Thorn releases 2024 research showing AI-generated child sexual abuse material now detected on major platforms

Thorn, a nonprofit technology organisation focused on child safety, has published its 2024 trends report documenting the emergence of AI-generated and AI-manipulated child sexual abuse material on mainstream platforms. The report, based on data from the National Center for Missing and Exploited Children and interviews with law enforcement and trust and safety professionals, found that synthetic and deepfake CSAM is being generated at scale using commercially available image models, often depicting real children whose likenesses have been extracted from legitimate social media profiles. Thorn warned that existing detection tools, which rely on hash matching of known material, are largely ineffective against novel AI-generated content.

Hash-based detection fails when every image is unique. Guardii's anti-CSAM module is trained to recognise the visual and contextual signatures of child sexual abuse material, including AI-generated and deepfake imagery, at the point of contact rather than after publication. The system operates in real time within direct messages on Instagram, Snapchat, Discord, Roblox and other platforms, intercepting the distribution of synthetic abuse material before it reaches the intended recipient and alerting the appropriate authority. Detection operates on pattern and content classification rather than fingerprint matching, addressing the operational gap created by generative models that produce novel material at scale.