A child doesn’t have to take an intimate photo for someone to create one of them anymore. Generative AI can take that everyday image and fabricate sexual content depicting a child who never participated, never removed their clothes, and may not even know the image exists.
New data from the Internet Watch Foundation (IWF) shows just how quickly this form of exploitation is evolving. And a major new UNICEF report reveals something equally important: behind the technology are children navigating coercion, unwanted sexual content, image-based abuse, extortion, shame, and barriers to getting help.
Together, the reports show AI hasn’t created the sexual exploitation of children. But it is giving perpetrators new ways to produce, manipulate, and scale abusive content—and sometimes they don’t even need to contact a child first.
IWF Recorded a 154% Increase in Reports Containing Realistic AI-Generated Child Sexual Abuse
In its 2025 Annual Data & Insights Report, the Internet Watch Foundation documented a rapid increase in realistic AI-generated child sexual abuse material. The report recorded 491 reports containing realistic AI-generated child sexual abuse imagery, up from 193 reports in 2024.
That’s a 154% increase in one year. And that’s just recorded reports.
Those 491 reports contained 8,029 AI-generated criminal images and videos, along with an additional 82 items that IWF classified as prohibited under a separate UK legal framework.
It’s important to understand what those numbers do—and don’t—mean. The 154% increase describes the number of reports containing realistic AI-generated abuse imagery, rather than a 154% increase in the total number of individual criminal files. IWF also cautions that its operational data reflects what its analysts locate through proactive investigations and what others report to the organization. It should not be treated as a measurement of the total amount of AI-generated child sexual abuse material worldwide.
Even within that limited window, however, one development stands out.
AI-Generated Abuse Video Exploded From 13 Videos to 3,443 in One Year
In 2024, IWF identified just 13 AI-generated child sexual abuse videos. In 2025, it identified 3,443. That’s more than 260 times as many. IWF says several factors may have contributed to the increase, including changes in where analysts looked for abusive material and what people reported to the organization. But its analysts also pointed to the expanding availability of open-source AI video models during 2025.
The technology itself changed quickly.
According to IWF analysts, AI-generated videos became dramatically more realistic over the year. Tools that once produced relatively simple clips can increasingly generate complex moving scenes and synthetic audio. And the material wasn’t simply becoming more plentiful. IWF classified 65% of the AI-generated abuse videos it identified as Category A, the UK’s most severe classification of child sexual abuse imagery.
A Child Doesn’t Need to Create Sexual Content to Become the Target of It
A perpetrator can now use a nonsexual photograph to fabricate sexual imagery. The targeted child doesn’t need to take an intimate photo, send one to anyone else, or participate in creating the resulting content. UNICEF’s new Through Children’s Eyes: How Digital Technologies Enable Child Sexual Abuse report describes this as an escalating challenge.
The report draws on the Disrupting Harm research project conducted by UNICEF Innocenti, ECPAT International, and INTERPOL. Researchers combined nationally representative surveys of approximately 21,000 internet-using children ages 12–17 across 21 countries with in-depth interviews involving more than 100 young people who experienced technology-facilitated sexual exploitation or abuse during childhood. Researchers also interviewed more than 500 frontline professionals and law enforcement personnel and reviewed national laws and policies. UNICEF Office of Strategy and Evidence – Innocenti. 2026, September. Through Children's Eyes: How digital technologies enable child sexual abuse. UNICEF Innocenti. https://www.unicef.org/innocenti/reports/through-childrens-eyes Copy
For the second round of the project, researchers added a particularly timely question. Children in nine countries were asked whether someone had used artificial intelligence to create fake sexual images or videos depicting them.
From those surveys, researchers estimated that more than 1.1 million internet-using children ages 12–17 across the nine countries had experienced this form of AI-enabled sexual abuse in a single year. That estimate should not be interpreted as a worldwide prevalence figure because the AI-specific question appeared only in the project’s second round and therefore covers nine countries: Armenia, Brazil, Colombia, the Dominican Republic, Mexico, Montenegro, North Macedonia, Pakistan, and Serbia. It’s likely higher.
But even within that limited geographic scope, more than a million children represent a striking reality. AI-generated sexual abuse isn’t merely a hypothetical risk experts expect to encounter someday. Children are reporting that it has already happened to them.
The Image Can Be Fake While the Violation Is Real
One young woman interviewed for the UNICEF research described receiving a digitally manipulated image created from a real photograph of her. Someone had taken her face and placed it onto another person’s body in a way that made the resulting image appear authentic.
When people hear “AI-generated,” it’s easy to focus on whether an image documents an event that actually occurred. But that question alone misses the experience of the person depicted. UNICEF emphasizes that fabricated sexual images can still produce serious consequences. Children interviewed through Disrupting Harm described shame, stigma, anxiety, social judgment, and fear surrounding sexual content circulated about them, including manipulated content.
The person is real. Their classmates are real. Their family is real. Their reputation, relationships, sense of safety, and uncertainty about where the image might spread are real. AI changes how the content gets created. It doesn’t erase the person whose identity someone has sexualized.
