AI-generated nude images of children on the rise, police warn

AI-generated sexual images of children have risen 16-fold in the past three years.

Twenty police forces across England and Wales logged 163 crimes returned by searches for terms including “AI-generated”, “deepfake” and “nudify” by July this year.

In 2023, the number was 10, rising to 46 in 2024 and 112 in 2025.

Prof Jessica Ringrose, of UCL’s Institute of Education, said AI was becoming routine in young people’s lives , increasing the risk that the tools could be used to abuse others.

“People using AI to undress women – it’s really scary how much AI can do,” said Lily, 18, a London sixth-form student interviewed for Prof Ringrose’s research.

“Maybe you just post an innocent photo of yourself. You don’t know what other people can do with that image.”

The records do not say how old every suspect was, so they cannot show how much of the abuse was committed by children and how much by adults. But they do include reports of pupils targeting pupils, and allegations involving ex-partners, strangers and suspects demanding money.

In Derbyshire, a group of boys was accused of creating indecent AI images of female classmates. Police identified suspects but decided that further investigation was not in the public interest.

In West Yorkshire, a student was accused of making a deepfake video showing child sexual abuse. The case was closed due to evidentiary difficulties.

Leicestershire Police recorded a case in which a photograph of an 11-year-old girl whose top had been pulled down was shared around a school. A separate AI image of her was then made and circulated. Another pupil was accused of setting up a fake TikTok account carrying an image depicting child sex abuse.

A 17-year-old boy in Wales was told a deepfake nude image would be distributed unless he paid. Another parent reported a demand for £700 to stop an AI pornographic video of their son from being published.

Humberside Police recorded an unknown social-media user demanding £50 in gift cards after sending a victim an AI-generated explicit image. Wiltshire Police disclosed reports involving gift-card demands and one threat to circulate further images unless £300 was paid.

Some cases ended in a community resolution or youth restorative disposal. Others were dropped because no suspect could be identified or the evidence did not meet the threshold. Many remain under investigation.

Prof Clare McGlynn, of Durham University, said deepfake technology had been made “easily and freely accessible” through Google and mainstream social-media platforms. “It spreads amongst young people because we have not been clear enough that this is wrong,” she said.

Schools faced the task of encouraging victims to report abuse while talking openly with boys the harm it causes and the culture behind it, Prof McGlynn said.

‘Our worst fears are coming true’

Kerry Smith, chief executive of the Internet Watch Foundation, said the technology had made it easier than ever to create lifelike child sexual-abuse images, sometimes of real and recognisable children.

“Steadily over the years, we have seen all our worst fears the potential abuses of AI to create ever more realistic child sexual abuse imagery come true,” she said.

Ms Smith said the foundation had dealt with criminals attempting to blackmail a school with AI-manipulated sexual images of pupils made from ordinary photographs put online. She called for the new ban on nudify apps to be enforced and for AI companies to build safety into their products.

Lewis Keller, of the NSPCC, said the figures chimed with what Childline counsellors were hearing from young people worried that AI was being used to create explicit images of them. Victims had described feeling ashamed and frightened, he said, with some reporting a serious effect on their mental health.

“Platforms should be required by law to thoroughly test their products to ensure they are safe for children and cannot produce harmful or illegal content,” Mr Keller said. “Child safety cannot be an afterthought.”

The true scale is likely to be higher. Surrey Police refused the request because it said 2,271 child-image records would have to be examined manually. Other forces warned that keyword searches could miss relevant cases and also sweep up records in which the term was incidental.

An offence criminalising the creation of intimate deepfakes of adults without consent came into force in England and Wales on Feb 6, 2024. The Government said existing child-image laws already covered indecent deepfake images involving under-18s.

The Protection of Children Act 1978 makes it illegal to make, distribute or show indecent photographs of children. In 1994, Parliament extended the law to computer-generated “pseudo-photographs”, a definition which encompasses photorealistic AI-generated images.

Separately, the Online Safety Act 2023 created offences covering the non-consensual sharing of intimate images and threats to them, including deepfakes. Those offences came into force on Jan 31, 2024. The act also imposes duties on regulated online services to tackle illegal content, with the main illegal-content safety duties taking effect in March 2025.

The Crime and Policing Act 2026 also introduced offences targeting so-called “nudification” tools, criminalising the making, adapting or supplying of tools for use in generating purported intimate images. The provisions came into force on June 29.