Meta secured a significant legal victory on Wednesday as a U.S. federal judge dismissed a high-profile copyright lawsuit brought by 13 authors, including Sarah Silverman, Ta-Nehisi Coates, and Junot Díaz, over the tech giant's use of their books to train its artificial intelligence system. The ruling marks the second win in a week for the AI industry, following a similar decision in favor of Anthropic.
Judge Vince Chhabria of the U.S. District Court in San Francisco ruled that the plaintiffs failed to present adequate evidence that Meta's AI model, LLaMA, harmed the market for their works. "Meta introduced evidence that its copying hasn't caused market harm," Chhabria wrote. "The plaintiffs presented no empirical evidence to the contrary... All the plaintiffs presented is speculation."
The lawsuit, filed in 2023, alleged that Meta used pirated versions of books to train its large language model without obtaining permission or providing compensation. The authors accused Meta of violating copyright law by pulling content from "shadow libraries" and downloading it via BitTorrent.
While Chhabria found that using copyrighted works to train AI models can be illegal "in many circumstances," he determined that in this particular case the plaintiffs failed to demonstrate that Meta's conduct constituted infringement. He emphasized that the ruling "does not stand for the proposition that Meta's use of copyrighted materials to train its language models is lawful," only that "these plaintiffs made the wrong arguments and failed to develop a record in support of the right one."
Meta welcomed the decision. "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," a company spokesperson said.
Attorneys for the plaintiffs at Boies Schiller Flexner LLP expressed strong disagreement with the decision. "Despite the undisputed record of Meta's historically unprecedented pirating of copyrighted works, the court ruled in Meta's favor. We respectfully disagree with that conclusion," a spokesperson for the law firm stated. The firm declined to comment on whether it would appeal the ruling.
Although the court declined to rule on the legality of Meta's use of torrenting to download the shadow libraries, that issue remains unresolved. The lawsuit is part of a broader wave of litigation by authors, publishers, and copyright holders against AI developers, including OpenAI, Microsoft, and Anthropic.
Judge Chhabria acknowledged the growing tension between creative professionals and AI firms. "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," he said.
However, he concluded that, in this case, the AI model's outputs were sufficiently distinct and transformative, and that the plaintiffs had not proven any meaningful harm to their ability to profit from their original works. "In the grand scheme of things, the consequences of this ruling are limited," Chhabria wrote.
