Survey finds 7% of adults across three countries reported AI intimate-image abuse

For people targeted with non-consensual AI intimate imagery, the consequences extend beyond the images: nearly 87% of affected survey respondents reported impacts on their lives. An RMIT University research account released November 3 describes a three-country study in which 7% of adults surveyed reported experiencing this abuse since turning 18.

The research moves beyond high-profile cases involving celebrities and public figures. Led by RMIT professor Nicola Henry, the team surveyed 7,230 adults in Australia, the United Kingdom and the United States in July and August 2024. The peer-reviewed paper was published online September 8, 2026, in the International Journal of Human-Computer Interaction.

Creation, sharing and threats

The survey counted three forms of abuse, rather than only images posted publicly. Respondents could report that someone had created or altered an intimate image using AI without consent, shared an AI-generated intimate image without consent, or threatened to share one. That definition includes abuse through threats even when no distribution is reported.

• Creation: 4.3% said someone had made AI-generated intimate images of them without consent.
• Sharing: 4.4% said AI-generated intimate images of them had been shared without consent.
• Threats: 3.6% said someone had threatened to share such images.

The broader measure was higher: 15.1% reported abuse involving digitally altered intimate images, including AI-generated fakes and images made with other editing methods. The distinction matters when reading the results: the 7% finding concerns AI specifically, not every form of manipulated intimate imagery.

Henry cautioned that some people may never discover that images of them have been created or shared, meaning the true scale could be greater. The study also does not answer how quickly victimization is increasing as generative AI becomes more widely available. Its responses describe experiences since adulthood, collected in 2024—not an annual abuse rate for 2026.

9.3%United Kingdom

UK respondents reported the highest rate of AI-generated intimate-image abuse among the three countries.

7.7%Australia

Australian respondents reported a 7.7% rate.

6.5%United States

US respondents reported a 6.5% rate.

Exposure and harm follow different patterns

Men reported experiencing AI-generated image-based abuse at more than twice the rate of women: 10.2%, compared with 3.8%. But women described more negative reactions and impacts. Among affected respondents, 42.4% of women reported fear for their safety, compared with 18.1% of men.

Women were also more likely to report mental-health impacts. Men were more likely to describe neutral or positive reactions, including finding the images funny, although those responses were uncommon overall. Across affected respondents, 90.3% reported at least one negative feeling. Reported consequences extended to health and wellbeing, relationships, work or finances.

Other differences cut across the sample. Respondents under 35, LGBTQ+ respondents, people of color and people with disabilities reported higher victimization. Together, the findings distinguish the likelihood of reporting abuse from its reported consequences: the group reporting more incidents was not necessarily the group reporting more severe reactions.

Preventing abuse before a complaint

Henry argued that technology companies should do more to prevent abuse before it occurs, rather than focus only on responding afterward. She pointed to Australia’s proposed digital duty of care as an opportunity to require online services to prevent foreseeable harm. She also called for education, accessible support for survivors, and safety built into AI tools and apps from the outset.

Google funded the research program, and co-author Rebecca Umbach is a staff user-experience researcher at the company. Umbach said understanding who experiences the abuse, how it happens and its effects is essential to designing better prevention and responses. The international team is continuing its research into understanding and preventing this abuse.