Judge rejects Meta’s claim that torrenting is “irrelevant” in AI copyright case

Meta largely prevailed in a key AI training copyright lawsuit from 13 authors, though a specific claim regarding its alleged unlawful torrenting of copyrighted books to train Llama models remains unresolved. A July 11 meeting is set to address this issue, with Judge Vince Chhabria indicating authors may face evidentiary challenges but also ruling that Meta's torrenting, potentially involving 80.6 terabytes from shadow libraries, is "at least potentially relevant" to "bad faith" in fair use analysis. This stems from allegations Meta resorted to piracy after failing to secure licensing deals, highlighting residual legal and reputational risks for the company and setting a precedent for AI data acquisition strategies amidst evolving fair use interpretations.

Analysis

Meta Platforms has achieved a significant legal victory by largely defeating a copyright infringement lawsuit from authors, substantially de-risking a key component of its Llama AI model development. The court's partial summary judgment in favor of Meta mitigates a major potential liability concerning the use of copyrighted books for AI training. However, a material overhang remains from the unresolved claim that Meta unlawfully torrented these works. A July 11 meeting is scheduled to address this specific issue, with Judge Vince Chhabria indicating that while plaintiffs may face evidentiary hurdles, Meta's alleged use of BitTorrent to acquire over 80.6 terabytes of data from shadow libraries is "at least potentially relevant" to establishing "bad faith." This is critical because the claim of bad faith arises from allegations that Meta resorted to piracy only after failing to secure content licenses. While the judge noted the law regarding bad faith's relevance to fair use is currently "in flux," this lingering claim represents a notable legal and reputational risk, highlighting the contentious data acquisition methods underpinning some large-scale AI projects.

Key Decisions for Investors

• The partial dismissal of the core copyright claim should be seen as a positive catalyst, reducing a major legal uncertainty surrounding Meta's AI strategy and potentially strengthening the investment thesis for its long-term AI monetization.
• Investors should monitor the outcome of the July 11 hearing on the torrenting allegation, as a finding of "bad faith" could create reputational damage and influence future regulatory scrutiny of AI data sourcing, even if direct financial penalties are minor.
• The case's resolution will set an important precedent for the AI industry's data acquisition practices, and a complete victory for Meta would further solidify the legal standing of a core operational strategy for large language model developers.