IWF finds more AI child abuse images in six months than in all of 2025: every school needs a plan for fakes

On 5 October 2026 the Internet Watch Foundation (IWF), the UK body that finds and removes child sexual abuse material online, reported that its analysts assessed 6,310 AI-generated child sexual abuse images in the first six months of 2026. That is more than the 4,512 identified in the whole of 2025, a rise of 40%. Of the images with a recorded gender, 98% showed girls; 79% depicted children aged 7 to 13, up from 70% in 2025, and 190 showed children under two. The IWF published the figures alongside a call for EU policymakers to agree the long-delayed Child Sexual Abuse Regulation. Its chief executive, Kerry Smith, said: "It is deeply concerning to see how quickly AI-generated child sexual abuse imagery continues to grow and spread online."

The figures that matter most to schools are the ones about children reporting themselves. The Irish Examiner and other outlets reported IWF figures showing 420 reports in the first half of 2026 to Report Remove, the service run with Childline that lets under-18s report sexual imagery of themselves, in which the child said the image was faked or manipulated. That is already above the 397 such reports in the whole of 2025. The IWF has also published guidance for parents on AI-manipulated images made from ordinary photographs of children. Its head of policy, Hannah Swirsky, said: "It is incumbent on tech companies to build tools which cannot be abused this way."

Dan Fitzpatrick, The AI Educator, says the story is not a reason to keep AI out of school. It is a reason to be honest with children about it. Safeguarding leads have handled incidents involving nude images for years; what has changed is that the child need not have taken any picture at all. A pupil who understands how image tools work, knows a fake can be made from a photograph, and knows what to do in the first hour after finding one is safer than a pupil who has only been told to be afraid.

Fitzpatrick also points to trust. Children now have to live with the knowledge that an image of them might exist that they never took, and adults have to respond without blame when a child says so. Teachers are already ahead of policy here: designated safeguarding leads are dealing with manipulated images before national guidance has caught up with the technology. There is an equity problem too. The schools best protected are those with trained staff and briefed parents, and the gap between them and everyone else is growing as fast as the numbers in this report.

For school leaders, the practical steps this term are concrete. Check that the safeguarding policy names AI-generated and manipulated imagery and says what staff do when a child reports one. Make sure every adult in the building, not just the safeguarding team, knows Report Remove exists and how a young person uses it. Brief parents in plain language about photographs of children shared online, without alarm and without judgement. Teach the technology itself, in computing and PSHE, so children learn how these images are made and why sharing one, even to show a friend, causes harm. For business leaders deploying image tools, ask the vendor what stops misuse and put a named person in charge of the answer.

Fitzpatrick works with schools, colleges, governing bodies and government organisations on adopting AI and building AI literacy with children's wellbeing at the centre, and writes regularly on what AI means for learning and work.

Dan Fitzpatrick is available for interview and comment on this story.