Meta Approved and Ran Paid Ads Containing AI-Generated Child Sexual Abuse Material for Nine Months
Summary
Updated Sep 8 — Follow-up TTP investigation found 332 Meta CSAM ads — including images of real, identifiable children — far exceeding the 50+ initially reported.
- • Meta reviewed, approved, and ran 50+ paid ads containing AI-generated CSAM across Facebook, Instagram, Messenger, and Threads
- • Tech Transparency Project found ads targeting U.S., UK, and over a dozen European countries over a nine-month period
- • At least one ad reached 2,563 accounts in Europe; ads linked directly to nudify and undressing apps
- • Meta only removed the ads after WIRED inquired — this is the second such discovery on Meta platforms in recent weeks
Updates
Follow-up TTP investigation found 332 total CSAM-containing ads on Meta — far exceeding the 50+ initially reportedrevised
A subsequent TTP investigation published September 2026 found 332 ads containing CSAM on Meta, dwarfing the 50+ figure from the initial August WIRED report; the original inquiry appears to have captured only a fraction of the violations
Real children identified in ads: a European royal family member, a 14-year-old Instagram influencer, and a preteen in stock photos
TTP matched multiple CSAM ads to photos of real, identifiable children: a young European royal's press photo was animated into a graphic sex act video; a 14-year-old influencer's sports uniform photo was animated into an oral sex video; a preteen in stock photos was depicted being molested by an adult male
Vast majority of Meta CSAM ads promoted Chinese-made AI nudify and undressing apps targeting men
TTP found that AI apps made in China — particularly 'nudify' apps that digitally alter images of children — were the primary beneficiaries of Meta's ad review failures; some ads used captions like 'This is the AI men actually use' and promised 'no restrictions'
DOJ has clarified AI-generated CSAM carries the same federal criminal liability as real CSAM under child pornography law
The Justice Department's position means creating, distributing, or facilitating AI-generated material sexualizing minors — including digitally altered images of real children — carries the same federal criminal exposure as traditional CSAM, heightening legal risk for advertisers and potentially Meta itself
Details
50+ paid Meta ads containing AI-generated CSAM ran for nine months across four platforms
Ads appeared on Facebook, Instagram, Messenger, and/or Threads between November 2025 and early August 2026; discovered by Tech Transparency Project researchers investigating nudify app advertising
Meta's ad review process approved the CSAM-containing ads before they ran
TTP Director Katie Paul: 'These are ads that were reviewed, approved, and allowed to run by Meta, never encountering interference while the company collected the ad dollars' — distinguishing these from user-posted content
At least one ad reached 2,563 accounts across eight European countries
Countries included France, Germany, Ireland, Italy, Netherlands, Spain, Sweden, and UK; Meta's ad library does not include reach data for the US, so total reach is likely higher
Ads removed only after WIRED inquiry; had been in Meta's public ad library undetected for months
Meta claimed the majority had minimal reach and many were disabled before WIRED contact; however, some ads were available in Meta's transparency library for months without detection or removal
TTP filed reports with NCMEC CyberTipline; Meta faces potential CSAM law liability
NCMEC does not comment on reports received; Meta's direct monetization of CSAM-containing ads exposes it to criminal referral risk under federal CSAM statutes and EU DSA enforcement
This is the second CSAM ad discovery on Meta platforms in recent weeks
A prior report identified a Chinese advertising partner posting nudify app ads on Meta's platforms; the back-to-back findings point to systemic ad review failures rather than isolated incidents
Meta cites new AI detection technology as future safeguard
A Meta spokesperson stated recently launched AI technology would better detect and block violating ads at upload; critics note the ads ran for nine months before any action, predating only some of the violations
Security Alert | Industry Update | Stat | Policy | Legal | Context | Strategy
What This Means
This is one of the most serious platform safety failures documented for Meta, precisely because it involves paid advertisements — not user content — that were reviewed, approved, and monetized by the company. The nine-month window during which these ads ran, some reaching thousands of accounts across Europe, will dramatically intensify regulatory scrutiny under the EU's Digital Services Act, where supervisory action against Meta is already ongoing. The fact that this is the second such discovery in recent weeks indicates a systemic breakdown in Meta's ad integrity systems, not an edge case. Meta's pattern of removing harmful content only after press inquiries rather than proactive detection is likely to become a central exhibit in legislative hearings, potential criminal referrals, and civil litigation. This story will have lasting consequences for how AI-generated content in advertising is regulated globally.
Sentiment
Overwhelmingly alarmed and critical of Meta's automated ad moderation failures
“TTP alerted @WIRED to over 50 image and video ads containing AI-generated child sexual abuse material (CSAM) that we discovered in Meta's ad library. We also reported the ads—which made no effort to hide what they were promoting—to @NCMEC.”
“Meta ran & hosted dozens of paid ads that featured AI-generated child sexual abuse material. This is absolutely repulsive. Big Tech is willing to do anything that helps their profits - even if it threatens kids' safety & wellbeing. The public should be outraged.”
“Meta's ad review is "primarily automated." That sentence is the whole story. TTP researchers found 50+ paid ads with AI-generated CSAM in Meta's ad library. Reviewed, approved, monetized. One ran for months, reaching 2,563 accounts in Europe.”
“meta's ad screening is currently blind. they ran dozens of ads with ai-generated csam, flagging none of them. trusting black-box moderation models to stop harmful content without actual humans in the loop is a fundamental engineering failure.”
Split
Unified criticism of Meta's automated ad systems (~100% negative; no defenses found).
Sources
- Meta Ran Ads That Contained AI-Generated Child Sexual Abuse ImageryWired
- Meta Failed to Catch Hundreds of AI Child Abuse Ads. Some Included Images of Real Kids - WIREDWired
- “This is the AI men actually use”: Meta ads pushed apps nudifying real teens - arstechnica.comArs Technica
- San Francisco Orders Meta to Stop ‘Allowing’ AI Child Abuse AdsWired