New Research Reveals AI-Generated CSAM Surged 300 Percent in 2024, Detection Tools Lag

The Internet Watch Foundation published research showing a 300 percent increase in reports of AI-generated child sexual abuse imagery in 2024 compared to the previous year, with analysts identifying more than 100,000 unique AI-created images in the twelve-month period. The report found that commercially available generative AI tools are being manipulated to produce realistic CSAM despite built-in safeguards, with offenders sharing techniques to bypass filters on forums and encrypted channels. The IWF noted that traditional hash-matching systems used to detect known abuse material are ineffective against novel AI-generated images, and called for development of detection technology capable of identifying synthetic CSAM at scale.

Synthetic abuse material evades hash databases because every image is unique, requiring detection systems that interpret content rather than match signatures. Guardii's anti-CSAM module, built to address both photographic and AI-generated material, applies pattern-recognition AI to identify sexual depictions of children in direct-message exchanges in real time, flagging newly created deepfake and synthetic imagery as it is shared across Instagram, Snapchat, Discord and other platforms before it can be traded or used for further exploitation. As a world-leading online-safety platform, Guardii treats AI-generated and photographic CSAM as functionally identical threats, applying the same point-of-transmission interception to both and escalating confirmed detections to law enforcement in the format child-protection agencies require.