Meta: Illegal Chinese AI Ads on Facebook

⚡

Key Takeaways

1Chinese companies have been running ads for illegal AIs on Facebook and Instagram, despite Meta's moderation efforts.

2GatherOne, a Chinese company, has been identified as having disseminated thousands of ads for face-swapping and nudity applications.

3Despite the removal of many ads, some remain active, highlighting the limitations of Meta's measures.

💡Why it matters — The persistence of these illegal ads underscores the challenges Meta faces in controlling harmful content and complying with international regulations.

Controversial Ads on Meta Platforms

Meta's social networks, namely Facebook and Instagram, have recently been the stage for the massive dissemination of ads for banned artificial intelligence (AI) applications. These ads, promoted by Chinese services, have highlighted the dark side of these platforms, where illegal content manages to slip through despite moderation efforts.

A Chinese company known as GatherOne is at the center of this issue. It is alleged to have orchestrated the distribution of tens of thousands of ads on Facebook and Instagram, promoting applications capable of generating images of naked women. Although Meta has begun to clean its platforms of such content, the company finds itself facing lawsuits for allowing other fraudulent ads to pass through.

A Massive and Organized Dissemination

The scale of the phenomenon is striking: approximately 31,000 ads for AI tools, including "face-swapping" applications, have been identified. These ads were promoted through 43 different Facebook pages, each contributing to the dissemination of over 7,600 distinct ads for controversial software.

Partially Effective Moderation Efforts

Meta has taken steps to remove these ads and disable the associated accounts. However, the Tech Transparency Project has pointed out that some of these ads remain active. A spokesperson for Meta stated that the company is taking "aggressive measures" against these applications, although they did not provide specific details in response to inquiries from the oversight group.

Internal data from Meta corroborates some of the allegations: out of the 210 identified pages, 179 had disseminated ads that were subsequently removed for violating rules regarding sexual exploitation, sexual solicitation, or nudity. Some of these ads featured women simulating sexual acts.

A Recurring Problem for Meta

This is not the first time Meta has faced this type of issue. In June 2025, the company filed a lawsuit against Joy Timeline HK Limited, a Hong Kong-based company, for disseminating over 87,000 ads through at least 135 Facebook pages. Meta had already begun taking action against this advertiser as early as 2023, in response to an Indian injunction in 2024 aimed at removing ads related to child abuse on Instagram.

The incriminating ads were explicit, promising users to "remove clothing" from any photograph. They primarily targeted countries such as the United States, Canada, the United Kingdom, Australia, and Germany. As early as 2024, CBS News had spotted hundreds of these ads, well before Meta took legal action.

Insufficient Countermeasures

Despite the efforts made, the measures taken by Meta have not been enough to curb the phenomenon. Researchers from the American Sunlight and Indicator projects identified an additional 4,215 ads for "nudity" applications in the three months following Meta's announcement of new detection technologies.

These ads are estimated to have reached around 359,547 users in the European Union and the United Kingdom. The operators of these ads circumvented blocks by creating new pages and accounts, sometimes registered under whimsical names like "ddd" or "111," in violation of the transparency rules imposed by the European Digital Services Act (DSA). The DSA requires platforms to disclose the identity of advertisers, a task made complex when these advertisers conceal their identities.