Meta’s advertising system allowed dozens of paid ads containing AI-generated child sexual abuse material (CSAM) and sexually suggestive images of minors to run across its platforms, according to details reviewed by WIRED from researchers at the Tech Transparency Project (TTP).
The ads appeared during the last nine months and were targeted at people in the United States, United Kingdom and more than a dozen European countries. Some reached only a small number of accounts, while at least one reached 2,563 accounts in Europe.
What Researchers Found In Meta’s Ad Library
TTP researchers found more than 50 image and video ads in Meta’s ad library that contained abusive material. Meta’s ad library is intended to provide transparency about ads shown on the company’s platforms, including Facebook, Instagram, Messenger and Threads.
Some of the ads linked to so-called nudify or undressing apps. The ads included explicit AI-generated CSAM, images of minors paired with sexually suggestive statements and material that Meta later said violated its own policies.
The ads were published between November last year and the start of August. Meta’s ad transparency data showed that many ran for several days, and some targeted only men.
Most of the ads were no longer actively being served when researchers found them. Even so, they remained visible in Meta’s ad library for months, apparently without being detected by the company’s systems.
Why The Approval Process Is Under Scrutiny
The central issue is not only that abusive content appeared online. It appeared as paid advertising, meaning it passed through Meta’s ad review process before publication.
Meta says ads are reviewed before they go live, with the process relying primarily on automated tools to check them against company policies. Those policies prohibit child sexual abuse and exploitation material. They also ban ads containing nudity, sexual activity, explicit or sexually suggestive positions and sexually suggestive activities.
Katie Paul, director of the Tech Transparency Project, told WIRED that the ads did not appear to hide what they were showing or promoting. She also emphasized that this was not ordinary user-posted content, but paid advertising that Meta approved and ran while collecting ad revenue.
Meta’s ad library now shows that the ads identified by TTP were removed for violating rules covering “child sexual exploitation, abuse and nudity” and “adult nudity and sexual activity.”
Meta Removed The Ads After WIRED Asked Questions
After WIRED contacted Meta, the company removed the ads from its advertising library. A Meta spokesperson said sexual exploitation is horrific and that the company works aggressively to keep it off its platforms.
The spokesperson said the majority of the ads had minimal reach and that many had already been disabled before WIRED shared them. Meta also said many of the ads predated new AI technology it recently launched to better detect and block violating ads at upload.
Meta pointed to broader enforcement efforts as well, saying it removed over 36 million pieces of child sexual exploitation content last year and has taken legal action against nudify app developers.
But the findings suggest that the newer detection technology did not stop all of the offending ads. TTP researchers initially found around two dozen ads. Hours before publication, they found around 30 more, including multiple ads that were live and being shown to accounts at the time they were discovered.
The Problem Appears To Be Repeating
The discovery follows another recent case involving Meta’s platforms. In early July, a BBC investigation found that Instagram had run ads promoting the sale of what appeared to be child sexual abuse material in India. Those ads reportedly used terminology such as “rape video” and linked to Telegram channels where illegal content could be purchased.
In response to that reporting, Meta published a lengthy statement saying it takes a “zero tolerance” approach to child sexual exploitation.
TTP’s new findings raise a separate but connected concern: whether Meta’s systems can reliably identify and block repeat violations. Paul said some ads in the cache appeared identical to ads Meta had already removed for policy violations.
That matters because repeated paid ads are easier to detect than one-off posts if the same creative material is being uploaded again. Paul said this raises significant questions about Meta’s handling of child exploitation in paid advertising.
She also noted a reporting gap inside Meta’s ad library. According to Paul, there is no way within the ad library itself to report content that may violate Meta’s policies when the ads are no longer running.
Advertisers And Accountability
The ads appeared to be posted mostly by Meta accounts with zero or very few followers. Where paying advertisers were disclosed, Paul said they were linked to Chinese companies.
In one case, involving an ad showing a girl reclining, TTP researchers said Meta’s data identified the advertiser as a Chinese firm called Meet Social. The company was once one of Meta’s ad resellers, which handle ads in China, where Meta’s social media platforms are blocked.
According to a 2019 New York Times profile cited by WIRED, Meet Social published thousands of ads on Facebook per day and expected more than $1 billion in sales. Meet Social did not respond to WIRED’s request for comment.
The larger question is how a major advertising platform reviews paid content when AI tools can generate abusive imagery and when advertisers can repeatedly upload material that violates clear policies. The TTP findings show that transparency archives can expose failures, but they also show that harmful ads may remain discoverable long after they stop running.