Anatomy of AI Risk Perception: Differentiating Privacy, Misinformation, and Technical Failures

This article examines the perception of risks associated with the use of artificial intelligence in business among the general population, with a focus on privacy breaches, misinformation, and technical failures. The aim of this study is to understand how sociodemographic and educational factors influence perceptions of these risks. Initially, respondents were presented with eight hypothetical scenarios, grouped into three risk categories: privacy, misinformation, and safety. Subjective perception of financial harm was measured as a percentage of the company’s total turnover, and the analysis was conducted using regression techniques. The results showed that respondents perceive AI risks as economically harmful. Gender consistently and significantly influenced perceptions, with men perceiving considerably lower risk than women. Higher education was associated with a lower risk perception, and other factors, such as knowledge and use of AI, generally had no significant impact. However, AI users perceived a lower risk of misinformation compared to non-users.