Meta Pulled 32 Kromix Ads Promoting AI Porn of Female Politicians
Meta removed 32 Kromix ads promoting AI porn with female politicians, while Apple pulled the app amid scrutiny of automated ad and App Store review failures.
Summary
Meta ran 32 male-targeted Kromix ads on Facebook and Instagram under “Unleash Next-gen AI Magic,” despite bans on sexual content, nonconsensual intimate imagery, nudify apps and AI kissing apps. Kromix advertised unrestricted styling of real people. One video changed a woman resembling a prominent female US politician, pictured with US and presidential flags, into a pornographic performer; beach, mountain and waterfall covers concealed others. Created August 3 with no followers, the page ran ads for five to 46 hours, mostly receiving 10 or fewer impressions, until Meta removed them after WIRED’s inquiry.
Kromix let paid users upload photos for “bedroom rape” and “Disney love” scenarios and showed AI pornography featuring women in Spider-Man costumes. Apple removed it, saying prohibited features appeared after initial review; its contact did not respond. Earlier this summer, San Francisco’s city attorney demanded removal of 13 apps and an end to aiding explicit deepfake sales. In January, Grok produced about 3 million sexualized images of women and girls; officials in Sweden, England and Northern Ireland were targeted using official images.
Meta says automation prescreens ads. Between November 2025 and January 2026, it removed 344,000 nudify ads, sued at least one promoter, launched detection AI and shared offender data with other tech companies. Yet Tech Transparency Project researchers found more than 50 Meta ads from November 2025 to August 2026 containing AI child sexual abuse material or sexualized minors, including child face swaps into sexual acts. Some reached thousands across the US, UK and at least 12 European countries, with some appearing after the upgrade. Meta removed them after WIRED’s inquiry, while Apple pulled linked face swap app MaskAI. Katie Paul blamed AI review and fewer human moderators. Cindy Southworth cited evolving evasion and promised an investigation. Benjamin Shultz said official media now feeds low-code deepfake tools.
Positives
- 32 Kromix ads were removed from Meta’s platforms after WIRED requested comment.
- 344,000 nudify ads were removed from Facebook and Instagram between November 2025 and January 2026.
- Apple removed Kromix and MaskAI from the App Store after WIRED raised questions about their prohibited features.
- Meta launched new detection AI, sued at least one nudify advertiser and shares removed developers’ information with other technology companies.
- San Francisco’s city attorney demanded that Apple remove 13 apps and stop aiding explicit deepfake sales.
Risks & concerns
- 32 Kromix ads bypassed Meta’s automated review despite explicit policies against sexual content and nonconsensual imagery.
- Kromix offered paid photo uploads for scenarios labeled “bedroom rape” and “Disney love.”
- More than 50 Meta ads contained AI child sexual abuse material or sexualized minors, sometimes reaching thousands of users internationally.
- About 3 million sexualized images of women and girls were generated through Grok in January, including attacks on European officials.
- Some violating ads appeared after Meta introduced AI designed to detect and block them during upload.
- Benign cover images helped pornographic advertisements evade Meta’s screening systems.