California federal judge dismisses portion of copyright infringement suit against Meta over Llama training data

A US federal judge in California dismissed part of a copyright lawsuit brought by authors including Sarah Silverman against Meta over its use of copyrighted works to train the Llama AI system. The judge expressed skepticism toward the remaining liability theories, stating "I don't understand even a little bit" the authors' core remaining claims. This follows a similar October ruling by another California judge rejecting copyright claims against Stability AI, Midjourney, and DeviantArt, finding generative AI output insufficiently similar to source works to constitute infringement. The rulings establish emerging precedent that unauthorized training-data use may not meet legal thresholds for copyright infringement, reducing the balance-sheet liability risk for AI vendors from copyright claims at the federal level.