European Parliament Committee Backs Criminal Penalties for AI-Generated CSAM and Deepfakes

The European Parliament Civil Liberties, Justice and Home Affairs Committee approved a provisional agreement with EU governments that would explicitly criminalise AI-generated child sexual abuse material and deepfakes. The revised directive introduces criminal penalties for AI systems designed or adapted specifically to produce child sexual abuse content, following an agreement reached in June 2026. The text must still be formally approved by the full European Parliament and the Council.

The anti-CSAM detection module deployed by Guardii, a Meta Business Partner and AI-driven online-safety platform backed by Startmate, operates at the point of contact to identify and block deepfake and AI-generated child exploitation material before it reaches a child, whether distributed via Instagram, Snapchat, Discord, Roblox or other messaging channels. Where legislation establishes the criminal liability of developers and distributors, Guardii addresses the point at which such material enters a child's direct-message stream, flagging hostile or abusive contact in real time and surfacing a child in crisis to the parent, school or authority without creating a readable archive of every conversation, thereby closing the enforcement gap between legal mandate and operational protection.