From Pillow Talk To Court Talk: Meta’s Adult Content Copyright Mess

The legal battle over what data tech giants can use to train artificial intelligence has taken an unexpected turn. While most headline-grabbing AI copyright disputes involve bestselling authors, major news outlets, or record labels, a recent federal court ruling proves that all copyright holders have equal standing when it comes to protecting their digital assets – even the makers of adult movies.

Because of this, on June 11, 2026, U.S. District Judge Eumi K. Lee in the Northern District of California denied a motion to dismiss filed by Meta Platforms, greenlighting a massive $359 million lawsuit brought by adult film producer Strike 3 Holdings and Counterlife Media.

The case, which alleges that Meta systematically pirated thousands of adult movies using BitTorrent to feed its AI models, offers critical insights into how courts are handling tech giants’ efforts to dodge AI-related copyright liabilities at the early pleading stage.

The Case: “Non-Human” Download Patterns on Corporate Networks

Strike 3 Holdings, the parent company behind major adult entertainment brands like Vixen and Blacked, is well known as one of the most aggressive copyright litigants in the United States. Utilizing proprietary digital tracking tools, Strike 3 monitors public BitTorrent networks to identify IP addresses illegally downloading and distributing its copyrighted films.

According to the lawsuit, Strike 3 traced 47 distinct IP addresses directly belonging to Meta’s corporate network. Between 2018 and 2025, these IP addresses were allegedly used to torrent nearly 2,400 of Strike 3’s films over 6,000 times.

The plaintiffs argue that the downloads did not look like employees slacking off on the clock. Instead, the data transfers showed highly coordinated, non-sequential, and “non-human” automated patterns, indicating that Meta was using automated scripts to scrape massive volumes of video data. The plaintiffs allege this content was gathered specifically to train Meta’s video-generation models, like Meta Movie Gen, which require extensive human-centric video data to learn realistic motion and continuity.

The Defense: Meta’s “Personal Use” and Pleading Technicalities

Meta moved to dismiss the lawsuit, calling the allegations legally insufficient. To avoid admitting to corporate-level data theft, Meta argued that even if the downloads occurred, the low volume over a seven-year period resembled “private personal use” by individual employees or network visitors rather than a structured corporate AI training project.

Furthermore, Meta argued that the case should be tossed because Strike 3 failed to specify exactly which AI models were trained on the stolen footage.

Judge Lee rejected Meta’s arguments completely. In her order, the judge noted that Meta was trying to apply a strict pleading standard borrowed from entirely different styles of AI copyright cases. Judge Lee clarified that Strike 3 was not just claiming that the training of the AI was the violation – they were arguing that the act of unlawfully downloading and distributing the files via BitTorrent was the direct infringement.

Because the tracking data plausibly tied the automated downloading behavior back to Meta’s corporate infrastructure, the court ruled that the case could proceed to discovery, where Meta will be forced to answer for what happened on its servers.

As of July 1, 2026, the parties entered into a stipulation that Plaintiff Strike 3 Holdings, LLC will seek leave of court to file an Amended Complaint with the following limited changes: (1) adding allegations of infringements of their motion pictures (“Works”) that occurred after the initial filing of the Complaint; (2) removing one previous allegation of infringement; and (3) making minor edits to the initial Complaint. On July 6, 2026, the Judge signed the Order granting the Stipulation. Plaintiff Strike 3 Holdings, LLC will now file an Amended Complaint, and Meta Platforms, Inc. may file another Motion to Dismiss or file an Answer and Affirmative Defenses.

Why This Matters for Intellectual Property Management

While the subject matter of this case is unique, the legal principles established by Judge Lee’s ruling apply broadly to all commercial businesses navigating the digital marketplace:

The Pleading Bar for AI Misuse Is Shifting. Tech companies have frequently had success dismissing AI lawsuits by arguing that creators cannot prove their exact work was reproduced in a final AI output. This ruling shows a critical shift: if a plaintiff can provide circumstantial or forensic evidence that a company unlawfully acquired and copied data on its corporate network, the suit can survive early dismissal.

Network Activity Equals Corporate Responsibility. Meta’s attempt to chalk up corporate network torrenting to “individual employee personal use” fell flat. Companies are ultimately responsible for the scripts, scraping tools, and data-gathering methods running on their corporate networks. Allowing automated data pirating to occur under a corporate IP address exposes an organization to massive vicarious and direct liability.

The Fair Use Defense Is Not a Blanket Shield. AI developers frequently rely on “fair use” arguments, claiming that scraping data to train models is inherently transformative. However, when the acquisition method involves alleged digital piracy or unauthorized software channels (like peer-to-peer torrenting networks), the legal framework becomes significantly harder to defend than scraping open-source, public internet text.

As this high-stakes lawsuit moves into the discovery phase, Meta faces the reality of a multi-million-dollar trial. The decision serves as a stark reminder to tech innovators and corporate entities alike: in the race to build the most advanced digital tools, cutting corners on data acquisition will inevitably lead to a courtroom.