The Internet Watch Foundation (IWF) has reported a 154% increase in reports of AI-generated child sexual abuse imagery (CSAM) in 2025 compared to the previous year. The charity received 491 reports containing realistic AI-generated CSAM in 2025, up from 193 in 2024.
In its annual report released on Thursday, the IWF highlighted new forms of AI CSAM, including content on AI companion sites and adverts on mainstream social media. Video content is a particular concern, with analysts reporting 3,443 AI-generated child sexual abuse videos in 2025, a 260-fold increase from just 13 in 2024.
The IWF noted that AI-generated imagery is more likely to be assessed as the most serious category (Category A) than non-AI imagery. Most AI-generated images assessed as criminal in the past two years have been Category C images. The foundation warned that such material can contribute to ongoing exploitation, cause lasting harm, and pressure safeguarding systems.
The charity called on tech companies to ensure safety is built into AI products, stating that while government plans to test AI models are welcome, there is no legal requirement for pre-deployment safety testing. A government spokesperson said the UK law is clear that creating, possessing, or distributing CSAM is illegal, including AI-generated content, and that platforms must proactively identify and remove it under the Online Safety Act.
