Meta tries to scapegoat staff who downloaded torrents to spare execs
Ars Technica Ashley Belanger ● Covered by 9 sources
Meta is being pushed to identify the execs behind its book-piracy claims in AI training lawsuits. A judge says it can’t hide the chain of command behind staff who downloaded torrents.
Based on reporting by Ars Technica, Ashley Belanger — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
Meta has spent weeks trying to pin its AI training copyright mess on rank-and-file staff who downloaded torrents. That story is getting harder to sell. A court order now pushes the company to name the executives who approved the decision to pirate books from shadow libraries such as Anna’s Archive and LibGen, according to TorrentFreak.
The ruling affects a class action brought by book authors including Richard Kadrey and Sarah Silverman, along with a separate case from publisher Entrepreneur Media. Magistrate Judge Thomas Hixon largely sided with the plaintiffs, saying Meta has to make the key decision-makers clear rather than leaving the chain of command blurry.
That matters because testimony from Meta employees already pointed upward. Staffers who downloaded torrents said supervisors and managers had greenlit the piracy, which is exactly the layer Meta seems most eager to keep out of view. The order could widen the list of names that end up in the record, including people involved in early AI training.
Meta’s own court filing tried to recast the whole operation as an “informal approach,” with no one in authority supposedly making decisions at all. In that telling, a “decision” existed only because some employees technically downloaded torrents. It’s a neat bit of corporate fog. Judges tend not to love fog.
The larger fight here isn’t really about torrents. It’s about who gets blamed when AI companies build giant training sets from material they don’t want to talk about too openly. Meta is betting on a very old Silicon Valley trick: if the paper trail gets uncomfortable, pretend the paper trail was never a real process.
My take — AI-written commentary, not fact-checked reporting
The best part of these cases is how quickly the “no one made decisions” defense turns into a scavenger hunt for adults in the room. Meta is discovering that ‘informal approach’ is not a magic phrase, just a very expensive one. When the training data is shady, the management structure usually is too.
Read more about this at: Ars Technica
Related stories
‘Feeding the beast’: Meta faces allegations of using its ‘perv glasses’ to train AI in new lawsuit
Fortune ·
51
Why is Meta destroying its engineering organization?
The Pragmatic Engineer · 4 months ago ·
46