Deepfake Sexual Abuse Survivors Face Unique Victim-Blaming

Deepfake sexual abuse is exactly what it sounds like: intimate imagery created or altered with AI, depicting a real person who never agreed to it. The technology first surfaced in celebrity fandom corners of the internet, but it has since spread to schoolyards and group chats, and its targets are overwhelmingly women and girls. Now a new study from researchers in Ireland says the damage runs much deeper than embarrassment, and names the victim-blaming myths that keep making it worse. The research, led by teams at University College Cork and the University of Limerick and reported by phys.org in October, is the first to identify victim-blaming patterns that are specific to deepfake abuse in accounts from celebrity survivors.

The myths surrounding deepfake sexual abuse are built around the technology itself. Victims were told the images cannot really be harmful because they are "fake." Celebrities were told they should simply "expect" this kind of content as a price of public life, an assumption the researchers say wrongly treats fame as a waiver of sexual privacy. Some victims described being pushed into a form of online sex work without their consent, as strangers created, distributed, and in some cases profited from sexual content generated from their likeness. According to the researchers, these myths compound the harm by minimizing what victims went through and by feeding the online harassment that follows.

Speaking out comes at a cost

Victims face a brutal choice when deciding whether to go public about the abuse. Some stay silent because they fear permanently attaching their name to sexually explicit deepfake content, and because publicity can drive more people to search for, view, or create abusive imagery of them. Others choose to speak out to challenge the abuse and protect other people, and then find themselves targeted with still more deepfake content. That creates a feedback loop at the center of deepfake sexual abuse: the people trying to stop it become its next targets. Dr. John Twomey, who led the research as part of his PhD studies at University College Cork and is now an assistant professor of psychology at the University of Limerick, said, "Speaking out about deepfake sexual abuse can come at a real cost. Our study found that celebrities who publicly challenged this abuse or became advocates against it could find themselves targeted with further deepfake content. This creates a deeply troubling situation where victims who speak out to challenge the abuse and protect others may risk exposing themselves to further harm."

Getting deepfake sexual abuse images taken down is its own ordeal. Victims described reporting and removal processes that demand substantial time and effort, including finding and documenting the abusive material themselves, often with limited results, according to phys.org's reporting on the study. Some turned to for-profit services to help identify and remove the content, while removal offered little protection against new images being created and distributed in their place. Identifying the perpetrators was another major obstacle, with victims often unable to determine who created the deepfakes at all. Gillian Murphy, professor of cognitive psychology at University College Cork and the study's supervisor, said, "This research shows that the harm of deepfake sexual abuse goes far beyond the creation of a single image. Victims can face repeated abuse, online harassment, victim-blaming and reputational harm, while also struggling to get abusive content removed or hold those responsible accountable."

Not just a celebrity problem

The celebrity cases are only the visible edge of deepfake sexual abuse. Sexualized deepfake imagery first emerged in twenty seventeen on Reddit, when a user uploaded fabricated images made by transposing female celebrities' faces onto the bodies of pornography actors, according to an analysis of deepfake abuse from Monash University. Since then the tools have gone mainstream: free and paid "nudify" apps can now generate a sexualized deepfake from any ordinary photo by digitally erasing the person's clothing. The Monash analysis describes substantial increases in reports of sexualized deepfake abuse around the world, including incidents involving schoolgirls and female teachers, plus a case in Sydney where police investigated deepfakes built from the faces of female students that had been circulated online. The same analysis reports that researchers interviewed ten perpetrators and fifteen victims in Australia, and none of the perpetrators interviewed had faced legal consequences for what they did.

What researchers want to change

The Irish team calls for responses that put less of the burden on victims, tackle the continued creation and spread of this material, and challenge the idea that deepfake sexual abuse is somehow less harmful simply because the images are fabricated, Murphy said. For Gen Z readers, the practical takeaway is blunt: if you encounter a deepfake of someone, do not share it, do not save it, and report it where you can. The idea that "fake" means harmless is the exact myth the researchers say keeps the cycle running. This pattern of blaming the victim rather than the system also shows up elsewhere, as GenZNewZ's reporting on scam victims argues, and the difficulty of reining in harmful AI tools is a familiar story in the debate over AI regulation. Consent does not stop applying just because software did the violating. Read the full study coverage at phys.org and the broader analysis of perpetrators at Monash Lens.