AI nudify apps have hit 90 schools across 28 countries. How it happens, what parents and schools must do, and how synthetic CSAM is detected. Read the guide.

✓ Fully verified. All statistics are drawn from primary sources: Wired/Indicator investigation; UNICEF/ECPAT/INTERPOL Disrupting Harm Phase 2 study across 11 countries (February 2026); IWF data release (November 2025); NCMEC CyberTipline data; Internet Matters UK survey (2024). Source links at the foot of this post.

⚠️ Content note: This post discusses AI-generated sexual imagery of children in the context of prevention, education, and child protection. If a child in your life has been targeted, Report Remove is available at childline.org.uk/report-remove. Childline is free on 0800 1111, 24 hours.

All it takes is one photo posted online. A school photo. An Instagram picture. A screenshot from a video call. Using freely available apps, many with minimal or no age verification, a teenage boy can produce a fake nude image of a girl in his class in minutes. The technology is not sophisticated. The harm is real. And in a UK survey of eight schools, students reported it was happening at their school while receiving no education on the subject.

The scale of the problem

  • 📊: 28 countries: Where AI deepfake nudes of schoolchildren have been reported since 2023 (Wired/Indicator)
  • 👥: 1.2M children: Affected globally in the past year, UNICEF/ECPAT/INTERPOL study across 11 countries
  • 🏫: 1 in 25 children: In some study countries, the equivalent of one child in a typical classroom
  • 📈: ×2.1 increase: IWF AI-generated CSAM reports doubled, 199 (Jan–Oct 2024) to 426 (same period 2025)

These numbers undercount the reality. Most children who have images made of them do not know. Many who do know do not tell anyone. A UK study of teenagers found that even when incidents were occurring at their school, pupils had received no education about what deepfakes are, whether it was legal, or what they could do about it. The shame is immediate; the knowledge of what to do is almost entirely absent.

"Deepfake abuse is abuse, and there is nothing fake about the harm it causes. When a child's image or identity is used, that child is directly victimised."

, UNICEF statement, Safer Internet Day, February 2026

How it is actually happening in schools

This is not an abstract technological risk. The mechanism is documented in police reports, court cases, and academic research across multiple countries. It follows a consistent pattern.

  • 01: Source images from social media: Boys harvest photos of girls from Instagram, Snapchat, school websites, yearbooks, and WhatsApp chats. A single face photograph is sufficient. Apps require no technical skill, a phone and a photo is all that is needed.
  • 02: Nudify apps are freely accessible: Dozens of "nudify" apps are available online with minimal or no age verification. Many have an age declaration requirement, a checkbox, and nothing more. The Wired investigation found these apps were used in every documented school case.
  • 03: Images circulate in peer networks: The images are shared via group chats, Snapchat, Discord, and gaming platforms. In documented cases, boys created organised folders with fake images of multiple girls from the same class. Circulation is rapid and largely invisible to school staff.
  • 04: Victims often do not disclose: UK research found that many girls who discovered images had been made of them did not report it, because they did not believe action would be taken, feared blame, or did not know it was illegal. Teachers often handled incidents inconsistently because they were unaware themselves.

⚠️ This is a crime in the UK. Creating or sharing a sexually explicit deepfake image of any person without consent is illegal under the Online Safety Act 2023. For images of children under 18, it also constitutes an offence under the Protection of Children Act 1978 regardless of whether the image is real or AI-generated. Schools and parents need to communicate this clearly, many children on both sides of this do not know the legal position.

The knowledge gap is part of the problem

A peer-reviewed UK study published in April 2026, drawing on interviews with students and teachers across eight schools, found a consistent picture: teachers underestimated how easy nudify apps are to use, did not consistently treat deepfake imagery the same way as real non-consensual images, and responded to incidents inconsistently. Students, meanwhile, associated AI with large language models like ChatGPT and had no conceptual framework for what had been done to them or their peers.

An Internet Matters survey found that 13% of UK teenagers had directly experienced nude deepfakes. Separately, 61% of children and 45% of parents said they do not know or understand the term "deepfake." The problem is happening at scale while the understanding needed to address it, across children, parents, teachers, and platforms, lags significantly behind.

💜 On the gendered dimension: research consistently shows that 98–99% of deepfake sexual imagery depicts girls or women. Boys are overwhelmingly the creators; girls are overwhelmingly the targets. Understanding this is essential for schools designing their response, both the support offered to victims and the education provided to pupils about consent, harm, and the legal consequences of creating this material.

What this means for parents, schools, and platforms

For parents

  • Review what photos of your child are publicly visible on social media. Instagram, Snapchat, Facebook, TikTok. Public posts are harvestable. Private accounts are not foolproof but are significantly harder to scrape.
  • Have the conversation before it happens. Explain what nudify apps are, that creating or sharing these images is illegal, and, critically, that if something has happened to them, your first response will be support, not anger.
  • If your child tells you an image has been made or shared: do not pay anyone, screenshot the evidence, contact CEOP, report to the platform, and contact Report Remove if images are circulating online.
  • Victims are not at fault. The shame children feel is real but it is misdirected, this is done to them, not by them.

