Meta Wins in Court, but Judge Signals Future AI Copyright Battles Ahead

Meta scored a legal victory Wednesday when a federal judge ruled that the company's use of 13 authors' books to train its AI models was legal under copyright law's "fair use" doctrine. But Judge Vince Chhabria made clear this wasn't a blank check for tech companies to raid copyrighted material.

"This ruling does not stand for the proposition that Meta's use of copyrighted materials to train its language models is lawful," Chhabria wrote. "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."

The case involved well-known authors including Sarah Silverman, Ta-Nehisi Coates, and Junot Díaz, who sued Meta in 2023 for using their books without permission to train its Llama AI models. Meta had downloaded the books from "shadow libraries"—essentially pirate sites—after licensing negotiations with publishers fell through.

The Authors Made the Wrong Case

The plaintiffs argued that Meta's AI could reproduce snippets of their books and that the company had harmed the market for licensing their works to AI trainers. Both arguments failed. Meta's tests showed that even when prompted to regurgitate training material, Llama could only produce a maximum of 50 words from any of the plaintiffs' books.

The licensing argument also fell flat. Courts have long held that plaintiffs can't claim harm from the loss of licensing fees for the exact use being challenged—otherwise every fair use case would automatically favor copyright holders.

But the judge saw a much stronger argument the authors largely ignored: market dilution.

The Real Threat: AI Flooding Markets

Chhabria warned that generative AI has "the potential to flood the market with endless amounts of images, songs, articles, books, and more" using "a tiny fraction of the time and creativity that would otherwise be required". "So by training generative AI models with copyrighted works, companies are creating something that often will dramatically undermine the market for those works, and thus dramatically undermine the incentive for human beings to create things the old-fashioned way."

The judge suggested that AI-generated romance novels could "successfully crowd out lesser-known works or works by up-and-coming authors." While AI books probably wouldn't affect Agatha Christie's sales, "they could very well prevent the next Agatha Christie from getting noticed or selling enough books to keep writing."

This case differed from typical copyright disputes because it "involves a technology that can generate literally millions of secondary works, with a miniscule fraction of the time and creativity used to create the original works it was trained on."

Tech Companies Shouldn't Celebrate Yet

Despite ruling for Meta, Chhabria emphasized that "in many circumstances it will be illegal to copy copyright-protected works to train generative AI models without permission". He predicted that "in cases involving uses like Meta's, it seems like the plaintiffs will often win, at least where those cases have better-developed records on the market effects of the defendant's use."

The judge also dismissed tech industry claims that copyright restrictions would kill AI development. "These products are expected to generate billions, even trillions, of dollars for the companies that are developing them," he wrote. "If using copyrighted works to train the models is as necessary as the companies say, they will figure out a way to compensate copyright holders for it."

What This Actually Means

The ruling only affects these 13 specific authors—not the "countless others whose works Meta used to train its models," since this wasn't a class action lawsuit. A separate case management conference is scheduled for July to discuss Meta's alleged distribution of copyrighted works during the downloading process.

This marks the second AI copyright victory for tech companies this week. On Monday, another San Francisco judge ruled that Anthropic's use of copyrighted books to train its Claude AI was also fair use, though the company still faces trial for allegedly pirating the books from illegal sources rather than buying them.

The Bigger Picture

The decision comes as dozens of similar lawsuits wind through courts, including cases by The New York Times against OpenAI and Disney against image generator Midjourney. The judge noted that "markets for certain types of works (like news articles) might be even more vulnerable to indirect competition from AI outputs."

Meta had initially tried to license books and discussed spending up to $100 million on licensing deals, but found that publishers generally don't hold the rights to license books for AI training—those rights belong to individual authors, and no collective licensing organization exists.

Why this matters:

• Authors won the argument but lost the case—Chhabria essentially provided a roadmap for future plaintiffs to build stronger market dilution claims
• The real copyright wars are just beginning—this narrow ruling suggests other cases with better evidence could easily go the other way, especially as AI models become more capable of flooding markets with competing content