A peer-reviewed article published in AI & Society on April 6, 2026 examines the profound challenges that generative artificial intelligence tools capable of producing synthetic child sexual abuse material present for child protection, law enforcement, and societal responses to exploitation. The research challenges arguments that AI-generated CSAM differs fundamentally from traditional CSAM due to a perceived absence of direct victimization, documenting that AI has been implicated in the creation of synthetic CSAM of children who have not previously been abused, the revictimization of known survivors through manipulated imagery, the facilitation of grooming, coercion, and sexual extortion, and the normalization of child sexual exploitation.
The research confirms what detection practitioners observe in live operations: synthetic CSAM functions as an active tool of abuse, not merely a representation of it. Offenders deploy AI-generated or deepfake imagery in grooming sequences to desensitise children to sexual content, in sextortion schemes to fabricate kompromat and coerce compliance, and in revictimization of known survivors whose likenesses are manipulated into new abuse scenarios. Detection technology that identifies both traditional and AI-generated CSAM within direct messages in real time—such as Guardii's anti-CSAM and anti-deepfake modules, which the company applies across Instagram, Snapchat, Discord and other platforms—intercepts these materials before they are used to manipulate a child, blocking the sender and alerting safeguarding authorities, preventing the psychological harm and further coercion that follows exposure.
