The Tech Transparency Project found Meta ran more than 300 paid ads containing apparent AI-generated child sexual abuse content for about nine months, renewing questions about the company`s ad review systems.
The Tech Transparency Project (TTP), an independent nonprofit research group, said on Tuesday that Meta approved and ran more than 300 paid advertisements containing suspected AI-generated child sexual abuse material across its platforms over a roughly nine-month period. The ads ran on Facebook, Instagram, Messenger and Threads between November 2025 and August 2026.
According to the report, which was shared with WIRED and covered by Bloomberg, the ads reached more than 29,000 accounts and were targeted at users in the United States, the United Kingdom and more than a dozen European countries. At least one advertisement reached 2,563 accounts in Europe. Researchers said the actual reach may be higher because Meta's ad library does not publish complete performance data for all countries, including the U.S.
What the ads contained
The advertisements openly depicted or promoted AI-generated images of children being sexually abused, or paired images of minors with sexually suggestive text. Many of them pushed so-called nudify or undressing applications — tools that use artificial intelligence to produce sexual images from ordinary uploaded photos or to swap faces into pornographic content. Some ads placed the faces of real children into sexual videos.
TTP Director Katie Paul said the ads made no effort to hide what they were promoting. They were paid placements that would have passed through Meta's pre-publication review process, rather than organic posts from third parties.
A moderation pipeline that failed
The findings raise questions about Meta's advertising controls. The company's own standards prohibit content that sexually exploits or endangers children, and Meta says all ads are reviewed before publication, primarily through automated tools. Despite those safeguards, hundreds of violating ads cleared the system.
TTP also documented repeat violations: identical ads that Meta had previously removed for breaching its policies were reposted and allowed to run again. In addition, roughly 30 more violating ads were published after WIRED first contacted Meta about the material, and some remained live when researchers checked shortly before the report's release. The ad library offers no mechanism for users to report a violating ad once it has stopped running.
Pattern of related incidents
The finding is not isolated. In July 2026, the BBC reported that Instagram ran paid ads in India offering what appeared to be child sexual abuse material, using search terms such as “rape video” and linking to Telegram channels. Also in July, Bloomberg reported that Facebook and Instagram had served thousands of ads for AI nudify apps through GatherOne, a Beijing-based advertising partner; TTP attributed roughly 7,600 ads for face-swap and undressing applications to that company.
In April 2026, the Consumer Federation of America filed a lawsuit against Meta over its handling of scam advertising.
Meta's response
Meta removed the offending ads after WIRED flagged them. A company spokesperson said sexual exploitation is horrific and that the company works aggressively to keep it off its platforms, adding that the majority of the ads had minimal reach and many were disabled before WIRED shared them. The spokesperson also said many of the ads predated new AI tools Meta launched to detect and block violating ads at upload, and that the systems are being improved. Meta said it removed more than 36 million pieces of child sexual exploitation content last year and has taken legal action against developers of nudify applications.
Critics, including TTP's Katie Paul, have noted that a number of the ads were posted after Meta introduced its new detection systems, and that some went live after journalists had already contacted the company. That timeline, they argue, raises doubts about how effectively Meta's automated review enforces its own child-safety policies.
TTP reported the findings to the National Center for Missing & Exploited Children (NCMEC) CyberTipline.