In one of the AI lawsuits faced by Meta, the company stands accused of distributing pirated books. The authors who filed the class-action lawsuit allege that Meta shared books from the shadow library LibGen with third parties via BitTorrent. Meta, however, says that it took precautions to prevent ‘seeding’ content. In addition, the company clarifies that there is nothing ‘independently illegal’ about torrenting.
Your example is exactly why meta didn’t seed.
It’s a weak defense because the clients still exchanged metadata with other clients, plus there’s the big issue of using the copyrighted works for their own profit, and not just archiving/preservation/personal use
It’s a solid defense, since the lawsuit’s about the sharing of the books. The metadata of the torrents isn’t part of the relevant IP, and how they used the content they downloaded is a separate issue.
Exchanging metadata does not violate copyright. I can stand on the sidewalk in front of MPAA headquarters, asking people to provide me with a copy of the latest blockbuster without infringing on anyone’s copyright. I can even offer to buy it.
So long as they are not further copying or distributing the work, they are not infringing on copyright.
Using it for their own profit is only an issue if they are claiming a fair use exemption. They aren’t claiming fair use. They don’t need to claim fair use, because requesting and receiving works is not a prohibited act under copyright law. The prohibited acts are copying and distribution, which can only be performed by the sender, not the receiver.