The key facts

  • TTP found 332 ads with AI-processed images of children published since the beginning of 2026
  • In 113 cases, the violation records did not directly indicate a violation of the rules concerning children
  • Most ads led to AI apps, a significant share of which were created in China

What the Investigation Revealed

The Tech Transparency Project (TTP), which studies the activities of major technology companies, reported in a report published on September 8, 2026, that Meta was not complying with its own rules when placing ads. Since the beginning of 2026, researchers have found 332 ads on Facebook and Instagram containing materials depicting the sexual abuse of children. In most cases, ordinary photographs of adults and children were turned into sexualized images or videos using artificial intelligence. These were not ordinary user posts but paid advertising. After researchers contacted the company, the materials were removed from the platforms.

How Prohibited Ads Passed Review

Meta claims that it reviews ads before publication and that its advertising standards prohibit content that leads to the sexual exploitation of people, particularly children, or creates a risk for them. However, as the TTP study shows, the review system failed to stop these 332 ads in a timely manner. Researchers found the ads through Meta's public Ad Library, which still allows viewing ads that were shown in certain regions.

Problems with Recording Violations

TTP also drew attention to another problem: in the violation records for 113 ads, it was not directly stated that a rule concerning children had been violated. In some cases, adult content rules were cited or a general formulation was given. According to the researchers, this shows that the automated review system does not always correctly recognize prohibited materials.

What the Ads Promoted

Most of the ads discovered by TTP led to apps that create images and videos using artificial intelligence. These tools made it possible to alter a person's photo with AI, depict them without clothing, or create sexualized videos. Meta noted that a significant share of these apps was developed in China.

The Essence of the Problem

According to TTP's assessment, the problem is not artificial intelligence itself, but the fact that the technology is being used for malicious purposes — using images of other people, especially children, without their consent. Publishing a child's photo on an open social network does not mean it can be used for any purpose; parents, when posting photos of their children, should keep in mind that they may later be copied or altered.