Celebrities who speak out about deepfake sexual abuse can become targets of new deepfakes

Every finding in this paper comes from things celebrities chose to say in public. Researchers at University College Cork (UCC) and the University of Limerick (UL) analysed public statements from celebrity victims of deepfake sexual abuse ^[1][3]. The abuse is intimate imagery made or altered with AI without the consent of the person shown ^[15]. The statements described significant psychological and emotional distress, along with social and reputational harm ^[3]. The press account does not give the number of celebrities included or how often each theme appeared.

Two excuses recur. One says the images do no harm because they are "fake"; the other says public figures should "expect" this kind of content ^[4]. The first exists only because the images are generated, and the second only because the victims are famous. According to the study, both make the harm worse by minimising what victims went through and by feeding online harassment ^[6]. Some victims described the experience as being placed into online sex work without their consent, while others made, shared and in some cases profited from sexual content built on their likeness ^[5].

The retaliation finding has the most direct bearing on policy. "Speaking out about deepfake sexual abuse can come at a real cost," said John Twomey, who led the research during his PhD at UCC and is now an assistant professor of psychology at UL ^[9]. "Our study found that celebrities who publicly challenged this abuse or became advocates against it could find themselves targeted with further deepfake content," he said ^[10]. Others stay quiet. They fear their name will be tied permanently to explicit content, or that publicity will send more people to search for, view or create the images ^[7].

I'd hold that finding with some care. The link between going public and being targeted again comes from victims' own accounts, and a sample of public statements is drawn from people who went public in some form ^[3]. Those accounts are good evidence that the pattern recurs. Measuring how much advocacy raises the risk would need a comparison with victims who stayed silent, and silent victims leave few public statements to analyse.

The private route has its own costs. Victims described reporting and removal processes that took substantial time and effort, including finding and documenting the material themselves, often with limited results ^[11]. Some turned to for-profit services to identify and remove content, and removal gave little protection against new images being made and distributed ^[12]. Victims often could not determine who created the deepfakes and had limited success holding anyone to account ^[13].

So the study describes two paths open to a victim. Public advocacy can bring fresh deepfakes ^[8]. Reporting asks the victim to find and document the material ^[11]. A response built on either path waits for the person already harmed to act first. "We need responses that place 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," said Gillian Murphy, professor of cognitive psychology at UCC and the study's supervisor ^[14].

The design is descriptive. It records what victims say happened, and judging which platform or legal remedy would lift the burden needs a study built to compare remedies. The subjects are also public figures, and one of the two myths depends on fame ^[4]. Private victims were outside the sample, which was drawn from celebrities ^[3].