Meta’s new AI ad screening was supposed to stop exactly this. Instead, researchers at the nonprofit Tech Transparency Project (TTP) counted more than 250 additional ads containing child sexual abuse material running across Facebook, Instagram, Messenger and Threads since the start of August 2026 — some of them identical to ads Meta had already taken down.
The trail goes back further. In early August 2026, after reporting by WIRED, Meta deleted around 50 ads that directly contained CSAM; TTP had initially flagged 53. Meta said most of the abusive ads predated new AI tools designed to “better detect and block” harmful ads at upload. In total, TTP has now documented more than 350 abusive video ads across Meta’s platforms and its wider advertising network since the end of 2025.
Real children, not only synthetic composites. That is what separates the newer batch. TTP traced the real-world identities of four minors, three of them in the US: two teenagers with public accounts on Meta’s own platforms, and a stock-photo model listed as pre-teen on stock photo sites. The fourth was a member of a European royal family, whose officially released photograph was turned into a video of a graphic sex act. The researchers are not naming the child.
The ads follow one template: an innocuous picture of a preteen or teenage girl, overlaid text saying it is “not just a photo” and that “there are no restriction” [sic] on what can be done with it, then a seconds-long clip that morphs the face into explicit content. A tap sends the viewer to an app store listing.
- The ads pointed people towards almost 50 apps combined across Google’s Play Store and Apple’s App Store
- Around 300 of the 350 ads directed people to AI image or video generator ads, most of those apps linked to China-based developers
- Reach recorded in Meta’s ad library: more than 29,000 accounts across the European Union and around 7,000 in the United Kingdom, with more than 250 of the ads appearing in the United States, 120 in Australia and 80 in India
“We don’t tolerate nudify apps or any kind of child exploitation, whether real or AI-generated,” says Meta spokesperson Tracy Clayton. The company says many of the ads had already been removed, that most drew fewer than 200 impressions, and that it received “under $5,000” in ad payments. TTP counters that in some cases Meta took a week to review live ads reported through its own tools.
Apple says 17 of the 20 apps named in the research were already off the App Store, that it has acted on the remaining three, and that it removed seven further apps WIRED identified. Google says it removed every app in the report and restricted store search terms such as “nudify”.
The fallout is regulatory. Senator Mark Warner wrote to Mark Zuckerberg and has had no reply; Michigan’s attorney general is “actively looking into the matter”; Florida attorney general James Uthmeier says his office is “investigating”; Australia’s eSafety regulator is assessing Meta’s compliance with its rules. All of it lands while Meta is paying up to $16.7 billion to settle lawsuits over its social networks allegedly harming children, and appealing a New Mexico court decision that would force it to improve CSAM reporting and add human review in the state.
