Home is where the hurt is, not the heart.
In the space of a single year, reports of AI-generated child sexual abuse material increased thirteenfold, a number that reflects not merely a technological failure but a civilizational one — the tools built to connect children to the world have been turned against them, in the rooms where they are meant to be safest. Researchers at the University of Edinburgh's Childlight institute have mapped a crisis of staggering breadth: across Western Europe, one in five children encounters online grooming before adulthood, and approximately five million are raped or sexually assaulted before they turn eighteen. The perpetrators are often not strangers but trusted figures, and the silence of victims — shaped by shame, fear, and misplaced self-blame — has allowed the harm to compound. What the evidence now demands is not only grief but accountability: from technology companies, from regulators, and from a society willing to see what is unfolding behind closed doors.
- AI-generated abuse images — real children's faces grafted onto synthetic sexual content — surged from 4,700 reported cases in 2023 to over 67,000 in 2024, a 1,300% rise in a single year.
- One in five children across Western Europe reports experiencing online grooming before age 18, while around five million are raped or sexually assaulted before adulthood — often by family members who exploit positions of trust both in person and through screens.
- Survivors like Rhiannon-Faye McDonald, groomed at thirteen and abused after an online predator appeared at her door, describe lasting psychological harm — the fear of being recognized, the weight of misplaced self-blame — that persists long after the abuse ends.
- Researchers identify the core failure as structural: mobile devices and platforms lack mandatory safety features, and technology companies have faced no meaningful regulatory accountability for the harm their products enable.
- Childlight and University of Edinburgh researchers are calling for enforceable safety standards built into devices, corporate accountability, and sustained investment in prevention — framing this not as an inevitable tragedy but as a preventable one.
The figures arrived in a research report, but they landed like a blow. In 2023, the National Center for Missing and Exploited Children logged 4,700 reports of AI-generated child sexual abuse material — images created by placing real children's faces into algorithms. By 2024, that number had surpassed 67,000. A 1,300% increase in a single year, spreading through the same devices meant to connect children to the world.
Childlight, a research institute at the University of Edinburgh, has been charting the full scale of this crisis. Their Into The Light study found that roughly one in five children across Western Europe — around 15 million young people — report experiencing online grooming before age 18. Approximately five million are raped or sexually assaulted before adulthood, with girls affected at more than twice the rate of boys. Many perpetrators are not strangers but family members, people who exploit closeness both in person and through screens.
Rhiannon-Faye McDonald was thirteen when a predator approached her online, posing as a peer. The grooming felt ordinary at first. Then the person appeared at her home. Today she campaigns for online safety, speaking from lived experience. 'For most victims and survivors, even with the right support, the impacts are significant and long-lasting,' she said. 'We live with misplaced self-blame and the fear of being recognised by those who have seen the images or videos of our abuse.'
Researchers point to a structural failure at the heart of this crisis: technology companies have not built adequate protections into the devices children use daily. Regulators have not demanded accountability. Paul Stanfield, CEO of Childlight, described the result as a preventable catastrophe — one hidden not because it is rare, but because victims stay silent, afraid of abusers, afraid of fracturing their families, convinced somehow that what happened was their fault.
Debi Fry, who leads the research at Edinburgh, insists that silence is itself part of the problem — and that the harm, though difficult to measure, registers in bodies and minds, affecting mental health and life expectancy for years. The path forward, she argues, requires investment in prevention: in education, in design, in enforcement, and in a willingness to look directly at what is happening in the homes and on the devices of millions of children.
The numbers arrived quietly in a research report, but they landed like a blow. In 2023, the National Center for Missing and Exploited Children logged 4,700 reports of AI-generated child sexual abuse material—images created by feeding real children's faces into algorithms, synthetic abuse with actual victims. By 2024, that figure had climbed to over 67,000. A 1,300% increase in a single year. The material was spreading through the same devices meant to connect children to the world, turning bedrooms into what researchers now call digital crime scenes.
