A federal judge has allowed an artificial intelligence-related copyright lawsuit against Meta to proceed but dismissed some of the claims. In Kadrey v. Meta, authors including Richard Kadrey, Sarah Silverman, and Ta-Nehisi Coates accused Meta of infringing their intellectual property rights by using their books to train its Llama artificial intelligence model, and that the company also removed copyright information from their books to cover up the alleged infringements.
Meta, meanwhile, claimed that its training qualified as fair use and argued that the case should be dismissed because the authors lacked standing to sue. During a court hearing last month, U.S. District Judge Vince Chhabria appeared to indicate that he opposed the dismissal, but he also criticized the "excessive" rhetoric of the author's legal team.
In Friday's ruling, Chhabria wrote that the copyright infringement claim "is clearly a concrete injury sufficient to constitute standing" and that the authors also "substantially allege that Meta intentionally removed CMI (Copyright Management Information) to conceal the copyright infringement."
"Taken together, the allegations raise a 'reasonable, if not particularly strong, inference' that Meta removed the CMI in an attempt to prevent Llama from exporting the CMI, thus revealing that it was trained on copyrighted material," Chhabria wrote.
However, the judge dismissed the authors' California Comprehensive Computer Data Access and Fraud Act (CDAFA) claim because they did not "allege that Meta accessed their computers or servers, only their data (in the form of books)."
The plaintiffs claim in court documents that Mark Zuckerberg allowed the Llama team to use copyrighted works to train models, and that other Meta team members also discussed using legally questionable content for AI training.
Courts are currently weighing multiple artificial intelligence copyright lawsuits, including the New York Times’ lawsuit against OpenAI.