AI Can Also Give Abusers Something to Threaten Children With
Synthetic imagery introduces another disturbing possibility: a perpetrator can manufacture the very material they later use for coercion. Traditionally, image-based sexual extortion often involves threatening to distribute an existing intimate image unless the targeted person sends more content, provides money, or complies with another demand.
With generative AI, an authentic, intimate image isn’t always necessary. UNICEF’s report notes that AI tools can fabricate compromising material that perpetrators may use as leverage, even when no real sexual image ever existed.
That development fits into a much broader pattern documented by the research. Across the 21 countries studied, UNICEF estimated that approximately 20 million internet-using children ages 12–17 experienced at least one measured form of technology-facilitated sexual exploitation or abuse in a single year—about one in five on average across those countries.
The forms measured included unwanted sexual content, unwanted sexual conversations or requests for images, offers of money or gifts for sexual content or contact, nonconsensual sharing of sexual images, and threats or blackmail involving sexual activity.
Researchers estimated that, in a single year:
- Nearly 15 million children were exposed to sexual content they didn’t want to see.
- Nearly 9 million were asked to engage in unwanted sexual conversations or share sexual images or videos.
- About 5.5 million were offered money or gifts in exchange for sexual content or sexual contact.
- About 4 million had someone share sexual images of them without permission.
These estimates come from surveys conducted at different points between 2020 and 2025, and UNICEF cautions against treating differences among countries as rankings. Social norms, stigma, awareness, and willingness to disclose abuse can all influence self-reported prevalence.The broader finding, however, remained consistent across countries: technology is already woven into many forms of sexual exploitation children experience.
AI adds another tool to that environment.
Most of the Abuse UNICEF Documented Happened on Familiar Platforms
One of the most important lessons from the UNICEF report is that technology-facilitated sexual abuse doesn’t exist only in hidden online communities. Almost 60% of the cases captured in the study happened on social media, while 14% happened in online games.
The five social media platforms children most frequently identified were Facebook, WhatsApp, Instagram, Snapchat, and TikTok. Because a single incident could move across multiple platforms, those platform percentages overlap. And perpetrators weren’t always anonymous strangers.
Across the 21 countries, 57% of cases involved someone the child already knew, including friends, relatives, acquaintances, or romantic partners. In 38% of cases, the child first met the perpetrator online. A child may be harmed by a stranger. But abuse can also begin with a classmate, friend, romantic partner, relative, or someone else who already has access to the child’s photographs and social world. The technology can then extend that person’s reach.
A private message can become a persistent contact. An ordinary photograph can become source material. A fake image can become a threat. A file can move from one person’s device into a group chat, and then somewhere the targeted child may never find.
More Than 4 in 10 Cases Were Never Disclosed
Technology can make sexual abuse easier to create and distribute. Shame can make it harder to talk about. UNICEF found that 41% of the technology-facilitated sexual exploitation and abuse cases captured by its research were disclosed to no one. When children did tell someone, they most commonly turned to a friend, sibling, or mother or female caregiver. Fewer than 1% of cases were reported to police, a social worker, or a helpline.
When researchers asked why children stayed silent, the most common response was that they didn’t know where to go or whom to tell. Others reported embarrassment, shame, fear of getting into trouble, or believing what happened wasn’t serious enough to report.
Some children interviewed by researchers raised another concern adults should pay attention to: they worried that telling a parent about something harmful online would result in losing access to their phone, internet, or social media. That can create a painful calculation.
A young person may need help while simultaneously fearing that asking for help will result in punishment. UNICEF’s recommendation is therefore bigger than simply telling children to report abuse. Families, schools, platforms, and institutions need to create environments where children can expect support when they speak.
The Mental Health Findings Are Serious, but They Don’t Establish Causation
UNICEF also examined how technology-facilitated sexual exploitation and abuse correlated with children’s well-being. Across the countries studied, children reporting these experiences were approximately four times more likely to report suicidal thoughts and behaviors and four times more likely to report self-harm tendencies. They also reported anxiety levels approximately 11 percentage points higher.
The analyses used cross-sectional survey data. That means researchers identified strong associations, but the study cannot establish that technology-facilitated sexual exploitation and abuse directly caused the differences in mental health. UNICEF explicitly cautions readers against making that causal leap. Still, researchers observed the pattern across diverse countries, and interviews added important context. Young people described fear, shame, isolation, loss of trust, anxiety, and ongoing distress following abuse.
The findings reinforce why fabricated sexual imagery shouldn’t be dismissed simply because software created it. For the child whose face appears in the image, the social and psychological consequences don’t disappear when someone says, “It’s fake.”
AI Is Making an Existing Consent Problem More Scalable
A child can’t consent to being sexually exploited. And AI now makes it possible for someone to manufacture sexual depictions of a child without requiring any sexual participation from that child at all.
