Over the past nine months, Meta has allowed dozens of paid advertisements containing AI-generated child sexual abuse material and images of minors alongside sexually suggestive text, according to findings shared with WIRED. The Tech Transparency Project (TTP) uncovered more than 50 offending image and video ads in Meta's ad library, reaching potentially thousands of users across the United States, the United Kingdom, and more than a dozen European countries.
Ads with explicit child images ran on Facebook, Instagram, Messenger, and Threads
Data from Meta's ad transparency tool shows the ads were displayed on Facebook, Instagram, Messenger, and Threads. Some linked to so-called "nudify" or "undressing" apps designed to digitally remove clothing from photos. Katie Paul, TTP's director, emphasized that these were not user posts but ads that Meta reviewed, approved, and allowed to run, collecting ad revenue in the process. The ads remained in Meta's ad library for months, often undetected, until WIRED contacted the company, which then removed them.
Sponsored Protocol
Video ad morphed child's face into a sexual act
One video ad used a thumbnail of a child sitting on the floor with text saying "Realizing Deep Fantasies with Generation AI. There is so much more than what is shown, use your imagination." When clicked, it played clips of adults in sexual acts and then inserted the child's face into a sex scene. Another ad showed a young girl reclining with legs spread and the caption "I can show you more." TTP immediately reported the findings to the National Center for Missing and Exploited Children's CyberTipline, which handles reports of online child sexual abuse material.
Sponsored Protocol
Meta removed ads but AI systems missed many
After WIRED reached out, Meta deleted the ads from its advertising library. A Meta spokesperson stated that most ads had minimal reach and many were disabled before WIRED shared them, also noting new AI technology launched recently to better detect and block violating ads at upload. However, TTP researchers found around 30 additional ads just hours before publication, many still live and shown to users. This raises questions about the effectiveness of Meta's ad review systems, which are supposed to check all ads before they go live, using automated tools primarily.
Ads linked to Chinese advertisers and an app with AI porn
Most ads came from accounts with zero or few followers, and where disclosed, advertisers were linked to Chinese companies. One ad's data indicated the advertiser was Meet Social, a Chinese firm that was once one of Meta's ad resellers serving China, where Meta's platforms are blocked. Some ads linked to an app called MaskAI, available on Apple's App Store, which contained only AI-generated porn and featured face-swapping tools. Apple removed the app after WIRED contacted it. This incident follows a BBC investigation in early July that found Instagram running ads promoting child sexual abuse material in India, despite Meta's "zero tolerance" policy. The recurring nature of such ads underscores the need for stronger regulatory oversight and more transparent ad review procedures to protect minors.
Sponsored Protocol
Source: https://www.wired.com/story/meta-ran-ads-that-contained-ai-generated-child-sexual-abuse-imagery
