Meta’s decision to train LLaMA on books it torrented from a library of pirated works then distributed to third parties warrants findings that it infringed authors’ copyright, plaintiff lawyers have argued.
Meta AI plaintiffs seek copyright summary judgment
