"There is nothing fake about the harm": The lived experiences and impacts of sexualized deepfake abuse

BackgroundSexualized deepfake abuse involves the non-consensual creation, distribution, or threat to create or distribute sexualized imagery that has been digitally altered or entirely fabricated, often with assistance from generative artificial intelligence (AI). Reports of this abuse are increasing globally, as AI tools become more accessible. Although sexualized deepfake abuse has been linked to image-based sexual abuse (IBSA) and technology-facilitated sexual violence (TFSV), limited research has examined whether it produces distinct harms or explored the perspectives of both victim-survivors and perpetrators.ObjectivesTo examine the lived experiences of victim-survivors and perpetrators of sexualized deepfake abuse, including its mental health and wellbeing impacts, and assess whether it produces harms that differ from those documented in the IBSA literature.DesignQualitative interview study.MethodsSemi-structured interviews were conducted with 25 Australian adults who had either experienced (n = 15) or perpetrated (n = 10) sexualized deepfake abuse. Interviews were undertaken via Zoom using a trauma-informed approach and were analysed thematically.ResultsSexualized deepfake abuse was associated with substantial psychological, social and occupational harms. Participants described impacts consistent with those reported in IBSA research, including distress, reputational harm, relationship difficulties and reduced wellbeing. However, the findings also identified harms linked to the synthetic nature of deepfakes. Victim-survivors reported anxiety about how others might perceive the authenticity of fabricated images and the burden of disproving false sexual representations. Perpetrator accounts highlighted how the accessibility of deepfake technologies and perceptions of anonymity and reduced accountability lowered barriers to abuse.ConclusionSexualized deepfake abuse produces significant harms that extend beyond those commonly associated with IBSA. The fabricated nature of deepfake content creates unique challenges relating to authenticity, credibility and control. Prevention and intervention efforts should address these distinctive features through platform-level safeguards, technological detection measures and strategies targeting the tools and infrastructures that enable abuse.