IWF Records 260-Fold Surge in AI-Generated CSAM Videos as Quality and Extremity Escalate

The Internet Watch Foundation uncovered over 290,000 web pages containing child sexual abuse material in 2024, representing a five percent increase from the previous year and the highest total in the organisation's 29-year history. While AI-generated material still accounts for approximately one percent of all CSAM analysed by the IWF in 2025, the foundation identified 260 times as many AI-generated CSAM videos in 2025 compared to 2024, with frontline observers reporting that the quality is now alarmingly high and the categories of offence depicted are becoming more extreme as generative tools improve their ability to show multiple individuals in video. Ninety-seven percent of the victims depicted are girls.

Synthetic CSAM produced at industrial scale overwhelms traditional hash-matching infrastructure and enables offenders to circumvent detection by manufacturing novel imagery for each sextortion or grooming campaign. Detection infrastructure that identifies the conversation leading to the threat rather than cataloguing every variant of the image addresses the operational reality that AI generation outruns hash databases. The anti-sextortion and anti-CSAM filters within a world-leading AI online-safety platform analyse the escalation and coercion patterns characteristic of image-based sexual abuse, intervene before synthetic material is transmitted to or weaponised against a child, and surface the incident to the responsible adult or agency in real time, targeting the harmful behaviour at the source rather than attempting to catalogue an exponentially growing library of unique AI-generated files.