Internet Watch Foundation reports huge rise in AI-generated CSAM in 2025
The charity said AI-generated child sexual abuse videos rose more than 260-fold on the prior year, with 65% of that video content in the most severe legal category.
- Security & misuse
- Notable
The Internet Watch Foundation, the UK charity that identifies and works to remove child sexual abuse material online, said it had assessed 8,029 AI-generated images and videos as depicting realistic child sexual abuse in 2025. Of those, 3,443 were videos — more than 260 times the 13 AI-generated videos it found in 2024 — and 65% of that video content fell into Category A, the most severe classification under UK law.
The IWF said the material was increasingly difficult to distinguish from real abuse imagery and was being found both on the dark web and on ordinary commercial platforms on the open internet. It set the AI figures against a record year overall: 312,030 total reports of child sexual abuse material, up 7% on 2024 and the highest total in the charity’s history. Chief executive Kerry Smith said generative tools meant “criminals essentially can have their own child sexual abuse machines to make whatever they want to see,” and the charity called for AI developers to build stronger safeguards against this kind of misuse into their products by design.
The report added AI-generated content to what the IWF described as an already deteriorating picture of online child sexual abuse, and came as the UK and other jurisdictions moved to tighten obligations on AI services: the IWF’s figures were published as part of a wider push for stricter regulation, and preceded new UK codes taking effect in March 2026 requiring AI services to limit children’s access to sexually explicit, violent and self-harm-related material.