Meta, the social media giant, is facing a copyright infringement lawsuit over its use of a pirated book dataset to train its AI models. The authors of the copyrighted works claim that the company’s decision to torrent terabytes of pirated books was unlawful and constitutes direct copyright infringement.
The lawsuit argues that Meta knew that downloading pirated content through torrents was previously deemed infringing, yet it took a risk by bulk-ing up its collections in order to catch up with rival tech firms. To cover its tracks, the company allegedly used Amazon Web Services to conceal its activities.
According to the authors, even if Meta limits seeding when downloads are completed, the act of “leeching” during the download process suggests that it was sharing pirated content with others. The authors claim this constitutes direct copyright infringement and demand summary judgment in their favor.
Source: https://arstechnica.com/tech-policy/2025/03/meta-mocked-for-raising-bob-dylan-defense-of-torrenting-in-ai-copyright-fight