• by Colin Kirkland , 2 hours ago

Following Meta’s $17.1 billion settlement for exploiting children on its social-media platforms, the tech giant is being accused of running ads including child sexual abuse material (CSAM) -- some of which featured real children -- according to a report from Wired.

Last month, an investigation led by the Tech Transparency Project (TTP) found that Meta had been running dozens of paid ads including explicit AI-generated CSAM material targeting users living in the U.S. and Europe. Some of the ads invited users to link to AI-powered undressing apps. 

According to Meta’s ad library, the CSAM ads were running across Facebook, Instagram, Messenger, and Threads. 

“Sexual exploitation is horrific, and we work aggressively to keep it off our platform,” a Meta spokesperson initially told Wired, which reached out to the company for information. 

“The majority of these ads had minimal reach and many were disabled before Wired shared them,” the spokesperson continued, adding that the ads pre-dated Meta’s newer AI-detection and blocking system.

However, Wired’s most recent report found that Meta’s improved AI system is failing to cut down on the spread of abusive ads, which researchers at TTP say are continuing to run. 

Over 250 additional ads featuring “graphic sexual scenes involving children -- including some that were identical to those Meta recently removed” -- have allegedly run on Meta-owned social platforms since the start of August. 

Furthermore, unlike the ads that TTP initially uncovered, this new batch featured images of real children, including “a minor belonging to a European royal family.” Other children depicted in CSAM ads are from the U.S., and one is a teen influencer. 

Meta’s review standards policy states that it reviews ads before they run to make sure ads align with the company’s ad standards, but TTP found 332 CSAM ads that went through the review process.

The ads detected by the TTP were shown to a total of 36,000 accounts in the EU and the U.K., according to the data in Meta’s ad library, but even more ads depicting child sexual abuse ran in additional countries.

According to the TTP investigation, nearly 80% of the CSAM ads (262)  ran in the U.S., and 84 ran in India (including 78 that ran after Meta was ordered by the Indian government to remove Instagram ads promoting CSAM), and 120 of the ads ran in Australia.

TTP believes that the majority of the CSAM ads were created by AI-image and AI-video generation apps created by Chinese developers, while two of the ad resellers were identified as Guandong Advertising Group Co. (GIMC) – a large ad agency controlled by the Chinese government –  and BlueFocus, an ad agency that works alongside the Chinese government’s propaganda unit.

“What we’re talking about here is not simply AI-generated child sex abuse material,” Katie Paul, the director of the TTP, told Wired. “There’s a lot of harm that can come from this all while Meta is raking in money in the process.”