Investigators discovered that Facebook and Instagram carried hundreds of sponsored posts featuring AI-generated child sexual abuse material, raising urgent questions about Meta's advertising safeguards and the promotion of AI 'nudify' applications.
Hundreds of Ads Promoted AI Nudify Apps
A technology watchdog group published an investigation into hundreds of sponsored posts containing AI-generated child sexual abuse material that appeared on Facebook and Instagram in 2026.
Researchers at the Tech Transparency Project (TTP) identified 332 advertisements containing child sexual abuse material, according to an investigation published on 8 September. The publication noted that most of the children depicted in the ads were digitally altered using artificial intelligence tools to make them appear as though they were engaging in sex acts, while some ads contained animations of real children.
Investigators identified a 14-year-old influencer and a young member of a European royal family among the children featured in the material, finding that many of these promotional posts pushed China-based applications capable of digitally 'nudifying' subjects.
The report noted that 298 of the 332 concerning ads discovered on Meta directed users to AI image- and video-generation applications available through Apple's App Store or Google Play, marketing software that could be used to produce or access child sexual abuse material, according to the researchers.
AI Apps Used Children in Explicit Promotions
Apple removed one promoted programme called Lumin - AI Partner from its marketplace after it advertised the ability to 'Bring Your Dream AI Partner To Life Instantly.'
Investigators noted that a promotional clip featured 'an elementary-school-aged girl who appeared to be still missing her adult teeth', accompanied by the phrase 'Realising Deep Fantasies with Generation AI.'
Researchers said that clicking on the ad caused the child's face to be superimposed onto an adult engaging in sex acts.
The report described other examples of child sexual abuse material, including images of preteen and elementary school-aged girls, featuring text that hinted at an opportunity to turn users' 'hidden desires into hypnotic videos'.
'There are no restrictions,' a voiceover reportedly said over a video of an underage girl engaging in a sex act. 'All the characters are real people, and there are many options. This is the AI men actually use.'
Apple Says Nudify Apps Violated Its Rules
'The App Store was designed to be a safe and trusted place for users, and we have zero tolerance for developers who deliberately attempt to evade our review process,' Apple spokesperson Adam Dema told SFGATE in a written statement Wednesday.
'In this case, the developers mentioned in this report initially submitted apps that are compliant with the App Review Guidelines, only to later surface nudify capabilities, which are against our rules.'
Apple said 17 of the 20 applications highlighted in the findings had already been removed, with the company taking action against the remaining three.
Meta Defends Its Content Moderation
Meta addressed the wave of sponsored content through a representative, Tracy Clayton, who informed SFGATE on Wednesday that 'We don't tolerate nudify apps or any kind of child exploitation, whether real or AI-generated.' Clayton pointed out that many of the flagged advertisements had already been removed, noting that most had fewer than 200 impressions, while total spending stayed below $5,000.
Clayton explained that bad actors 'constantly shift tactics to evade detection', making ongoing improvements to safety systems necessary.
Meanwhile, leaders at the Tech Transparency Project voiced doubts regarding the platform's policing efforts. TTP Director Katie Paul told SFGATE in an emailed statement Wednesday that Meta claims it takes a hard stance against child sexual abuse material, but TTP found hundreds of ads containing CSAM that ran on Facebook and Instagram.
Paul pointed out that Meta took a week or more to respond when researchers reported CSAM advertisements and often declined to remove them, saying they did not violate Meta standards, leaving observers doubtful about the firm's capacity to handle synthetic explicit imagery effectively.
