Meta is facing a new lawsuit that accuses the company of downloading and distributing thousands of copyrighted adult films through BitTorrent networks. According to the complaint, this activity was part of an effort to accelerate the development of its artificial intelligence systems. The suit was filed by adult content producer Strike 3 Holdings and alleges that Meta used corporate IP addresses to obtain and share pirated videos as a strategy for improving data acquisition for AI training. The lawsuit claims these practices began as early as 2018 and continued despite clear evidence linking Meta to the infringement.
Filed in a California federal court, the lawsuit places these claims within a broader pattern of behavior by Meta. Earlier this year, a group of authors accused the company of downloading more than 81 terabytes of content from shadow libraries to train its AI models. Meta denied uploading or distributing that material through BitTorrent. However, Strike 3 now says it has tracked the distribution activity using proprietary tools that detect repeat offenders. The lawsuit states that several of the identified IP addresses were registered directly to Meta or were linked to Meta employees.
Strike 3 argues that this method of using BitTorrent to acquire high-quality content gave Meta a competitive advantage. The BitTorrent protocol rewards users who upload popular files. The complaint alleges that Meta deliberately shared its most in-demand titles in order to gain faster access to other large volumes of content. According to Strike 3, this approach not only violated copyright law but also contributed to widespread access to explicit content without age restrictions. The company believes these actions damaged both its business and its reputation.
Strike 3 Alleges Strategic Seeding of Adult Content
According to the lawsuit, Meta’s file-sharing activity was not random. Strike 3 Holdings claims it was a calculated strategy focused on widely viewed adult videos. Proprietary tracking tools reportedly detected patterns of distribution involving at least 2,396 titles. The company says Meta shared these files for extended periods after downloading them, often over the course of several days, weeks, or even months. Strike 3 believes this tactic was intended to maximize the efficiency of downloading other data by taking advantage of BitTorrent’s reciprocal sharing model.
The lawsuit also claims Meta acted quickly to seed content that had just been released on adult platforms. This timing, according to Strike 3, suggests a deliberate attempt to participate in active torrent swarms from the moment new content became available. The complaint argues that this strategy helped Meta reduce server costs while still gathering the large datasets necessary to train its AI systems at scale.
In addition to the copyright issues, the lawsuit raises concerns about the accessibility of adult content to underage users. Since BitTorrent has no built-in system for age verification, Strike 3 believes Meta’s activity may have made explicit material available to minors. This would represent a violation of content distribution laws in several states and could also compromise the company’s legal standing as a provider of age-restricted media.
Evidence Points to Meta Infrastructure, According to Filing
To support its claims, Strike 3 Holdings cites data collected using its VXN Scan and Cross Reference tools, which are designed to monitor digital piracy. According to the complaint, the company identified 47 IP addresses linked to Meta’s infrastructure. Some of these addresses were reportedly tied to Meta’s main network, while others used Virtual Private Clouds to hide their origins. The lawsuit describes this system as a stealth network created to distribute content discreetly.
Even after Meta was notified of the alleged activity, the lawsuit says the seeding continued. Further analysis reportedly showed that Meta’s torrent activity extended across multiple addresses and involved additional content types beyond adult videos. These included e-books, television shows, and software. The complaint suggests that this behavior aligns with automated content gathering, not individual use, and points to a large-scale system intended for AI training.
In a statement responding to the lawsuit, a Meta spokesperson said the company is reviewing the complaint but does not believe the claims are accurate. The court has not yet ruled on whether the evidence is admissible. Meta has not confirmed or denied if any of the material mentioned in the suit was used to train its AI models. If the case moves forward, it could shape how courts evaluate data sourcing practices in the context of AI development and copyright law.
Legal and Industry Implications for AI Training Practices
This lawsuit arrives during a broader conversation about how companies collect data to train generative AI models. Meta, which has heavily invested in AI research, is already involved in other legal cases related to the unauthorized use of books, images, and software code. If a court determines that Meta distributed copyrighted adult content for the purpose of AI training, it could establish a new legal standard for handling similar claims.
Strike 3 is asking the court for monetary damages, a permanent injunction to stop any further use of its content, and the deletion of all material that may have been included in Meta’s AI datasets. The company also argues that its content, which features high-end production and unique human expression, is particularly valuable for visual AI training. If Meta used that content, the complaint alleges, the company may now be able to recreate similar material without ever compensating the original creators.
At the heart of this lawsuit is a larger issue of fairness in AI development. Strike 3 contends that companies like Meta gain an unfair advantage by ignoring copyright law, while smaller creators are forced to comply with costly legal standards. The outcome of this case could influence not only how AI companies gather training data, but also the legal benchmarks required to prove infringement in complex digital environments.