Children deserve to understand that using another person’s photograph to fabricate sexual content violates that person’s boundaries. Young people also need age-appropriate education about consent, coercion, image-based abuse, healthy relationships, and what to do when something harmful happens.
Fight the New Drug’s school presentations do just that. Bring awareness to your school, community or organization today and see the difference education on the harms of pornography and AI imagery can make. Click below to learn more.
UNICEF stresses that education alone can’t become another way to place the burden on children. Children can learn digital skills. They can use privacy settings. They can block accounts, recognize warning signs, and tell trusted adults. This is important, critical today, but it’s just as important that they cannot redesign an AI model, they can’t control whether a platform recommends them, and they cannot force a company to remove an abusive tool. And they cannot single-handedly build a legal and technological environment capable of responding when new forms of exploitation emerge.
Safety Has to Be Built Into the Technology
UNICEF’s recommendations place substantial responsibility on the adults, institutions, governments, and companies with the power to change the environment around children. The report calls for platforms to adopt safety-by-design approaches, including stronger privacy defaults for children, restrictions on unsolicited adult contact, effective moderation, rapid removal of abusive material, and proactive detection of child sexual exploitation.
For AI specifically, UNICEF calls for guardrails that prevent AI-powered tools from generating sexual content depicting children or otherwise facilitating their sexual abuse. IWF reaches a similar conclusion. Its 2025 findings show what can happen when increasingly powerful generative tools become accessible to people seeking to create abusive material. IWF has called for AI products to incorporate safety by design and for lawmakers to address tools built to generate abusive sexual imagery.
Trying to remove every abusive image after someone creates it will always leave children absorbing some of the consequences. Prevention has to begin upstream. AI developers can make choices about training data, model safeguards, abuse testing, access controls, and how their systems respond to attempts to sexualize children. Platforms can make choices about whether they allow services that generate this material to advertise, operate, or distribute content. Governments can update laws as methods of abuse change.
Those choices determine whether technology makes exploitation easier or creates meaningful barriers against it.
New Technology Doesn’t Change the Value of Consent
The numbers from IWF show a technology evolving at an extraordinary speed. The testimony and survey data from UNICEF show why that evolution matters.
AI can now create sexual imagery of a child who never posed for it, never sent an intimate image, and never interacted with the person generating it. It can manipulate the likeness of real children, manufacture material for harassment or extortion, and produce enormous quantities of abusive content faster than previous technologies allowed. But beneath all that technological complexity lies a principle that hasn’t changed.
A child is a person, not raw material for someone else’s sexual gratification. And we’d argue that a person is a person, capable of thoughts, feelings, and depth, not an object, and shouldn’t be consumed as such.
Your Support Matters Now More Than Ever
Most kids today are exposed to porn by the age of 12. By the time they’re teenagers, 75% of boys and 70% of girls have already viewed itRobb, M.B., & Mann, S. (2023). Teens and pornography. San Francisco, CA: Common Sense.Copy —often before they’ve had a single healthy conversation about it.
Even more concerning: over half of boys and nearly 40% of girls believe porn is a realistic depiction of sexMartellozzo, E., Monaghan, A., Adler, J. R., Davidson, J., Leyva, R., & Horvath, M. A. H. (2016). “I wasn’t sure it was normal to watch it”: A quantitative and qualitative examination of the impact of online pornography on the values, attitudes, beliefs and behaviours of children and young people. Middlesex University, NSPCC, & Office of the Children’s Commissioner.Copy . And among teens who have seen porn, more than 79% of teens use it to learn how to have sexRobb, M.B., & Mann, S. (2023). Teens and pornography. San Francisco, CA: Common Sense.Copy . That means millions of young people are getting sex ed from violent, degrading content, which becomes their baseline understanding of intimacy. Out of the most popular porn, 33%-88% of videos contain physical aggression and nonconsensual violence-related themesFritz, N., Malic, V., Paul, B., & Zhou, Y. (2020). A descriptive analysis of the types, targets, and relative frequency of aggression in mainstream pornography. Archives of Sexual Behavior, 49(8), 3041-3053. doi:10.1007/s10508-020-01773-0Copy Bridges et al., 2010, “Aggression and Sexual Behavior in Best-Selling Pornography Videos: A Content Analysis,” Violence Against Women.Copy .
From increasing rates of loneliness, depression, and self-doubt, to distorted views of sex, reduced relationship satisfaction, and riskier sexual behavior among teens, porn is impacting individuals, relationships, and society worldwideFight the New Drug. (2024, May). Get the Facts (Series of web articles). Fight the New Drug.Copy .
This is why Fight the New Drug exists—but we can’t do it without you.
Your donation directly fuels the creation of new educational resources, including our awareness-raising videos, podcasts, research-driven articles, engaging school presentations, and digital tools that reach youth where they are: online and in school. It equips individuals, parents, educators, and youth with trustworthy resources to start the conversation.
Will you join us? We’re grateful for whatever you can give—but a recurring donation makes the biggest difference. Every dollar directly supports our vital work, and every individual we reach decreases sexual exploitation. Let’s fight for real love:
