A federal judge has permitted significant portions of an AI copyright lawsuit against Meta to proceed while dismissing other elements of the case. The lawsuit, filed by authors Richard Kadrey, Ta-Nehisi Coates, and Sarah Silverman, alleges Meta violated copyright laws by using their books to train AI models without permission.

US District Judge Vince Chhabria ruled that the authors' copyright infringement claims constitute "obviously a concrete injury sufficient for standing." The judge also found that the authors "adequately alleged that Meta intentionally removed CMI [copyright management information] to conceal copyright infringement."

Internal documents revealed through discovery show concerning discussions within Meta about copyright:

• Meta CEO Mark Zuckerberg allegedly approved training on copyrighted content through "legally questionable" means while halting licensing discussions with publishers
• Research engineer Xavier Martinet suggested purchasing retail ebooks rather than securing proper licensing
• Senior manager Melanie Kambadur proposed using LibGen, a platform previously fined for copyright infringement, to access copyrighted works

While the judge dismissed claims related to the California Comprehensive Computer Data Access and Fraud Act (CDAFA), the core copyright infringement allegations remain intact. The case highlights growing tensions between AI companies and content creators over training data usage rights.

The ruling suggests potential legal consequences for AI companies training models on copyrighted materials without proper authorization, setting an important precedent for future AI copyright cases.

Related Articles

Previous Articles