US District Judge Vince Chhabria ruled in favour of Meta in a fair use case, dismissing the plaintiffs’ claims of market harm related to the company’s use of their works in training the Llama AI. The case involved several well-known authors, including Sarah Silverman, Ta-Nehisi Coates and Junot Diaz, who accused Meta of infringing their copyrights by using their books without permission. Judge Chhabria rejected the plaintiffs’ argument that Meta’s use of their works harmed the market for their books, criticising their failure to provide meaningful evidence to support their claims.
The plaintiffs' main arguments were that Meta’s AI, Llama, could reproduce snippets of their works and that Meta's unauthorised use of their books had diminished their ability to license their works for training AI models. Judge Chhabria dismissed both points, calling them unconvincing.
While the court ruled in Meta's favour on fair use, the authors’ claims regarding the downloading of pirated books and copyright infringement are still pending. Judge Chhabria also indicated that Meta’s use of copyrighted materials for AI training was transformative, a key factor in fair use decisions, but emphasised the importance of analysing market harm when assessing fair use.
Meta welcomed the decision, emphasising that the fair use of copyrighted material is vital for the development of transformative AI technologies. However, the case is not final and appeals are expected as it progresses through the courts.
