In a bid to train its expansive large language models, social media behemoth Meta downloaded a staggering 100-plus terabytes of pirated e-books through BitTorrent from covert library platforms like Z-Library and LibGen. In a legal action brought forth by book authors, Meta's legal representatives contended that both uploading pirated e-books to unfamiliar parties via BitTorrent and the act of file-sharing during the download process are covered under the umbrella of fair use. Meta further underscored the significance of these data in cementing the United States' preeminence in the global AI landscape, asserting that utilizing BitTorrent for file-sharing is an indispensable method for data acquisition. Despite a prior court ruling that deemed the use of pirated e-books for training large language models as fair use, Meta is still facing scrutiny for downloading and sharing e-books, as book authors maintain that Meta was complicit in copyright infringement.