A controversial AI image styling tool called Kromix recently promoted its services through video advertisements across social media. One of the video ads, reviewed during the campaign, featured a woman closely resembling a prominent American politician standing in front of U.S. presidential and national flags. The caption asked what would happen if she moved, followed by a wink and a transition into a pornographic video. Promotional voice-overs for the tool marketed it as having no restrictions and being an AI specifically used by men.
App Store Presence and Prohibited Content
Before facing scrutiny and subsequent removal, the Kromix app was hosted on Apple's App Store. The platform invited paying users to upload personal photographs and insert them into scenarios described as bedroom rape and Disney love, alongside AI-generated pornographic depictions of individuals wearing Spider-Man costumes. Contact inquiries sent to the developers went unanswered. Apple maintains strict policies prohibiting non-consensual intimate imagery and pornography, yet enforcement remains challenging. Earlier in the summer, the city attorney of San Francisco issued a cease-and-desist letter demanding the removal of thirteen similar applications accused of facilitating explicit deepfakes.
Targeted Meta Ad Campaigns
According to Meta’s ad library data, the promotional campaign was exclusively targeted toward male users under the tagline Unleash Next-gen AI Magic. The account linked to the ads ran thirty-two individual items that remained active between five and forty-six hours, with most garnering ten or fewer impressions. Katie Paul, director of the Tech Transparency Project, noted that these campaigns consistently sexualize and place women into AI-generated pornography, raising questions about platform oversight when political figures are targeted by active advertisements.
Broader Context of Political Deepfakes
Elected officials, particularly women, have faced non-consensual deepfake media for years. Earlier in the year, tools like Grok generated millions of sexualized images targeting officials in Sweden, England, and Northern Ireland using official portraits. Benjamin Shultz, lead researcher at the American Sunlight Project, stated that as the production of AI content democratizes, benign public media assets routinely become raw material for low-barrier exploitation tools.
Platform Ad Review Policies and Evasion Tactics
Meta's public policies state that advertisements undergo automated and manual review before publication to check for prohibited sexual content, nudity, or exploitation services. However, advertisers frequently bypass these safeguards. The page running the Kromix ads was created on August 3 with zero followers, a potential red flag that went unflagged. Company representatives emphasized that bad actors continuously evolve their tactics, often disguising pornographic content behind benign cover images of beaches, mountains, or waterfalls to evade detection.
Company Responses and Enforcement Actions
Cindy Southworth, head of women’s safety at Meta, stated that the company works aggressively to combat sexual exploitation and removes hundreds of thousands of violating ads while pursuing legal action against developers. Similarly, an Apple spokesperson confirmed that Kromix was removed from the App Store for introducing prohibited features after its initial review, reiterating that the company strictly bans pornography and nudification applications.
