Internet Watch Foundation reports huge rise in AI child sexual abuse material
The Internet Watch Foundation (IWF) says it has assessed more AI-generated child sexual abuse images in the first half of 2026 than in all of 2025. The rise has renewed calls for binding rules requiring technology companies to make AI systems safer.
The IWF assessed 6,310 images that met the legal definition of child sexual abuse, about 40% more than its total of over 4,500 last year; the figures cover still images, not videos. Most depicted girls aged seven to 13. Report Remove has also received 420 reports from children about intimate images they believed were faked or manipulated, already above its 397 reports in 2025. UK law makes such material and certain related AI models illegal, while authorities advise families to limit public sharing of children’s photos.
- IWF has assessed 6,310 AI-generated abuse images this year.
- The count is already 40% above 2025’s total.
- Children have reported more manipulated images than last year.
New here? Start with this
AI-generated child sexual abuse material refers to images created by artificial intelligence that depict minors in sexual situations. This material is illegal in the UK regardless of whether real children were involved in creating it, because of the harm such images cause and how they can be used to groom or exploit children.
The Internet Watch Foundation, which monitors illegal content online, has found a significant increase in such images in recent months. It assessed more AI-generated child sexual abuse images in the first six months of 2026 than in all of 2025, indicating how rapidly the problem is growing. Additionally, there has been a rise in reports of real photographs of children being manipulated or faked to create intimate images, causing direct harm to the young people involved.
The rise has prompted technology companies and regulators to consider how to address the issue. Currently, there is no binding international requirement forcing companies to implement safeguards in their AI systems to prevent this misuse, and child protection experts worry that without such protections the problem will continue to worsen.
Both sides, in good faith
The strongest fair case each way — we don't pick a winner.
The case for
Those calling for binding rules argue that the explosion in AI-generated child sexual abuse material demonstrates a genuine crisis requiring urgent legal action. Companies have known about these risks for years yet chosen insufficient voluntary measures, making clear that regulatory obligation is necessary to drive consistent safeguarding across the industry. The comparison to regulated industries like pharmaceuticals or aviation is instructive: society accepts binding rules when children's safety is at stake, and AI systems capable of generating such material should be no exception.
The case against
Those sceptical of binding regulation argue that such rules risk unintended consequences if poorly designed, potentially stifling legitimate AI research and development whilst pushing innovation to less regulated jurisdictions where safeguards may be weaker. They contend that technology companies are already investing substantially in detection and prevention, and that what's needed is time for market-driven solutions to mature alongside targeted law enforcement and parental education—not blunt legislative instruments that could freeze beneficial AI progress. The real problem, they suggest, lies with criminal misuse of AI tools rather than the tools themselves.