Judge Orders Meta to Identify Employees Behind Torrenting Shadow Libraries
Meta faces legal pressure to name staff who authorized BitTorrent usage for Llama training.
1Court Mandates Employee Identification
Magistrate Judge Thomas Hixson has formally ordered Meta to disclose the identities of employees who approved the use of BitTorrent to acquire shadow library datasets. This ruling follows ongoing litigation from Entrepreneur Media, which accuses the tech giant of contributory copyright infringement by seeding protected works to third-party users during data collection.
2Meta Claims Consensus Over Authority
Meta has struggled to name specific individuals responsible for the decision to use torrenting protocols for Llama model training. The company argues that these choices were the result of team consensus rather than individual authorization. They further claim that the technical execution of downloads was handled by employees as routine operational tasks.
3Internal Warnings Regarding Seeding
An internal Meta logbook explicitly warned that utilizing torrents would involve sharing copyrighted material externally. The document noted that seeding files constitutes sharing content outside the organization, which was flagged as legally problematic. Despite these internal warnings, the company proceeded with the acquisition of datasets from Anna's Archive and LibGen via BitTorrent.
4Legal Review and Data Acquisition
Meta maintains that its data acquisition from Anna's Archive underwent an internal legal review in early 2024 involving in-house counsel. However, the company has failed to identify any high-level executives who formally approved the specific use of BitTorrent. This lack of clear accountability remains a central point of contention in the current legal proceedings.
5Quantifying the Seeding Impact
Meta’s own expert witness estimated that uploads from its AWS servers accounted for approximately 30% of the total data downloaded during a specific period in 2024. While Meta claims they implemented measures to limit seeding, the court remains unconvinced by these assertions, demanding further clarity on how the distribution of copyrighted material was managed.
6Challenging the Corporate Narrative
Entrepreneur Media has countered Meta’s defense, noting the absurdity of the claim that no one at the company decided to torrent anything despite repeated activity from 2022 to 2024. The publisher insists that the systematic use of shadow libraries requires clear internal authorization, rejecting the notion that these actions occurred without specific management oversight.
7Scope of the Judicial Order
Judge Hixson’s order requires Meta to identify every individual involved in the decision-making process regarding the downloads. This includes those who determined whether or how to seed the files, as well as those responsible for mitigating the distribution of copyrighted content. The court seeks to establish a clear chain of responsibility for these actions.
8Implications for AI Training Data
This case highlights the growing legal risks associated with sourcing training data for large language models from unauthorized repositories. As courts demand transparency regarding data provenance, companies face increased scrutiny over their internal processes. The outcome of this dispute could set a significant precedent for how tech firms handle copyright compliance in AI.