The key facts

  • 332 ads containing child sexual abuse material were found on Facebook and Instagram since the start of the year
  • Most ads showed ordinary photos of adults and children undressed with AI and turned into sexual content
  • The ads promoted AI image- and video-generating apps, most of them developed in China

The Tech Transparency Project, which studies the activities of major technology companies, has found that Meta failed to comply with the rules when placing ads in its operations. According to the project's report announced on September 8, 2026, researchers have identified 332 ads on Facebook and Instagram containing material depicting sexual violence against children since the beginning of the year. In most of them, an ordinary photo of adults and children was undressed using artificial intelligence and turned into a sexually explicit video or image. These materials were not just ordinary user posts, but paid advertisements. They were removed from the platforms after researchers reported them to the company.

How did the banned ad get onto the platform?

Meta says it reviews ads before they are placed, and the company's advertising standards prohibit content that leads to the sexual exploitation of individuals, particularly children, or endangers them. However, the 332 ads identified in the TTP study suggest that this review system failed to stop them in time. The researchers identified the ads through Meta's public Ad Library. This tool still allows viewing ads distributed in certain regions.

TTP also drew attention to another problem: in the violation records for 113 ads, it was not explicitly stated that a rule concerning children had been broken. In some cases, rules regarding adult content were cited, or the record was limited to a general comment. This shows that the automated review system cannot always correctly identify prohibited material.

What did the ads promote?

Most of the ads identified by TTP directed users to apps that create images and videos using artificial intelligence. They were tools that allow altering a person's photo with AI, depicting them without clothes, or creating sexually explicit videos. The company noted that most of these ads were for apps developed in China.

The problem here is not in artificial intelligence itself, but in using the technology to serve harmful purposes — using images of other people, especially children, without their consent. A child's photo being posted on an open social network does not mean it can be used for any purpose. When parents post photos of their children, they must also take into account the purposes for which those photos may later be copied or altered.