Thousands of videos linked to child abuse networks
A new investigation by Spanish fact-checking organisation Maldita has raised serious concerns about AI-generated sexualised content involving minors circulating widely on TikTok, despite platform rules that explicitly prohibit such material.
According to the report, more than 20 TikTok accounts were found publishing over 5,200 videos depicting young girls in revealing clothing or suggestive poses, including bikinis, school uniforms and tight outfits. Collectively, these accounts have amassed over 550,000 followers and nearly six million likes, indicating significant reach and engagement.
Links to child pornography sales
More alarmingly, Maldita found that comment sections under these videos contained links directing users to Telegram groups selling child sexual abuse material. The organisation said it identified 12 such Telegram groups and reported them to Spanish police.
The report highlights how TikTok is allegedly being used as a traffic funnel, drawing viewers to encrypted platforms where illegal material is sold.
Monetisation through subscriptions
Maldita also found that some accounts were profiting directly by selling AI-generated images and videos via TikTok’s subscription feature, which allows creators to charge a monthly fee for access to exclusive content. Under TikTok’s revenue-sharing model, the platform reportedly takes around 50 per cent of subscription earnings.
This has raised questions about platform accountability, especially when harmful content is not detected or labelled appropriately.
Gaps in AI labelling and moderation
TikTok requires creators to label AI-generated content, and states that videos may be removed if deemed “harmful to individuals”. However, Maldita’s analysis found that most of the videos lacked AI watermarks or disclosures, making it difficult for moderators and users to identify synthetic content.
Some videos carried TikTok’s “AI Alive” watermark, a tool that converts still images into videos within the app, but many others showed no indication of AI use.
Platform responses
In statements to Euronews Next, both TikTok and Telegram said they are committed to preventing child sexual abuse material.
Telegram said it scans all media uploaded to public channels and claimed it removed more than 909,000 groups and channels containing child sexual abuse material in 2025 alone. The company argued that criminals relying on private groups and external platforms shows its public moderation systems are effective.
TikTok said 99 per cent of harmful content involving minors is removed automatically, while 97 per cent of offending AI-generated content is taken down proactively. The platform added that it works with the US National Center for Missing and Exploited Children (NCMEC) and acts swiftly to suspend accounts involved in child exploitation.
TikTok also told CNN that between April and June 2025, it removed over 189 million videos and banned more than 108 million accounts for policy violations.
Wider regulatory context
The findings come amid growing global scrutiny of social media platforms. Countries including Australia and Denmark, as well as the European Union, are either enforcing or debating age-based restrictions on social media to better protect minors.
Maldita’s report has renewed calls for stronger AI content detection, clearer labelling, and faster cross-platform enforcement, as synthetic media becomes increasingly difficult to distinguish from real imagery.