Meta's ad systems approved AI-generated child sexual abuse material, researchers say

Meta’s own ad-review systems approved more than 50 paid ads containing AI-generated child sexual abuse material and sexually suggestive content involving minors, researchers at the Tech Transparency Project said.

The image and video ads ran across Facebook, Instagram, Messenger and Threads between November 2025 and August 2026, reaching thousands of accounts in the United States, the United Kingdom and more than a dozen European countries, according to the group. Several promoted “nudify” apps that digitally undress people in photos.

That Meta’s automated review cleared the ads before they went live is the core of the finding. The AI-generated CSAM ads passed the same systems the company has said keep exploitative content off its platforms, pointing to a gap between that assurance and what shipped to real users for months.

Meta removed the ads from its ad library after reporters flagged them. “Sexual exploitation is horrific, and we work aggressively to keep it off our platform,” the company said in a statement. It added that most of the ads had minimal reach, many were disabled before they were flagged, and many predated AI-detection tools it recently launched.

The count and reach come from the advocacy group’s own review and from outlet reporting rather than an independent audit, and Meta disputed the scale of the ads’ exposure. Even so, the case adds to evidence that automated ad approval is struggling to catch synthetic abuse imagery, and it lands as regulators in the US and Europe weigh tighter rules on AI-generated sexual content.