Judge rejects Meta's claim that torrenting is "irrelevant" in AI copyright case
10 months ago
- #Meta
- #AI-training
- #copyright
- Meta has largely beaten an AI training copyright lawsuit raised by 13 book authors, including Sarah Silverman and Junot Diaz.
- The remaining issue is whether Meta violated copyright laws by torrenting books to train Llama models.
- Judge Vince Chhabria confirmed a meeting on July 11 to discuss the torrenting claim.
- Authors may struggle to win due to lack of evidence, but Meta's torrenting isn't irrelevant to fair use.
- Torrenting over 80.6TB from LibGen is 'at least potentially relevant' to fair use analysis.
- Meta's decision to pirate books from shadow libraries is relevant to 'bad faith' in fair use.
- Authors argued Meta abandoned licensing talks and used BitTorrent to acquire books instead.
- Judge noted uncertainty in law about whether bad faith affects fair use.