A U.S. magistrate judge has ordered Meta to identify people involved in decisions to acquire shadow-library material for AI training, expanding disclosure in a copyright case brought by Entrepreneur Media, LLC.
The Oct. 19 discovery order from U.S. Magistrate Judge Thomas S. Hixson requires Meta to give approximate dates for each person’s involvement and connect them to the decisions and datasets at issue. It also covers choices about obtaining LibGen Scimag and Anna’s Archive material through BitTorrent, including whether and how to seed files or restrict seeding. A torrent client may upload file pieces to peers during a download or afterward.
Meta told the court that decisions were often made by teams and that it knew of no individual who decided to upload or distribute the datasets. Hixson said Meta must identify participants even when a group made the decision. He also found that Meta’s earlier response focused more on how it used the material than on who decided to obtain it.
The order resolves a dispute over information Meta must provide, not whether it infringed copyright. The cited section sets no deadline for Meta’s answers. In a separate case brought by authors, a judge ruled on June 25, 2025, that Meta’s copying of the named plaintiffs’ books to train AI models was fair use on the record in that case. A claim alleging distribution through torrenting remained unresolved.
