Judge Dismisses Key Claims in Sarah Silverman’s AI Copyright Lawsuit Against Meta

U.S. District Judge largely dismissed Sarah Silverman’s lawsuit against Meta on copyright infringement, building on findings from another federal judge overseeing a lawsuit against AI art generators.

U.S. District Judge Vince Chhabria has dismissed most of comedian and actress Sarah Silverman's lawsuit against Meta, regarding the use of copyrighted books to train its generative artificial intelligence (AI) model. This ruling marks another significant development in the ongoing legal debates over intellectual property rights in the age of AI.

Silverman, along with other authors, had accused Meta of copyright infringement, asserting that the company's AI system, specifically the LLaMA model (Large Language Model Meta AI), constituted an infringing derivative work. They argued that this system, used for various applications including chatbots, was built upon information extracted directly from copyrighted materials.

However, Judge Chhabria has rejected this core theory, stating in his order that viewing the LLaMA models as a recasting or adaptation of the plaintiffs' books was "nonsensical." He further dismissed the argument that every output produced by Meta's AI tools should be considered a copyright infringement, due to a lack of evidence that these outputs transformed or adapted the original books. He emphasized that to prove derivative infringement, there must be a substantial similarity between the AI's outputs and the original copyrighted material.

This decision builds upon a similar ruling by U.S. District Judge William Orrick in a case involving artists suing AI art generators. Orrick questioned the feasibility of proving copyright infringement when the material created by AI tools isn't identical to the copyrighted content. The ruling also addressed other claims, including unjust enrichment and violation of competition laws, dismissing them largely because they were dependent on the surviving copyright infringement claim.

Nevertheless, the judge allowed Silverman and her legal team a chance to replead this claim, along with five others. Meta did not dispute the allegation that the copying of books for AI training could constitute copyright infringement, leaving some room for further legal exploration.

As the legal battles continue, the boundaries of copyright law in the age of AI remain a contested and evolving area. Meta has yet to comment on the ruling, and Sarah Silverman has also joined a class action against OpenAI on similar grounds, demonstrating the growing unease among creators regarding the use of their work in AI development.