Meta has won a key legal victory in a copyright lawsuit filed by authors including Sarah Silverman and Ta-Nehisi Coates, who claimed the company used pirated books to train its Llama AI system without permission.
U.S. District Judge Vince Chhabria ruled the plaintiffs failed to show Meta’s actions violated copyright law, particularly that it harmed the market for their work. However, he emphasized the decision was procedural, not a blanket endorsement of Meta’s practices.
“This ruling does not stand for the proposition that Meta’s use of copyrighted materials is lawful,” Chhabria wrote. “It stands only for the proposition that these plaintiffs made the wrong arguments.”
The decision marks the second win for the AI industry this week, following a ruling in favor of Anthropic. Yet unlike that case, where a judge found the use of copyrighted works to be “fair use,” Chhabria noted that unauthorized AI training may be illegal in “many circumstances.”
The lawsuit is part of a growing wave of copyright cases targeting AI companies like Meta, OpenAI, Microsoft, and Anthropic. Creators argue their work is being copied to generate competing content, threatening livelihoods. Tech firms claim fair use protects their ability to train generative AI models on public data.
A Meta spokesperson praised the ruling and called fair use “vital” for AI innovation. The authors’ legal team said it disagreed, calling Meta’s use of copyrighted works “historically unprecedented piracy.”
The broader legal debate over how AI systems may legally learn from copyrighted material is far from settled, with more trials expected later this year.
