A federal magistrate judge has ordered Meta to name the executives who authorized or approved torrenting copyrighted books for AI training, rejecting the company’s attempt to blame individual staff members. The discovery order, reported by TorrentFreak and covered by Ars Technica, applies to class actions from authors including Richard Kadrey and Sarah Silverman, as well as a separate suit from Entrepreneur Media. Judge Thomas Hixon largely agreed that Meta must clearly identify key decision-makers and produce command history logs for every server used in the torrenting process.
Meta had argued that its early AI dataset collection was managed through an “informal approach” where no one with authority made decisions. The company claimed that a “decision” was made only in the sense that some staffers “technically” downloaded torrents. But internal messages and testimony show staff discomfort, including one worker who said “torrenting from a corporate laptop doesn’t feel right,” and another who said downloads required director or VP approval. At least one unredacted message string allegedly showed that a LibGen download occurred after “a prior escalation to MZ” — Mark Zuckerberg.
The judge was not persuaded by Meta’s attempts to narrow the request, including its claim that it was “not currently aware” of anyone who approved seeding. Authors have argued that Meta’s torrenting of 80.6 terabytes from LibGen and 35.7 terabytes via Anna’s Archive necessarily involved seeding, since torrents depend on swarms. The newly ordered log files could reveal whether Meta’s claimed “mitigations” against seeding actually worked, and could provide the closest thing yet to smoking-gun evidence of willful infringement.