One in 15 adults has experienced AI-generated image-based abuse, study finds

New research reveals that one in every fifteen adults in the United States, United Kingdom, and Australia has experienced AI-generated image-based abuse, according to a study conducted by a team of international researchers. The comprehensive survey, which involved over seven thousand adults aged eighteen and above, has uncovered the prevalence of this emerging issue.

The study, published in the International Journal of Human-Computer Interaction, found that seven percent of the respondents had encountered at least one form of AI-generated image-based abuse, which included the creation or alteration of intimate images without their consent, the unauthorized sharing of such images, or the threat of image sharing. Almost nine out of ten of the victim-survivors reported significant negative impacts on their lives, affecting their health, well-being, relationships, work, and finances.

Professor Nicola Henry, the lead author from RMIT University in Australia, conducted the research with support from Google, and highlighted that the findings shed light on the extent of AI-generated image-based abuse affecting the general population, not just high-profile individuals. She emphasized that the true scale of the problem may be even greater due to the possibility that some people might remain unaware of the creation or sharing of AI-generated images of them.

Notably, the study found that higher rates of AI-generated abuse were reported by younger individuals, people from racial and ethnic minority backgrounds, men, and those with disabilities. While men reported experiencing the abuse at a rate of 10.2%, which is more than twice the rate of women at 3.8%, women reported more severe negative consequences, including fear for their safety and adverse effects on their mental and physical health.

This discrepancy in impacts may be attributed to broader gender inequalities, stigma, and safety risks that women frequently encounter.

The lead author, Nicola Henry, pointed out that the findings demonstrate that anyone can be affected by AI-generated image-based abuse, and that the risk of experiencing such abuse, as well as its consequences, is not evenly distributed across all demographics. The research also underscores the growing potential for technology companies to misuse generative AI tools for creating, sharing, or threatening to share intimate images without consent as the technology becomes more widely available.

While the study brings valuable insights into the issue, it also highlights the need for greater responsibility from technology companies to prevent AI-generated image-based abuse before it occurs. Australia's proposed digital duty of care, which aims to shift the focus from responding to abuse after it happens to requiring online services to prevent foreseeable harm, is seen as an opportunity to address this issue more effectively.

However, the researchers stress that regulation and new laws alone are insufficient. They also advocate for better education, accessible support for victim-survivors, and a greater sense of responsibility from technology companies and developers to incorporate safety measures into AI tools and applications from the outset.

The international team behind the study intends to continue their work to gain a deeper understanding of AI-generated image-based abuse and develop effective prevention strategies. The researchers are collaborating with Google researchers to identify where industry-wide prevention efforts are most needed and explore how technology itself can contribute to harm prevention.

The study marks a significant step towards understanding and mitigating this emerging form of abuse and its far-reaching impacts on individuals and society.