Meta Approved and Profited from AI-Generated CSAM Ads for Nine Months, Investigation Reveals

Meta's Ad Platform Approved Dozens of AI-Generated CSAM Ads

Meta ran more than 50 paid advertisements containing AI-generated child sexual abuse material across Facebook, Instagram, Messenger and Threads over a nine-month period, according to an investigation by the Tech Transparency Project1

1

. The ads, discovered in Meta's own ad library, targeted users in the United States, United Kingdom, and more than a dozen European countries between November 2025 and early August 20262

2

. At least one advertisement reached 2,563 accounts across Europe, including France, Germany, Ireland, Italy, the Netherlands, Spain, Sweden, and the United Kingdom1

1

.

The sexually exploitative content made no attempt to disguise what it promoted. Katie Paul, director of the Tech Transparency Project, emphasized that these were not posts from third parties slipping through content moderation, but ads that Meta's ad platform reviewed, approved, and allowed to run while collecting ad revenue1

1

. Several advertisements linked directly to AI nudify apps or undressing services that manipulate images4

4

.

Automated Ad-Review System Failed to Block Harmful Content

Meta's automated ad-review system, which the company relies on to screen advertisements before publication, failed to detect the child exploitation content2

2

. Researchers initially found approximately two dozen violating ads, but discovered around 30 more in the hours before WIRED published its investigation1

1

. Several of these newly discovered ads had gone live only after WIRED contacted Meta for comment, raising questions about the effectiveness of the company's AI detection systems3

3

.

Many of the offending Meta ads got past the company's new AI technology designed to better detect and block violating ads at upload1

1

. Most ads were not actively showing to users when researchers found them, but they remained available and apparently undetected in Meta's ad library for months1

1

. The repeated appearance of such advertisements suggests problems extending beyond individual moderation failures, as Meta reviews, distributes, and amplifies paid advertisements through its commercial systems while earning revenue from their placement4

4

.

Pattern of Systemic Failures in Child Safety Enforcement

This incident marks the second time in recent weeks that paid ads linked to AI-generated child sexual abuse material appeared on Meta's platforms1

1

. A July 2026 BBC investigation found Instagram serving paid advertisements in India using terms like "rape video" and "child video," directing users to Telegram channels selling illegal content for as little as Rs 994

4

. Following that investigation, India's Ministry of Electronics and Information Technology summoned Meta and ordered it to turn off CSAM-related advertisements within seven days4

4

.

Meta's history with child safety issues extends back years. In 2023, a joint investigation by the Wall Street Journal, Stanford, and UMass Amherst found Instagram's algorithms actively connecting and promoting a network of pedophile accounts3

3

. In March 2026, a jury ordered Meta to pay $375 million after New Mexico's Attorney General sued the company for endangering children and misleading the public3

3

.

Meta's Response and Ongoing Concerns

Meta removed the advertisements after WIRED raised questions, stating that sexual exploitation is horrific and the company works aggressively to keep it off its platform1

1

. The company noted that the majority of these ads had minimal reach and many were disabled before WIRED shared them, while pointing to removing over 36 million pieces of child sexual exploitation content last year2

2

. However, Meta's ad library now states the removed ads violated rules on child sexual exploitation, nudity, and sexual activity4

4

.

The Tech Transparency Project reported the material to the US National Center for Missing and Exploited Children4

4

. Several advertisements identified by researchers linked to an app called MaskAI, which allowed users to place faces into AI-generated sexual content. Apple removed the app from the App Store after WIRED contacted the company4

4

. The findings raise critical questions about whether Meta's safeguards can effectively stop illegal content before it is approved, paid for, and delivered to users on Facebook, Instagram, Threads, and Messenger.