Meta Platforms has formally requested a U.S. federal court to dismiss a lawsuit accusing the tech giant of pirating and utilizing thousands of pornographic videos to train its artificial intelligence systems—a claim Meta categorically denies as baseless and illogical.
In a motion filed earlier this week with the U.S. District Court for the Northern District of California, Meta argued that the plaintiffs presented no credible evidence linking the company’s AI models to any copyrighted adult material. The filing described the lawsuit’s assertions as “nonsensical” and “unsupported by facts,” firmly rejecting the notion that AI training involved any infringing content.
The lawsuit in question alleges that Meta illegally downloaded and redistributed copyrighted pornographic content to enhance the capabilities of its AI tools, such as large language models and image generators. However, Meta countered in its legal filing that any adult content obtained was for internal research purposes and not used in any capacity to train AI systems.
“Plaintiffs have fabricated a speculative narrative built on assumptions, not proof,” the company stated in its court filing. Meta emphasized that the complaint relies heavily on conjecture, lacking any direct evidence that its AI models were exposed to or learned from the pornographic material in question.
Meta also noted that while the company may have accessed certain adult videos, these were not redistributed or incorporated into any training datasets used by AI development teams. The company clarified that the content in question was viewed solely for personal use by individual employees, not as part of any structured AI training protocol.
The motion to dismiss comes amid increasing scrutiny over how major tech firms source data to train powerful AI models. Content creators, publishers, and rights holders have raised concerns about unauthorized use of copyrighted material, prompting a wave of legal challenges targeting companies like Meta, OpenAI, and Google.
Despite the broader industry trend, Meta maintains that its AI systems are developed within strict ethical and legal frameworks. The company reiterated that its internal policies prohibit the use of copyrighted content—particularly adult material—without proper licenses or permissions for any training-related activities.
Legal experts note that the burden of proof lies on the plaintiffs to demonstrate that copyrighted videos were indeed used in training datasets and that this use resulted in direct infringement. So far, no such evidence has been made public, which may weaken the plaintiffs’ case.
In addition to denying the infringement claims, Meta also criticized the legal team behind the lawsuit for attempting to conflate AI development with unrelated activities. According to the tech firm, the plaintiffs’ case stretches the definition of AI training beyond recognition, making it difficult to establish a coherent legal argument.
The case reflects a growing tension between AI innovation and intellectual property laws. As generative AI models become more advanced and data-hungry, tech giants find themselves in legal gray areas where the boundaries of fair use, licensing, and data ethics are hotly contested.
At the heart of the issue is the question of whether AI developers can legally use publicly accessible or privately obtained content—such as images, text, or videos—as training material without explicit permission from rights holders. Courts have yet to establish clear precedents, leaving both creators and tech companies in a state of legal uncertainty.
This lawsuit also underscores the reputational risks faced by companies involved in AI development. Allegations involving adult content, even if unproven, can tarnish a brand’s image and fuel public skepticism about how transparently and responsibly AI systems are being trained.
In response to growing concerns, Meta has previously stated that it is committed to transparency around its AI development practices. The company claims it actively works to minimize legal and ethical risks by sourcing training data from licensed, open-source, or public domain materials whenever possible.
Furthermore, the case calls attention to the need for clearer regulations governing AI training data. As lawsuits like this become more frequent, industry watchers suggest that lawmakers may need to establish new legal frameworks to balance innovation with content ownership rights.
While the court has yet to rule on Meta’s motion to dismiss, the case is likely to serve as a bellwether for future legal battles over AI and copyright. Observers say the outcome could set important precedents for how tech companies handle sensitive content in the age of machine learning.
Until then, the legal landscape remains uncertain, and companies like Meta will continue to walk a fine line between technological advancement and legal accountability.