Meta wins copyright case against authors; future cases though not ruled out

A federal judge has dismissed a high-profile copyright infringement lawsuit against Meta Platforms, ruling in favor of the Facebook parent company in a case brought by 13 authors who accused it of illegally using their books to train its AI model, Llama. However, the ruling left the door open for future, more successful claims.

According to Associated Press, US District Judge Vince Chhabria emphasised that his ruling was narrowly focused, applying only to the plaintiffs in this case and not granting Meta broad legal cover. 

"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." 

It was not a definitive summary judgment on whether Meta's actions were legal under copyright law but rather a procedural loss for the authors due to weak arguments and insufficient evidence.

The lawsuit, brought by prominent writers including Sarah Silverman, Ta-Nehisi Coates, Jacqueline Woodson, and Junot Díaz, alleged that Meta had scraped their copyrighted books from pirate websites—so-called "shadow libraries" — to train its generative AI system without paying or obtaining permission. 

Chhabria acknowledged the seriousness of such accusations, stating, "In many circumstances, it will be illegal to copy copyright-protected works to train generative AI models without permission."

While Meta argued that the outputs of Llama are "quintessentially transformative" and do not replace the authors' works, the judge did not fully accept that defence. Instead, he indicated that a stronger case could be made by other plaintiffs in the future. 

"Although the devil is in the details, in most cases the answer will likely be yes" to the question of whether AI firms are engaging in unlawful use of copyrighted material, Chhabria said.

The ruling comes just two days after a related case was decided in the same San Francisco courthouse. 

Also Read: US judge vindicates Anthropic in copyright case

District Judge William Alsup found that Anthropic did not break copyright law by training its Claude chatbot on millions of books, calling the AI's use of those texts "quintessentially transformative" under the fair use doctrine. 

However, Alsup allowed that case to proceed to trial over the issue of whether Anthropic acquired the training data illegally from pirate sources rather than legitimate vendors.

Chhabria, in contrast, took issue with Alsup's focus on transformative use while "brushing aside concerns about the harm it can inflict on the market for the works it gets trained on." 

The judge suggested that economic harm to authors could be demonstrated and used as the basis for successful copyright claims against AI companies.

Meta contended that no one could use Llama to read or replicate the original works, citing that the system doesn't output verbatim content and serves a fundamentally different purpose. 

However, internal company discussions, disclosed during discovery, revealed that executives, including CEO Mark Zuckerberg, were aware of the risks associated with using pirated content for training purposes.

The authors' lawyers argued that Meta's actions were knowingly unlawful. "Whatever the merits of generative artificial intelligence, or GenAI, stealing copyrighted works off the Internet for one's own benefit has always been unlawful," they wrote.

Ultimately, Chhabria's 40-page ruling tossed the case but signaled that future lawsuits—if better argued and supported by evidence—may be more successful. 

He concluded, "This is not a class action, so the ruling only affects the rights of these 13 authors — not the countless others whose works Meta used to train its models."