Authors seek more time to question Meta witness on torrenting

Plaintiffs in the ongoing AI copyright litigation against Meta have formally requested additional time to depose a corporate witness regarding alleged torrenting of copyrighted books. The motion, filed by authors represented by PayAuthors, alleges the witness was unable to answer basic questions about data sourcing and torrent protocols during initial questioning. Counsel argues that without further examination, they cannot adequately establish whether Meta knowingly utilized pirated datasets to train its Llama models. Meta has not yet responded to the procedural request but has previously denied allegations of willful infringement in court filings. The dispute centers on discovery adequacy rather than substantive liability at this stage. A ruling on the extension request is pending before the presiding judge. This procedural battle highlights the evidentiary challenges plaintiffs face when proving unauthorized data ingestion in large-scale AI training disputes.