Meta Ran Ads That Contained AI-Generated Child Sexual Abuse Imagery

Meta, led by Mark Zuckerberg, has run numerous paid advertisements containing explicit AI-generated child sexual abuse material (CSAM) and images of minors engaging in sexually suggestive behavior, according to research shared with WIRED. These ads were targeted at individuals residing in the United States, United Kingdom, and more than a dozen European countries.

Researchers from the Tech Transparency Project (TTP) discovered over 50 image and video ads containing abusive content in Meta's ad library. The ads were published between November last year and the beginning of August, often running for several days and occasionally targeting men exclusively. One video advertisement utilized a thumbnail depicting a child sitting on the floor, with text above it stating "Realizing Deep Fantasies with Generation AI."

Upon clicking, the ad displayed video clips of adults engaging in sexual acts and inserted the child's face from the thumbnail into a sexual act. Another ad repeatedly published showed a young girl lying back with her legs spread, accompanied by text that read "I can show you more." Most of the ads had minimal reach, with at least one reaching 2,563 accounts in Europe.

Some of these ads were not actively shown to users when discovered but remained undetected in Meta's ad library for months. Meta acknowledged the issue, stating they work aggressively to prevent such content on their platforms. However, some of the offending ads had bypassed Meta's new AI technology designed to detect and block violating ads at upload.

Despite removing the ads from their library, Meta emphasized their ongoing efforts to combat child sexual abuse and exploitation material, including removing over 36 million pieces of such content last year and taking legal action against developers of nudify apps.