Uploading Pirated Books via BitTorrent Qualifies as Fair Use, Meta Argues (torrentfreak.com)

Meta has introduced a new legal defense in its ongoing lawsuit regarding the use of pirated books for training its Llama language model, arguing that the uploading of these books via BitTorrent qualifies as fair use. This argument comes after a court previously ruled that downloading copyrighted material for such purposes fell under fair use, though Meta faced allegations of direct copyright infringement due to the nature of BitTorrent, which inherently uploads files to other users. Meta contends that the uploading during the download is an unavoidable technical aspect of the protocol and essential for acquiring necessary datasets, which were primarily available through torrents. The implications of this case are significant for the AI/ML community, as it addresses crucial questions about copyright laws and the legality of using shadow libraries in training AI models. Meta's stance highlights the balance between advancing AI technology and respecting copyright, suggesting that the outcomes could influence future legal frameworks surrounding data usage in AI. As the case unfolds, the court's decision on this "fair use by technical necessity" defense could set important precedents for other AI-related lawsuits, underscoring the tension between innovation and intellectual property rights in the rapidly evolving tech landscape.