Overview
- An amended class-action complaint alleges a Nvidia data strategy staffer contacted Anna’s Archive for high‑speed access to its collection to pre‑train large language models.
- The filing says Anna’s Archive warned the trove was illegal, Nvidia management approved proceeding within a week, and roughly 500 terabytes of data were offered.
- Plaintiffs also allege Nvidia pulled from other shadow‑library sources including LibGen, Sci‑Hub, Z‑Library, and facilitated downloads of The Pile via scripts shared with customers.
- The suit expands legal theories to direct, contributory, and vicarious infringement and adds named authors including Abdi Nazemian, Brian Keene, Stewart O’Nan, Andre Dubus III, and Susan Orlean.
- Reporters note this appears to be the first public disclosure linking a major U.S. tech company to Anna’s Archive, the complaint does not state whether any payment was made, and Nvidia’s fair‑use defense remains on the record as the case proceeds in the Northern District of California.