Canadian Police Report AI-Generated CSAM Now Used in One-Third of Child Exploitation Cases, Increasingly Weaponised in Sextortion
Guardii Analysis
The RCMP in Prince Edward Island reported on 15 November 2026 that AI-generated or manipulated material now appears in approximately one-third of online child abuse investigations on the island, a finding echoed across Atlantic Canada. Jacques Marcoux, director of research for the Canadian Centre for Child Protection, which operates the national Cybertip.ca reporting line, stated that offenders are increasingly using AI-generated child sexual abuse material in sextortion attempts, creating fabricated explicit images of minors when they cannot trick a child into providing real ones, then threatening to release the images online and claiming they are authentic. Summerside police confirmed at least one local case this year in which a suspect deployed AI-generated CSAM as part of a sextortion scheme, illustrating the operational shift from coercion to outright fabrication as generative tools become accessible.
The weaponisation of generative AI to fabricate evidence of abuse that never occurred creates a coercion vector that traditional image-hashing detection cannot address. Guardii's anti-CSAM module, which includes detection of AI-generated and deepfake material, identifies the distribution, sharing and coercive use of such content in direct-message conversations, flagging the threat before a child is confronted with a fabricated image or believes the offender's claim that it will be published. The world-leading platform, a Meta Business Partner, applies pattern recognition to the shape and escalation of the conversation rather than relying solely on matching known images, enabling it to surface novel AI-generated material being used to extort a teenager in real time. The approach addresses the reality Canadian police and child-protection researchers described: offenders no longer need a real image to coerce compliance, so detection must focus on the coercion itself, not the provenance of the file.
