Meta platforms recently ran advertisements for an AI pornography-generation tool that appeared to encourage users to create deepfaked videos resembling female US politicians, according to a report published by WIRED. The campaign represents the latest instance in which Meta has failed to prevent promotions for tools that produce nonconsensual intimate imagery from appearing on its platforms, despite company policies explicitly banning ads containing sexual material. The tool, called Kromix, marketed itself as an "AI image styler" with "no restrictions" and was described in one advertisement as "the AI that men actually use."

The advertising campaign, which was discovered by the Tech Transparency Project, ran 32 ads over periods ranging from five to 46 hours, with most receiving 10 or fewer impressions before removal. Meta's ad library shows the ads were exclusively targeted to male users with the tagline "Unleash Next-gen AI Magic." One video advertisement featured a woman closely resembling a prominent female US politician standing in front of a US flag and the presidential flag above the caption "What if she moved?" before cutting to a scene showing a woman with the same face performing in a pornographic video. The Kromix app, which was hosted by Apple's App Store before WIRED requested comment, invited paid users to upload photos for use in scenarios including "bedroom rape" and "Disney love." The page running the advertisements was created on August 3, shortly before the ads began running, and had no followers.

Katie Paul, director of the Tech Transparency Project, notes that "all of the ads that we've seen so far broadly on the nudify issue are exclusively sexualizing and nudifying and putting women into AI-generated porn." According to the report, Meta's own count shows it removed 344,000 ads on Facebook and Instagram that were promoting nudify services between November last year and January. Meta states in its public documentation that each ad is "reviewed" against its policies before publication, with the company's advertising standards noting that "our ad review system relies primarily on automated tools to check ads and business assets against our policies." Most of the video ads featured cover photos of benign images like beaches, mountains, or waterfalls, disguising the pornographic content within.

The report explains that politicians and elected officials, overwhelmingly women, have been targets of nonconsensual deepfake images and videos for years. Benjamin Shultz, the lead researcher at the American Sunlight Project who is investigating how harmful deepfake technology is used to target politicians, observes that "as the production and distribution of AI-generated content have democratized, formerly benign media—official portraits, video from a press conference—have become raw material for low- and no-code tools that anyone with an internet connection can use." In January, as Grok was used to create around 3 million sexualized images of women and girls, officials in Sweden, England, and Northern Ireland were targeted by people using the technology, often using official images as a starting point. Cindy Southworth, head of women's safety at Meta, describes sexual exploitation including promoting nudify apps as "horrific" and says the company "work[s] aggressively to fight it," pointing out that "this is a highly adversarial space where bad actors constantly evolve their tactics."

The report notes that some violating ads were published after Meta told WIRED it had introduced new AI technology to "better detect and block violating ads at upload." An Apple spokesperson told WIRED that the company has removed Kromix—which appeared to have added prohibited content and features after initial review—from the App Store, stating that "'nudification' apps are against our App Review Guidelines and we aggressively work to identify and remove apps—as well as the developer accounts behind them—that attempt to evade our processes and policies." Paul says that Meta has increasingly relied on AI to flag and vet advertisements while laying off human moderators, adding that "ads are Meta's primary product, and there appears to be zero quality control." The continued appearance of these ads raises questions about whether automated enforcement can keep pace with actors determined to circumvent detection systems. The tension between platform scale and content safety suggests that even substantial investment in detection technology may struggle without corresponding human oversight capacity.