The complaint accused the tech company of using adult films to assist its AI model training. Similar claims have been made by other rightsholders, including many\u00a0book authors<\/a>.<\/p>\n This latest case, with over 350 million dollars in potential damages, specifically focuses on Meta\u2019s BitTorrent activity that was recorded in detail through proprietary torrent tracking software. That\u2019s no surprise, as plaintiff Strike 3 is the\u00a0most active<\/a>\u00a0copyright litigant in the United States, known for targeting thousands of\u00a0alleged BitTorrent pirates<\/a>\u00a0based on similar evidence.<\/p>\n Meta responded in October by filing a\u00a0motion to dismiss<\/a>, arguing the sporadic downloads were consistent with ordinary\u00a0\u2018personal use\u2019<\/a>\u00a0by employees and visitors on the corporate network. It was certainly not a coordinated AI training effort, Meta countered.<\/p>\n The motion to dismiss remains pending and, meanwhile, the case is heating up in other areas. Last week, the parties filed their joint discovery plan, which Strike 3 used to raise a rather eye-popping allegation.<\/p>\n Meta said that it prefers to delay written evidence discovery requests in this case until the court ruled on its motion to dismiss. However, Strike 3 would like to start gathering evidence right away, fearing that key data may otherwise disappear.<\/p>\n Strike 3\u2019s legal team points out that, at a February 5 hearing in the unrelated\u00a0Kadrey v. Meta<\/a>\u00a0book-authors case, lawyers revealed that a Meta employee had recently deleted over nine terabytes of torrented files. Fearing more deletions, Strike 3 asks the court to allow discovery in the present case to begin immediately.<\/p>\n \u201cBecause of the tangible risk that relevant evidence may be deleted by Meta\u2019s employees, Plaintiffs respectfully request that they be allowed to conduct discovery immediately,\u201d the plaintiffs write.<\/p>\n <\/p>\n In the same filing, Meta\u2019s legal team immediately tried to defuse the deletion claim. Meta says that no data was spoiled and clarified that it will preserve all evidence as it is legally obliged to do.<\/p>\n \u201cPlaintiffs mischaracterize the Kadrey record. There was no spoliation in Kadrey, which is an unrelated case, and in any event Meta has an appropriate hold in place and is abiding by its preservation obligations,\u201d Meta writes.<\/p>\n The discovery plan also provides the clearest picture yet of what Strike 3 actually wants to find. Among the targets is Meta\u2019s Machine Learning Hub \u201cML Hub,\u201d including downloaded digital media files, torrenting-related metadata, and labeling data for content acquired from BitTorrent.<\/p>\n Strike 3 also wants logs of Meta servers communicating over \u201cPySpark or Fairspark protocols,\u201d suggesting it believes these tools were used to coordinate downloads across Meta\u2019s infrastructure. Separately, the company is seeking records tying Meta\u2019s alleged hidden \u201coff-infra\u201d IP addresses to Amazon Web Services instances.<\/p>\n The discovery list is broad by design, and the above are just a few examples. In essence, Strike 3 wants all policies, directives, and algorithms related to torrenting. They hope that this information will help to back up their copyright infringement claims.<\/p>\n While Strike 3 references thousands of downloads, Meta stresses that the complaint only mentions 157 downloads from Meta\u2019s corporate IP addresses over seven years. They note that this is illustrative of personal use, rather than an organized data collection effort.<\/p>\n Meta also explains that the alleged downloads began years before it started researching generative video AI, making a coordinated training effort even more implausible. In addition, Meta says that Strike 3 has \u201cno facts whatsoever\u201d linking it to the thousands of additional third-party IP addresses that are named in the complaint.<\/p>\n While Meta\u2019s motion to dismiss is still unresolved, both parties are also looking ahead. While they differ on the exact timing of various deadlines, both believe that an eventual trial can take place in the first half of 2028, if it gets to that.<\/p>\n \u2014<\/em><\/p>\n A copy of the parties\u2019 26(f) discovery plan, filed at the U.S. District Court for the Northern District of California, is\u00a0available here (pdf)<\/a>. We will add a copy of the transcript as soon as we notice that it is publicly posted.<\/em><\/p>\n Source: \u00a0\u2018Meta Employee Deleted 9TB of Torrented Files\u2019<\/h2>\n
![\"kadrey](\"https:\/\/torrentfreak.com\/images\/kadreydeleted.png.webp)
<\/picture><\/center> <\/p>\nTorrent Evidence<\/h2>\n
Meta\u2019s Defense & Trial Date<\/h2>\n
![\"TorrentFreak\"](\"http:\/\/worldjusticenews.com\/news\/wp-content\/uploads\/2016\/11\/torrentfreak.png\")
<\/a> TorrentFreak.com<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"