Meta Accused of Pirating Adult Content to Train AI Models
348It sounds like something out of a sci-fi movie, but now Meta is facing serious legal trouble over allegations of copyright theft. A company that manages adult movies, Strike 3 Holdings, along with an adult film studio called Counterlife Media, has sued Meta, claiming the tech giant illegally downloaded nearly 2,400 adult videos. All of this supposedly happened while Meta was working on artificial intelligence research.
The lawsuit was filed in a California federal court and is based on investigations that trace Meta’s online activities. The companies say they discovered, through IP and email tracking, that Meta began downloading their copyrighted films via Bittorrent as early as 2018. The lawsuit also claims that Meta not only downloaded these videos but also shared them with others, possibly even minors, by seeding the content on torrent networks. The companies say this activity went on for years, often on the very day new adult movies were released.
How Did They Find Out?
The adult companies got wind of Meta’s activities after following the digital trail. They referenced an ongoing lawsuit from earlier this year involving authors whose work was allegedly pirated by Meta. This led them to uncover that about 47 IP addresses connected to Meta, including one belonging to a Meta employee’s home, had downloaded their videos. Interestingly, the pattern of data movement suggested automated, non-human activity—more like a bot or a program rather than a person.
While Meta has long banned nudity and adult content from its main platforms, the lawsuit suggests that the company might have used the stolen videos for AI training purposes. That’s a big deal because it raises questions about what kind of data Meta is feeding into its AI systems. The adult companies believe their videos, which show natural human interactions and detailed body parts, could be particularly useful for training AI to understand human movement and facial expressions.
Why Would Meta Use Adult Content for AI?
The lawsuit hints that Meta might have used the adult videos to help improve its AI video generators, like Meta Movie Gen. These tools aim to recreate human movement and facial expressions with high realism. The companies argue that such data is unique because it contains continuous scenes without cuts, which is valuable for AI to learn human continuity and motion.
The idea is that if AI models are trained on these adult videos, they could eventually produce realistic human images or videos at little or no cost. This could have serious consequences for the market, potentially threatening the livelihoods of the content creators. The adult companies are demanding that Meta delete all the stolen content, stop torrenting their work, and pay damages of up to $150,000 per video. With nearly 2,400 videos involved, damages could total over $350 million.
While it’s unusual to think about a major tech company like Meta using adult content for AI training, the evidence shows that Meta has admitted to pirating other types of content in the past. Given this history and the detailed digital footprints uncovered, it’s possible that Meta did indeed download and use these videos. Whether or not it planned to release them, the potential use for AI training raises a lot of questions about data ethics and copyright in the age of AI.
Meta’s pursuit of the latest AI tech seems relentless, and this lawsuit points to a darker side of innovation—using stolen content to push forward. As AI continues to evolve, legal battles like this highlight the importance of respecting copyright and understanding what data powers the next generation of technology.
What do you think?
It is nice to know your opinion. Leave a comment.