Childlight, a research institute based at the University of Edinburgh, has been mapping the scale of this crisis. Their Into The Light study paints a landscape of abuse so widespread it strains comprehension. Across Western Europe, roughly one in five children—about 15 million young people—report experiencing online grooming before they turn 18. That means unwanted sexual advances, pressure, manipulation conducted through screens in the supposed safety of home. And the abuse doesn't stop at the digital boundary. Around five million children across the region are raped or sexually assaulted before adulthood, with girls affected at more than twice the rate of boys. Many of these perpetrators are not strangers. They are family members, people who occupy positions of trust, who exploit that closeness both in person and online.
Rhiannon-Faye McDonald was thirteen when someone approached her online, posing as a teenager. The conversation felt normal at first—the careful grooming that predators use to lower a child's guard. Then the person showed up at her home. What followed was abuse. Today, McDonald campaigns for better online safety, speaking from the weight of lived experience. She has watched technology companies prioritize profit over protection, watched the number of victimized children climb as a result. "For most victims and survivors, even with the right support, the impacts are significant and long-lasting," she said. "We live with misplaced self-blame and the fear of being recognised by those who have seen the images or videos of our abuse." The trauma doesn't end when the abuse stops. It persists in the body, in the mind, in the constant low-level terror of being identified.
The researchers at Childlight have identified the mechanism of this failure. Technology companies have not built adequate safety features into the devices children use daily. Mobile phones, designed to be personal and intimate, have become vectors for harm. The platforms and devices lack the protections that might slow or stop the creation and distribution of abuse material. Regulators have not held companies accountable. The result is what Paul Stanfield, CEO of Childlight, describes as a preventable catastrophe. "Home is where the heart is," he said, "but sadly for too many children, home is where the hurt is." Behind closed doors, in bedrooms where children should feel safest, an avalanche of abuse accumulates. It is hidden not because it is rare, but because victims stay silent—afraid of the abuser, afraid of breaking up their family, convinced somehow that what happened was their fault.
Debi Fry, who leads the research at the University of Edinburgh, emphasizes that silence is part of the problem. Many cases of child sexual abuse are never identified because children do not come forward. They may not even recognize what happened as abuse. The trauma, though, registers in the body. It affects mental health, physical health, even life expectancy. The evidence is limited to a handful of countries, and prevalence estimates vary, but what researchers know is clear: this violence is hardest to measure and hardest to confront. Yet it is also, Fry insists, preventable. It requires investment in stopping abuse before it starts—in education, in design, in enforcement. It requires technology companies to choose safety over convenience, regulators to demand accountability, and a society willing to look directly at what is happening in the homes and devices of millions of children.
Notable Quotes
For too long technology companies have favoured profit over safety. A rising number of children being abused is a direct result.— Rhiannon-Faye McDonald, abuse survivor and online safety campaigner
It's a hidden emergency in places where children should be safest, an avalanche of abuse behind closed doors—but it's preventable, not inevitable.— Paul Stanfield, CEO of Childlight
The Hearth Conversation Another angle on the story
Why does the number jump so dramatically from 2023 to 2024? Did the abuse suddenly get worse, or did reporting improve?
Both, probably. The technology got easier to use—AI tools became more accessible. But also, more people started reporting it. Still, researchers think the real surge is in creation and distribution. The tools are cheaper, faster, harder to detect.
When you say one in five children experience grooming, does that mean they all know it's happening to them?
Not necessarily. Grooming is designed to be subtle. A child might not recognize pressure as pressure. They might think an adult is just being friendly. The manipulation is the whole point—making the child unsure of what's normal.
The perpetrators are often family members. How does that change the dynamic?
It changes everything. A child can report a stranger. But reporting a parent or uncle means risking the family itself. They blame themselves. They fear the consequences of speaking up more than they fear the abuse continuing.
What does Childlight mean by "digital crime scenes in bedrooms"?
They mean the devices in those rooms are collecting evidence of abuse—images, videos, messages. The bedroom is supposed to be a refuge. Instead it's become a place where abuse happens and gets documented, stored, shared.
If this is preventable, what would prevention actually look like?
Better design in phones and platforms—features that detect and block abuse material before it spreads. Companies choosing safety over profit. Regulators with teeth to enforce it. Education so children recognize grooming. And adults who listen when children speak up.