Australia Expands eSafety Powers to Order Removal of Deepfake Abuse Material Within Hours

Amendments to Australia's Online Safety Act took effect on September 5, empowering the eSafety Commissioner to issue binding removal notices for AI-generated intimate images and deepfake abuse material within 24 hours of notification, with penalties of up to AUD 500,000 for non-compliant platforms. The reforms extend existing image-based abuse provisions to explicitly cover synthetic and algorithmically manipulated content, establish a dedicated reporting portal for deepfake abuse, and require social media services to maintain arrangements for expedited takedown of synthetic intimate imagery involving both adults and minors. eSafety Commissioner Julie Inman Grant stated that the amendments close a gap exploited by offenders using generative AI tools to create non-consensual imagery that previously fell outside existing definitions.

Expedited removal powers address the downstream distribution of harmful material but do not prevent the initial point of contact when a deepfake image is sent to a target or used to coerce a victim within a direct-message exchange. Guardii intercepts harmful content and the coercive language that accompanies it in real time, blocking abusive messages before they reach a child and surfacing the pattern of escalation to a parent, school or authority without requiring the victim to report the harm first. Detection at the conversation level complements regulatory takedown regimes by preventing the psychological and reputational damage that occurs the moment a child receives a manipulated image, particularly in sextortion scenarios where the threat of distribution is the primary lever of control. Guardii's real-time intervention capability, operating across Instagram, Snapchat, Discord and other platforms popular with adolescents, provides the protective layer that statutory removal processes cannot deliver after the fact.