Meta platforms recently ran advertisements for an AI-powered porn-generation tool that encouraged users to create deepfaked videos of female US politicians. Despite Meta’s explicit policies against sexual material, the tool, named Kromix, was promoted to male users through ads that bypassed automated safeguards.
A Targeted Campaign
According to a WIRED investigation, one video advertisement featured a woman closely resembling a prominent female US politician standing before the American flag. The ad transitioned into a pornographic scene featuring the same face. Kromix marketed itself as having "no restrictions," explicitly stating that "all the characters are real people" and inviting paid users to upload photos for nonconsensual scenarios.
Katie Paul, director of the Tech Transparency Project, noted that these ads exclusively target women for AI-generated pornographic exploitation. "This raises a lot of questions about how Meta is going to deal with this kind of content... when the company is actively running ads containing this content on their platform," Paul stated.
Evasive Tactics and Algorithmic Failures
Meta’s ad review system relies primarily on automated tools, which failed to flag the Kromix campaign. Advertisers used "adversarial tactics" to evade detection, such as using benign cover photos of landscapes or waterfalls to mask the pornographic nature of the underlying video. While Meta claims to have removed 344,000 similar ads between November and January, the Kromix ads ran for up to 46 hours before being taken down.
Apple also faced scrutiny, as the Kromix app was hosted on the App Store until WIRED requested comment. An Apple spokesperson confirmed the app was removed for violating guidelines against pornography and "nudification." This follows a cease-and-desist letter from San Francisco’s city attorney earlier this summer, which accused Apple of "aiding and abetting" the distribution of explicit deepfakes through similar applications.
