling contrasts with a separate fair use decision earlier this week involving Anthropic, where the judge focused more on whether the AI’s use was “transformative.” Chhabria, however, dismissed “market dilution” as a standalone argument against fair use and said that more evidence would be needed in future cases.
Although Meta welcomed the ruling, the authors’ legal team expressed disappointment, stating that Meta had still used copyrighted works without consent. Judge Chhabria did note that copying protected material for AI training could be illegal in other contexts, and that companies may need to compensate creators in the future.
This decision adds to a growing body of legal precedents on AI and copyright. Other lawsuits are still pending, including one against Microsoft and another involving Getty Images and Stability AI. Courts have so far largely sided with AI companies under fair use, but the legal debate is still evolving.
