Meta’s ad platform has been caught running promotions for an AI porn-generation tool that explicitly encouraged users to create deepfaked videos of female US politicians — a direct violation of the company’s own advertising policies. The tool, called Kromix, marketed itself as an “AI image styler” but its ads told a different story. One video, viewed by WIRED, boasted the software had “no restrictions” and described it as “the AI that men actually use.” The ad showed a woman resembling a prominent female politician winking in front of the presidential flag before cutting to pornographic footage featuring the same face.
Kromix was hosted on Apple’s App Store until WIRED requested comment, at which point it disappeared. The app invited paying users to upload photos for use in scenarios including “bedroom rape” and “Disney love” — categories that should have triggered immediate rejection from any platform with functioning content moderation. Apple has faced its own reckoning here: San Francisco’s city attorney sent a cease-and-desist letter this summer demanding removal of 13 similar apps, accusing the company of “aiding and abetting” the sale of explicit deepfake images.
Meta’s ad library shows the Kromix campaign exclusively targeted male users under the tagline “Unleash Next-gen AI Magic.” The account ran 32 ads total, with each live between five and 46 hours before Meta removed them after WIRED asked questions. Most received 10 or fewer impressions — a detail that cuts both ways. On one hand, the reach was minimal. On the other, that a campaign so blatantly violating policy could launch at all, and stay live for nearly two days, suggests the review process is either automated to the point of uselessness or easily gamed by bad actors who know exactly what language to obscure.
This isn’t an isolated slip. Meta has repeatedly failed to keep nonconsensual intimate imagery tools off its platforms, and each incident chips away at the argument that AI safety systems are improving fast enough to outpace abuse. The fact that Kromix also slipped through Apple’s App Store review — a process famously more stringent than Meta’s ad checks — points to a systemic problem across the entire mobile ecosystem. When two of the largest tech companies in the world can’t catch an app offering “bedroom rape” scenarios, the burden falls squarely on victims to discover and report their own exploitation.
