Meta Claims Downloaded Porn at Center of AI Lawsuit Was for ‘Personal Use’
Meta, the global technology behemoth, has formally requested a US district court to dismiss a contentious lawsuit alleging that the company illegally torrented a vast collection of adult films to train its advanced artificial intelligence models. This legal maneuver comes amid growing scrutiny over AI training data and copyright infringement, placing Meta at the center of a particularly salacious and high-stakes legal battle. The lawsuit, initiated by adult film producer Strike 3 Holdings, challenges Meta not only on potential financial penalties but also on its reputation and ethical commitments in the rapidly evolving landscape of AI development.
Strike 3 Holdings launched the legal action after an internal investigation purportedly uncovered unauthorized downloads of its copyrighted adult films. These downloads were allegedly traced to Meta’s corporate IP addresses. More strikingly, Strike 3 claims to have unearthed a sophisticated "stealth network" comprising approximately 2,500 "hidden IP addresses," which they allege Meta employed to further conceal illicit downloading activities. The core accusation from Strike 3 is grave: Meta, they contend, engaged in a clandestine operation to amass a massive dataset of adult content. The purpose, according to Strike 3, was to train an unannounced, adult-oriented version of its AI model, potentially an explicit variant of its "Movie Gen" technology. The financial stakes are substantial, with Strike 3 seeking damages that could potentially surpass $350 million, as initially reported by TorrentFreak, underscoring the severity of the alleged copyright infringement and the perceived intent behind Meta’s actions.
In a robust counter-response, Meta filed a comprehensive motion to dismiss the lawsuit on Monday, categorically refuting Strike 3’s claims as resting solely on "guesswork and innuendo." Meta vehemently denies any corporate involvement in, or knowledge of, the alleged illegal torrenting. Furthermore, Meta launched a direct counter-offensive against Strike 3 itself, leveling the contentious accusation that the adult film producer is a "copyright troll." This term is often used to describe entities that primarily engage in aggressive litigation over copyright infringements, sometimes perceived as aiming to extract settlements rather than genuinely protecting intellectual property rights. Meta’s filing specifically stated that Strike 3 "has been labeled by some as a ‘copyright troll’ that files extortive lawsuits," a significant rhetorical move designed to undermine the plaintiff’s credibility and standing in court.
The tech giant explicitly requested the court to dismiss all copyright infringement claims, asserting a profound lack of concrete evidence linking Meta as a corporate entity to the alleged downloads of approximately 2,400 adult movies owned by Strike 3. Meta contended that there was no indication it directed, sanctioned, or was even aware of any illegal activity occurring on its network involving Strike 3’s content. Meta’s legal filing further emphasized that Strike 3 had provided "no facts to suggest that Meta has ever trained an AI model on adult images or video, much less intentionally so," directly challenging the central premise of the lawsuit. A Meta spokesperson, in a succinct statement to Ars Technica, dismissed the allegations outright, declaring, "These claims are bogus."
A cornerstone of Meta’s defense hinges critically on the timing and scale of the alleged downloads. The company highlighted that the unauthorized torrenting reportedly commenced in 2018 and spanned a period of seven years. Crucially, Meta points out that this timeline significantly predates its publicly announced and concentrated efforts in "researching Multimodal Models and Generative Video" by approximately four years, with those initiatives beginning around 2022. This substantial temporal discrepancy, Meta argues, renders the notion that these downloads were intended for AI training "implausible" from the outset. The company contends that it would be illogical to collect data for a purpose that had not yet become a strategic focus for its AI development.
Adding another compelling layer to its defense, Meta cited its internal corporate policies, which explicitly "prohibit generating adult content" through its AI tools. This policy, Meta contends, directly "contradict[s] the premise that such materials might even be useful for Meta’s AI training," as it would be fundamentally antithetical to their operational guidelines and public commitment to responsible AI development. The company implies that intentionally training an AI on explicit material would violate its own core principles and user safety guidelines, making the accusation inherently unbelievable.
Instead, Meta asserts that the available evidence "is plainly indicative" that the flagged adult content was torrented for "private personal use" by individuals. The company elaborates on this, pointing to the relatively "small amount" of content linked to Meta IP addresses and employees—described as "a few dozen titles per year intermittently obtained one file at a time." This sporadic and limited pattern, Meta argues in its filing, strongly suggests that "disparate individuals downloaded adult videos for personal use" rather than a coordinated, large-scale corporate effort to build a training dataset.
To further underscore this point, Meta drew a stark comparison with other high-profile AI copyright lawsuits, such as those brought by authors whose works form part of "enormous datasets" involving millions of copyrighted texts used for AI training. In contrast, the activity on Meta’s corporate IP addresses allegedly amounted to only about 22 downloads per year from Strike 3’s catalog. Meta claims this figure is "nowhere near the concerted effort to collect the massive datasets Plaintiffs allege are necessary for effective AI training." The argument here is that the sheer quantity and inconsistent nature of the downloads are fundamentally inconsistent with the systematic, high-volume data acquisition strategies typically employed for robust and effective AI model development. Such a small, uncoordinated collection would be largely useless for training a sophisticated AI.
Meta further complicated Strike 3’s claims by questioning the direct link between the alleged activity and any specific Meta employee or corporate directive. The company’s legal team emphasized that Strike 3 "does not identify any of the individuals who supposedly used these Meta IP addresses, allege that any were employed by Meta or had any role in AI training at Meta, or specify whether (and which) content allegedly downloaded was used to train any particular Meta model." This profound lack of specific attribution forms a critical part of Meta’s argument that Strike 3’s case is built on mere conjecture rather than concrete, actionable evidence. Without identifying the individuals, their roles, or the specific use of the content, Meta argues, the claims remain unsubstantiated.
