Overview
- Magistrate Judge Thomas Hixon issued the discovery order requiring Meta to identify every person who made, authorized, or approved the use of BitTorrent for pulling shadow-library datasets, with the order dated Oct. 20, 2026.
- The command reaches two Northern District of California cases — the Kadrey v. Meta class action by authors and a separate suit from publisher Entrepreneur Media — and follows a prior September order that compelled server command-history logs for machines used in the torrenting.
- Plaintiffs say the identities and server logs could prove that Meta seeded (uploaded) large volumes of copyrighted books during 2022–2024, which would support claims of contributory or willful infringement because BitTorrent both downloads and uploads pieces of files.
- Meta has defended the practice as an informal, team-level technical approach without formal executive sign-off, while internal messages and staff testimony disclosed in the suits say some downloads required director- or VP-level approval and showed employee discomfort.
- The disclosure fight shifts the litigation from whether model training is fair use to how Meta acquired and handled the data, and the evidence could affect corporate accountability and industry norms for building AI training sets.