Ex-Meta Researcher Torrented Pirated Books, Deleted It 'Due to Litigation'

Meta faces multiple lawsuits over using pirated books from the Books3 dataset to train its Llama AI models, with former researcher Stephen Roller testifying during depositions that he torrented the dataset while employed at Meta between 2017 and 2023. Roller has since deleted his copy of Books3, citing litigation concerns, and the plaintiffs claim Meta possessed multiple copies of the pirated dataset, though Meta argues Roller's download may have been personal rather than for the company. The case centers on how Meta obtained Books3, as a court previously ruled that training on the books constituted fair use.