The sheer scale of Meta’s operations also plays a pivotal role in its defense. Meta argued that "tens of thousands of employees," alongside "innumerable contractors, visitors, and third parties access the internet at Meta every day." Given such extensive and diverse network traffic, Meta contended that while "it’s possible one or more Meta employees" downloaded Strike 3’s content over the past seven years, "it is just as possible" that a "guest, or freeloader," or "contractor, or vendor, or repair person—or any combination of such persons—was responsible for that activity." This broad range of potential users on its network makes it exceedingly difficult, Meta claims, to definitively assign corporate responsibility for individual, unauthorized actions without specific proof. The company cannot be held liable for every single byte of data traversing its network by every person who might connect to it.
The lawsuit also included a specific claim involving a Meta contractor allegedly directed to download adult content from his father’s house. Meta dismissed this, too, as "plainly indicative of personal consumption." The company pointed out that this contractor worked as an "automation engineer," a role with "no apparent basis provided for why he would be expected to source AI training data." Meta reiterated that "no facts plausibly" tie "Meta to those downloads," and the fact that the torrenting allegedly ceased when his contract ended "says nothing about whether the alleged torrenting was performed with Meta’s knowledge or at its direction." The implication is clear: even if a contractor was involved in such downloads, it was an individual act for personal gratification, not a corporate mandate for AI data collection.
Perhaps one of the most puzzling aspects of Strike 3’s complaint for Meta, as outlined in its motion, is the allegation concerning a "stealth network" of hidden IP addresses. Meta framed this as "yet another conundrum" that Strike 3 "fails to address." Meta’s argument challenges the fundamental logical consistency of Strike 3’s claims: "why would Meta seek to ‘conceal’ certain alleged downloads of Plaintiffs’ and third-party content, but use easily traceable Meta corporate IP addresses for many hundreds of others?" If the intent was truly to secretly train AI, a consistent and robust concealment strategy would be expected, not a mix of hidden and easily identifiable activity.
Meta’s filing concludes that "The obvious answer is that it would not do so," unequivocally dismissing Strike 3’s "entire AI training theory" as "nonsensical and unsupported." This line of defense aims to highlight what Meta perceives as fundamental flaws and inconsistencies in Strike 3’s narrative, suggesting a lack of a coherent and believable strategy if Meta were indeed engaged in such illicit activities. The alleged inconsistency in using both hidden and traceable IPs, according to Meta, severely undermines the credibility of the "stealth network" claim and, by extension, the entire premise of the lawsuit.
Finally, Meta addressed the implicit suggestion within the lawsuit that it should have been more vigilant in "policing" its vast network for illegal activity. Meta argued that "Monitoring every file downloaded by any person using Meta’s global network would be an extraordinarily complex and invasive undertaking." Citing legal precedent, Meta maintained that it is only required to employ a "simple measure" to monitor such activity, not an exhaustive and intrusive surveillance system. This argument touches upon the practical limitations, immense costs, and significant privacy concerns associated with implementing extensive network monitoring, especially for a company of Meta’s size and global reach, which serves billions of users. The legal framework, Meta implies, does not demand an impossible standard of policing.
The tech giant is now hoping the court will concur that Strike 3 has demonstrably failed to provide sufficient evidence linking Meta directly to the alleged illegal downloads or their purported use in AI training. Strike 3 has been given two weeks to submit its response to Meta’s motion, as reported by TorrentFreak, setting the stage for the next phase of this intricate legal skirmish.
For Meta, the stakes in defeating this lawsuit extend far beyond merely avoiding substantial financial damages, which could run into hundreds of millions of dollars. The case also represents a crucial defense of its commitment to ethical AI development and its stated policy of ensuring that its AI video tools do not generate explicit content—a domain that is increasingly subject to stringent regulation globally. In its filing, Meta emphatically asserted that Strike 3 provided no evidence that Meta trained its AI models on Strike 3’s content "because there was none." This categorical denial aims to preserve Meta’s reputation as a responsible AI developer.
A Meta spokesperson reiterated the company’s firm stance, stating, "We don’t want this type of content, and we take deliberate steps to avoid training on this kind of material." This statement underscores Meta’s efforts to align its AI development practices with societal norms, user safety, and regulatory expectations, making the allegations of using pirated pornography for AI training particularly damaging to its public image and future AI ambitions. The outcome of this case could therefore have significant implications for how AI companies approach data acquisition, content moderation, and corporate responsibility in the face of widespread internet usage and the complexities of monitoring vast networks.
This legal battle highlights the complex challenges emerging at the intersection of intellectual property law, corporate responsibility, and the rapid advancement of artificial intelligence. As Meta seeks to clear its name and dismiss what it calls "bogus" claims, the court’s decision will be closely watched by tech companies, copyright holders, and policymakers alike, potentially setting new precedents for how copyrighted material can—or cannot—be used in the training of AI models. The focus remains squarely on whether Strike 3 can move beyond "guesswork and innuendo" to provide concrete evidence that Meta, as a corporate entity, deliberately engaged in or directed the alleged copyright infringement for its AI initiatives, or if these were indeed isolated acts of "personal use" by unidentified individuals on a sprawling corporate network.
