AI-generated videos purportedly depicting underage girls in sexually suggestive clothing and poses are racking up millions of likes on TikTok despite the platform banning such content, according to a new study by online safety groups.
Researchers identified more than a dozen accounts that posted videos of AI-generated girls wearing tight clothing, lingerie, and school uniforms, sometimes in suggestive poses. These accounts have hundreds of thousands of followers in total. Many comments under the video posted links to Telegram chats offering to purchase child pornography.
As of Wednesday evening, 13 accounts remained active after 15 of them were reported using TikTok's reporting tools last week, said Carlos Enriquez Echevarria, who led the investigation as deputy director and head of public policy at Maldita.es. A Spanish non-profit organization that researches online misinformation and promotes media transparency. The report was released on Thursday.
The report said TikTok is facing increasing regulatory pressure regarding online safety, particularly in the wake of Australia's law banning minors from using the social network, which came into effect this week.
“There's no nuance to this at all,” said CNN's H. Enriquez-Echevarria. “I don't think anyone would actually think this is acceptable and want it removed.”
Impact on platform policies and TikTok's response
TikTok says it does not tolerate any content that “depicts, promotes, or solicits pedophilia or the exploitation of minors,” and bans accounts that focus on AI-generated images of minors in adult clothing or sexually suggestive poses or facial expressions. Another policy section prohibits “sexual content involving minors, including anything depicting or encouraging abuse or sexual acts,” as well as “AI-generated images” and anything that sexualizes or fetishizes young people's bodies.
The company notes that content management involves a human team and uses a combination of visual, audio, and text tools. From April to June 2025, TikTok removed more than 189 million videos and banned more than 108 million accounts. Our data shows that 99% of content that violates our policies on nudity and exposure of young bodies, including minors, was proactively removed, and 97% of content, including AI-generated content, was also proactively blocked.
A TikTok spokesperson did not comment separately on this report.
“Telegram is committed to preventing child sexual abuse (CSAM) content on our platform and adheres to a strict zero-tolerance policy,” Telegram spokesperson Remi Wong said in an interview with CNN. “Telegram checks all media uploaded to our public platforms against the CSAM database and is removed by moderators to prevent it from spreading. Although our encrypted platform does not allow proactive screening of content in private groups, Telegram accepts reports from NGOs around the world to ensure compliance with our Terms of Service.”
In 2025, Telegram removed more than 909,000 public groups and channels related to child sexual exploitation content, the group said.
Maldita.es said it identified the TikTok video using a test account that it uses to monitor potential misinformation and other harmful content as part of its work.
“One of our team members noticed a trend of (AI-generated) videos featuring young children dressed as adults, and the comments seemed particularly financially motivated,” Enriquez-Echevarria said.
According to the report, some accounts described their videos as “cute schoolgirl” or “young model” on their profiles. “A more sophisticated video featuring a young girl licking ice cream is filled with vulgar sexual comments,” the document states.
In some cases, Enriquez-Echevarria said, account holders used TikTok's AI Alive feature that animates static images to turn AI-generated images into videos, but Enriquez-Echevarria said other videos may have been created using third-party AI tools.
Comments on many videos contained links to private Telegram chats where child pornography was promoted, the report said.
“There's no way anyone can look at this and not understand what's going on,” Enriquez-Echevarria said. “The comments are incredibly crude and full of the worst people on the planet who write things like that.”
By mid-week, at least one account and one video that had not previously violated the rules was disabled. Enriquez Echevarria added that it is unclear why it was not removed immediately after the initial report.
The report follows another Global Witness study released in October that found TikTok led young users to sexually explicit content through search prompts. In response, TikTok said it has already removed content that violates its rules and implemented improvements to its search prompt system.
