A United States federal judge on Wednesday ruled in favour of tech giant Meta in a closely-watched copyright lawsuit, finding that the company’s use of copyrighted books to train its artificial intelligence (AI) model constituted “fair use” under US law.
The decision, delivered by District Judge Vince Chhabria in San Francisco, dismissed claims brought by a group of authors, including comedian Sarah Silverman and novelist Junot Díaz, who accused Meta of illegally using pirated versions of their works to train its open-source LLaMA generative AI system.
Judge Chhabria noted that although the authors raised important concerns about potential market harm, their legal arguments did not sufficiently demonstrate that Meta’s conduct violated copyright protections.
“No matter how transformative [generative AI] training may be, it’s hard to imagine that it can be fair use to use copyrighted books to develop a tool to make billions or trillions of dollars while enabling the creation of a potentially endless stream of competing works,” the judge said in his ruling. However, he concluded that the plaintiffs had not advanced the correct legal approach to succeed in this instance.
The decision marks the second major courtroom win for AI firms this week, following a separate ruling by US District Judge William Alsup, who also sided with AI company Anthropic in a similar lawsuit filed by authors over the use of copyrighted books to train the Claude chatbot.
In Meta’s case, the authors alleged that the company downloaded pirated copies of their books without permission or payment. Titles cited in the case included The Bedwetter, a memoir by Silverman, and Díaz’s Pulitzer Prize-winning novel The Brief Wondrous Life of Oscar Wao.
Meta welcomed the ruling, stating: “Open-source AI models are powering transformative innovations, productivity and creativity for individuals and companies, and fair use of copyright material is a vital legal framework for building this transformative technology.”
While the ruling offers legal reprieve to Meta, the judge clarified that it did not amount to a blanket endorsement of the company’s practices. “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,” he wrote.
The court’s decision highlights a growing debate over the boundaries of fair use in the age of generative AI, where vast quantities of online content—books, music, art, and journalism—are increasingly being fed into large language models without explicit permission from creators.
Similar lawsuits have been filed in the United States and elsewhere by musicians, artists, and media outlets, many of whom argue that AI companies are profiting from creative work while bypassing established licensing frameworks.
