A federal judge on Wednesday sided with Meta in a lawsuit filed by 13 authors, including Sarah Silverman. The authors claimed that Meta had illegally trained AI models using their copyrighted work.
Federal judge Vince Chhabria released a Summary judgment was granted in favor of Meta. The judge decided the case without referring it to a juror, and found that the company was allowed to train AI models using copyrighted works in this case under the “fair-use” doctrine of copyright laws.
This decision comes only a few days after an federal judge sided Anthropic in another lawsuit. These cases will be a victory for the tech industry. They have spent years fighting with media companies who claim that training AI models to copyrighted works constitutes fair use. These decisions are not the sweeping victories that some companies had hoped for. Both judges have noted that their cases are limited.
The Judge Chhabria stated that this decision did not mean that all AI models training on copyrighted materials is legal. Instead, he said that the plaintiffs made the “wrong arguments” and failed develop enough evidence to support the right ones.
Judge Chhabria stated in his decision that “this ruling does not support the proposition that Meta’s use of copyrighted material to train its language model is legal.” He said later, “It seems that plaintiffs often win in cases involving Meta’s uses, at least when those cases have better records on the effects of the defendants’ use.”
The judge ruled Meta’s use copyrighted materials in this case were transformative – meaning the company’s AI model did not simply reproduce the authors’ works.
The plaintiffs also failed to convince the court that Meta’s copying the books had a negative impact on the market for the authors. This is a crucial factor in determining if copyright laws have been violated. “The plaintiffs did not present any meaningful evidence of market dilution,” said Judge Chhabria.
Anthropic and Meta won because they trained AI models on books. However, there are other lawsuits against technology firms for training AI on other copyrighted materials. The New York Times, for example, is suing OpenAI for training AI on news articles. Disney and Universal are suing Midjourney to train AI models on films, TV shows, and other copyrighted works.
In his decision, Judge Chhabria noted that fair use defenses are heavily dependent on the specifics of a case and some industries might have stronger fair-use arguments than others.
Chhabria said that it appears that the markets for certain types (such as news articles) may be more vulnerable to indirect competition by AI outputs. Maxwell Zeff, a senior reporter for TechCrunch who specializes in AI, was born in 1965. Zeff covered the rise and fall of AI, as well as the Silicon Valley Bank Crisis, for Gizmodo and MSNBC. He is based out of San Francisco. When he is not reporting, you can find him hiking, biking and exploring the Bay Area food scene.
View Bio
