Meta Denies Sharing Pirated Books for AI Training

Meta claims it didn't seed a torrent of pirated books used for AI training, despite admitting to downloading it. In a court filing, Meta stated it took precautions to avoid seeding the downloaded files, arguing authors can't prove distribution occurred during the torrenting process. While admitting to downloading the dataset from sources like LibGen and Z-Library, Meta contends downloading itself isn't illegal, merely accessing publicly available data. This case involves copyright infringement claims, with authors alleging Meta engaged in large-scale data piracy and violated California's Computer Data Access and Fraud Act (CDAFA).