Earlier this year Strike 3 Holdings and Counterlife Media filed a federal lawsuit against Meta, alleging the tech giant pirated 157 adult films via BitTorrent to train its AI video generation models. The plaintiffs (who operate major adult brands including Blacked, Tushy, Vixen, and Deeper) claim Meta used 47 corporate IP addresses to download their copyrighted content between 2018 and 2025, then fed this material into Meta’s “Movie Gen” AI suite and other generative video projects.

Meta has filed a motion to dismiss, arguing that linking IP addresses to downloads fails to prove direct corporate involvement. This defense strategy highlights evidentiary challenges facing content producers in AI copyright cases and could become a template for other tech companies facing similar claims.

The case tests whether traditional BitTorrent piracy enforcement methods can adapt to AI training contexts, where the alleged infringement involves data processing rather than redistribution. Key legal questions include whether IP address evidence meets the heightened proof standards courts increasingly demand in corporate piracy cases, how fair use defenses apply when adult content trains AI models that explicitly prohibit pornographic outputs, and whether adult studios can claim the same licensing market harm as mainstream media companies in AI training disputes.

This lawsuit emerges amid intensifying legal battles over AI training data, following recent high-profile cases like Authors Guild v. OpenAI and Getty Images v. Stability AI. The adult industry faces unique hurdles in these disputes since many platforms ban adult material, limiting the ability to detect unauthorized AI outputs, while mainstream legal precedents may not fully address the industry’s specific licensing models and distribution challenges.

The case builds on the landmark Authors Guild v. Google decision, where courts established that fair use analysis for large-scale digitization must weigh transformative purpose against market harm. More recently, in Andy Warhol Foundation v. Goldsmith (2023), the Supreme Court emphasized that commercial use weighs heavily against fair use — a principle that could undermine Meta’s potential defense that AI training constitutes transformative use.

The evidentiary challenges highlighted by Meta’s motion to dismiss reflect broader enforcement difficulties emerging across AI training disputes. Recent cases involving OpenAI and book publishers have similarly struggled with proving direct corporate involvement in data acquisition, while comedian Sarah Silverman’s parallel suit against Meta faces comparable hurdles in demonstrating how copyrighted works were specifically used in training datasets.

The case also highlights the adult industry’s unique position in AI development; while Meta claims its terms prohibit pornographic outputs, the underlying training process may still benefit from exposure to adult content’s cinematography, lighting, and production techniques. This creates a complex dynamic where adult content adds value to AI systems even when explicit outputs are restricted.

The discovery phase, if the case survives dismissal, could expose internal Meta documents about AI training practices. Understanding how major tech companies source and process training data will be crucial for content producers developing licensing frameworks and enforcement strategies.

The case emerges as regulatory scrutiny intensifies over AI training practices:

• FTC Investigation: The Federal Trade Commission is actively investigating major AI companies’ data collection practices, including how they acquire training datasets, with particular scrutiny on whether companies adequately disclose their data sources to consumers and content creators.
• Congressional AI Oversight: House Judiciary Committee hearings have focused on copyright implications of AI training, with pending legislation like the COPIED Act that would require AI companies to disclose training data sources and obtain explicit consent for copyrighted material.
• DOJ Copyright Enforcement: The Department of Justice has signaled increased enforcement priority for large-scale copyright infringement cases involving emerging technologies, particularly where commercial entities systematically acquire content without authorization.

Recent regulatory scrutiny from the FTC over AI training transparency, combined with the EU’s emerging AI Act requirements for dataset documentation, suggests courts may become more receptive to claims demanding disclosure of training methodologies and data sources. The adult entertainment industry’s aggressive copyright enforcement history, including thousands of successful BitTorrent lawsuits, positions these plaintiffs uniquely well to challenge tech companies’ data acquisition practices.

For industry operators, this lawsuit signals the need for proactive strategies beyond traditional enforcement. Content producers may need to audit existing licensing agreements to explicitly address AI training uses, implement more sophisticated tracking beyond IP addresses, and develop standardized AI training licenses that allow monetization while maintaining control over how premium content is utilized in machine learning applications.