Meta CEO Mark Zuckerberg has invoked YouTube's approach to copyright enforcement as part of his defense in an ongoing AI copyright case, Kadrey v. Meta. He argues that using a dataset of copyrighted e-books for training AI models may be justified as 'fair use.' Despite internal concerns, Zuckerberg reportedly approved the use of the LibGen dataset for Meta's Llama AI models. New allegations suggest Meta used LibGen to assess licensing agreements by cross-referencing pirated e-books with available titles. The case highlights the ongoing legal tensions between AI development and intellectual property rights.