Facebook, Instagram Ran Thousands of Ads for AI Apps That Digitally Undress Women, Simulate Sex Acts

Meta Platforms approved thousands of adverts for artificial intelligence apps that digitally undress women and generate explicit sexual imagery on Facebook and Instagram, despite its own ban on such content, according to an investigation published on Saturday.

The Tech Transparency Project (TTP), a Washington-based non-profit that monitors major technology firms, said the adverts were traced to GatherOne, one of Meta's authorised advertising partners in China.

The findings raise questions over how prohibited AI-generated sexual material continued to appear on Meta's platforms via a partner with preferred access to its advertising network.

Meta's Advertising System Carried Thousands of Prohibited AI Ads

TTP identified roughly 7,600 adverts promoting AI 'nudify' applications capable of removing clothing from photographs or placing people's faces onto nude or pornographic bodies and videos.

Researchers linked the campaigns to 43 Facebook pages connected to GatherOne. The adverts promoted applications offering features such as digital undressing, AI-generated intimate imagery and face-swapping tools.

Meta's advertising standards prohibit adverts promoting adult sexual exploitation, non-consensual intimate imagery and nudify services. Many of the adverts documented by researchers remained active before they were later removed.

Investigation Expanded Beyond Individual Apps

As researchers broadened their review, they identified a wider network of 210 Facebook pages connected to GatherOne that collectively ran about 30,800 adverts promoting AI image-generation and face-swap tools.

According to TTP, about 85 per cent of those pages later had adverts removed after Meta found they violated company policies on adult sexual content and exploitation.

The findings indicate that prohibited AI-generated sexual content continued to circulate through multiple advertising accounts rather than through isolated campaigns.

One App Prompted Child Abuse Complaints

Among the applications highlighted in the investigation was BAfter, which advertised features including 'One-Click Undress' and subscription services providing access to additional AI-generated imagery.

Researchers said the application included a public gallery where users shared AI-generated images. Reviews posted on Google's Play Store said some user-generated content appeared to depict children in sexually explicit images.

One reviewer highlighted what they described as 'ILLEGAL CHILD PORN' inside the application.

The app stayed on Google Play with an 'Everyone' age rating until shortly before publication of the investigation.

Meta later confirmed it had banned BAfter from its advertising platform.

GatherOne Held Preferred Meta Partner Status

The investigation also focused on GatherOne's position within Meta's advertising business.

GatherOne, based in Beijing and Hong Kong, is one of a small group of Chinese advertising agencies authorised by Meta to help businesses in China reach international users through Facebook and Instagram.

According to TTP, the company manages more than 30,000 adverts each month and received four performance awards during Meta's 2026 Agent Summit in Hong Kong.

GatherOne said it has suspended new advertising accounts promoting nudify applications, face-swap services and AI companion products while it strengthens its compliance procedures.

AI Deepfake Enforcement Faces Growing Scrutiny

Meta said advertisers that breach its advertising policies risk campaign removal, financial penalties and permanent account suspension. The company said it had removed the adverts identified in the investigation and banned BAfter after reviewing the findings.

The investigation comes as governments increase scrutiny of AI-generated intimate imagery and deepfake abuse. The US Take It Down Act, which criminalises publication of non-consensual intimate images, came into force earlier this year.

The findings add to scrutiny of how major technology companies enforce advertising policies as AI-generated sexual content becomes more widely available and harder to moderate at scale.