June 26 (UPI) — A federal judge let Meta off the hook for the use of books to train its artificial intelligence model, but it still might face legal challenges by the authors in other ways.
United States District Judge Vince Chhabria ruled Wednesday that it wasn’t unlawful for Meta to feed copyrighted material to its large language models, or “Llama,” because of the doctrine of “fair use.”
Chhabria wrote in his ruling that the use of copyrighted works by companies to train generative AI models, are using the materials in a creative fashion, which in copyright law is considered a “transformative” use as the Copyright Act says the use of copyrighted materials to teach, research, criticize, comment or report news is not infringement.
Conversely, the judge also noted that creating a new product by way of copying a protected work and then marketing that product could potentially harm the markets that carry the original materials.
“So by training generative AI models with copyrighted works, companies are creating something that often will dramatically undermine the market for those works,” Chhabria wrote.
He then questioned if companies that feed copyright-protected materials into their AI platforms without first getting permission from the copyright holders or paying to use the originals are doing something illegal.
“Although the devil is in the details, in most cases the answer will likely be yes,” wrote the judge.
Chhabria then referred to a similar case that concluded on Monday in which U.S. District Judge William Alsup ruled the Anthropic artificial intelligence company didn’t violate any copyright laws when it used millions of copyrighted books to train its own AI.
“Judge Alsup focused heavily on the transformative nature of generative AI while brushing aside concerns about the harm it can inflict on the market for the works it gets trained on,” Chhabria wrote.
“No matter how transformative [Llama] 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,” Chhabria wrote. “While enabling the creation of a potentially endless stream of competing works that could significantly harm the market for those books.”
However, part of the lawsuit filed by the authors also alleges that Meta had removed copyright management information, which would violate the Digital Millennium Copyright Act, or DMCA, and this will be overseen separately.
A case management conference on how the court will proceed in regard to the DMCA is scheduled for July 11.