Meta ceo Mark Zuckerberg himself knew that the books used to train the company’s AI tool were pirated, according to newly-public documents in one of the California class-action lawsuits against the tech company. Attorneys for a group of author plaintiffs (including Richard Kadrey, Sarah Silverman, Andrew Sean Greer, Ta-Nehisi Coates, and Jacqueline Woodson) assert that Zuckerberg approved the use of a dataset from shadow library LibGen to train their Llama LLM, despite concerns from other employees about it provenance. “Meta has treated the so-called ‘public availability’ of shadow datasets as a get out of jail free card, notwithstanding that internal […]
