Judge Tosses Authors’ Copyright Suit Against Meta Over AI Training

Background: The Lawsuit Against Meta

A federal judge in San Francisco has dismissed a high-profile copyright lawsuit brought by a group of authors against Meta, marking another legal win for AI companies training large models on copyrighted works.

The case, led by comedian Sarah Silverman and writers including Jacqueline Woodson and Ta-Nehisi Coates, accused Meta of using pirated versions of their books to train its Llama AI system without permission or payment. The authors argued this amounted to copyright infringement and threatened the market for their works.

Court’s Reasoning and Ruling

But U.S. District Judge Vince Chhabria was unconvinced—at least, by the arguments presented. In his ruling, Chhabria found the authors “made the wrong arguments and failed to develop a record in support of the right one,” ultimately tossing the case. He noted that while Meta’s use of the works was “transformative” enough to potentially qualify as fair use, the ruling applies only to these plaintiffs and does not establish that Meta’s AI training practices are definitively lawful.

Implications for AI and Copyright Law

Chhabria’s decision is the second such courtroom victory for AI firms this week, following a similar outcome for Anthropic. However, the judge’s opinion leaves the door open for future challenges. He acknowledged that training generative AI on copyrighted books could, under different arguments, be found to harm authors and violate copyright law, especially if it enables “a potentially endless stream of competing works that could significantly harm the market for those books”.

Meta’s Response and the Road Ahead

Meta welcomed the ruling, emphasizing that open-source AI models drive “transformative innovations, productivity, and creativity,” and that fair use is a “vital legal framework” for building such technology. The authors’ legal team did not immediately respond to requests for comment.

While Meta prevailed this round, Judge Chhabria’s 40-page ruling reads more like a warning shot than a sweeping endorsement. As he put it, “This ruling does not stand for the proposition that Meta’s use of copyrighted materials to train its language models is lawful. It stands only for the proposition that these plaintiffs made the wrong arguments and failed to develop a record in support of the right one”.

Conclusion: Ongoing Legal Uncertainty

With the legal landscape around AI and copyright still unsettled, the decision is unlikely to be the last word. The invitation is clear: authors may yet succeed if they bring the right case.