For schools and safeguarding staff

  • Update your safeguarding policy before an incident happens. Deepfake sexual imagery of pupils should be treated identically to real non-consensual imagery, same process, same seriousness, same support for the victim.
  • Train staff on how nudify apps work and what pupils may be experiencing. Teachers who underestimate the ease of the technology respond to incidents less consistently.
  • Victims often disclose to a peer before an adult. Ensure there are clear routes for young people to bring this to trusted staff, and that those staff know what to do when they do.
  • Include this in RSHE education. Consent, the legal position on deepfakes, and the harm caused are all teachable. Students who understand the law are more likely to report, and less likely to participate.
  • Review your school's public image use. School websites and social media feeds can be scraped for student images. UK Safer Internet Centre guidance recommends auditing public-facing imagery.

For platforms and developers

  • Nudify apps with minimal age verification are the primary tool in every documented case. Age verification that goes beyond self-declaration is the single most direct intervention available to platforms in this ecosystem.
  • Detection infrastructure matters. Images are circulating on the same platforms children already use. Discord, Snapchat, Instagram, gaming chats. Identifying AI-generated imagery before it spreads is technically possible and increasingly a regulatory expectation.
  • UNICEF has called explicitly for platforms to proactively prevent the circulation of these images, "instead of merely removing them after they have been shared." Detection before distribution is the standard being demanded.

The conversation starters that matter

These are not alarmist. The goal is to name the thing clearly, communicate the legal position, and make clear that disclosure leads to support rather than blame.

"You've probably heard of deepfakes, AI-generated videos and images. There are apps now that can take a photo of someone and create a fake nude image of them. This is illegal, for anyone, any age, and it's happening to girls in schools. If you hear of it happening, or it happens to you or a friend, I want you to come to me."

"If someone you know has made an image like this of you, or threatens to share it, you've done nothing wrong. My first response would be to help you, not to be angry. The person who made it is the one who has broken the law."

"Have you heard of anything like this at school? I'm not looking to get anyone in trouble, I want to understand what's going on."

How Tuteliq detects synthetic child sexual imagery

The platforms where these images are created and shared, nudify apps, messaging platforms, social networks, gaming chats, are the environments where detection infrastructure can intervene. Tuteliq's synthetic content detection is designed specifically for this.

Every endpoint is designed to surface AI-generated child sexual imagery before it spreads, aligned with UNICEF's call for proactive prevention, not reactive removal.

  • 🔍: AI-Generated Image Detection: A 6-signal forensic pipeline: vision AI analysis, EXIF metadata extraction, pixel statistics, C2PA Content Credentials, watermark detection, and perceptual hashing. Identifies nudify app output with high confidence even after resaving or minor edits.
  • 🛡️: Synthetic CSAM Classification: Returns classification across the full severity spectrum, confirmed_synthetic, suspected_synthetic, unknown. Category tags include SYNTHETIC_CSAM and AI_MANIPULATED_MEDIA. Age calibration multiplier applied for under-18 content context.
  • ⚠️: Unsafe Content Detection: Identifies sexually explicit content in images, voice, text, and video. Detects the sharing of non-consensual intimate imagery in chat environments, the primary distribution vector for deepfakes in school-age peer groups.
  • 🔌: Platform Integrations: Tuteliq integrates with Discord, Roblox, Slack, and Telegram, the platforms where peer sharing of these images occurs. For nudify apps and image generation platforms, detection can run at the point of creation rather than distribution.

For children. Report Remove

If images of you have been created or shared, they can be removed.

The IWF's Report Remove service allows children in the UK to report sexual imagery of themselves, including AI-generated deepfake images, and have it blocked and removed from the open web. The service is free, confidential, and can act proactively to prevent sharing before a criminal does so.

Visit Childline Report Remove or go directly to iwf.org.uk. In the US, the NCMEC's Take It Down service provides the same function.

Support is available

---

Sources: Wired / Indicator investigation, nearly 90 schools and 600+ pupils across 28 countries since 2023 · UNICEF/ECPAT/INTERPOL Disrupting Harm Phase 2 study, February 2026, 1.2 million children across 11 countries · IWF data release, November 2025, 199→426 AI CSAM reports · Internet Matters survey (2024), 13% of UK teenagers experienced nude deepfakes · Sexualized Deepfakes in UK Schools, MDPI peer-reviewed study, April 2026 · US News/AP, December 2025, NCMEC reports soared from 4,700 (2023) to 440,000 (first half 2025). UNICEF quote drawn directly from press release of 4 February 2026.

Deepfake Nudes in Schools: 28 Countries, 1.2M Children

Wired/Indicator investigation found nearly 90 schools and 600+ pupils affected across 28 countries since 2023. UNICEF/ECPAT/INTERPOL study puts the global figure at 1.2 million children — 1 in 25 in some countries.

IWF AI-generated CSAM reports doubled (199→426). NCMEC reports surged from 4,700 (2023) to 440,000 (first half 2025). 13% of UK teenagers have experienced nude deepfakes. 61% of children don't know what a deepfake is.

Get Started Free · Read Documentation · View Pricing