The AI Stress and Anxiety Scale (AISAS): Development, Validation, and Insight on the Diffusion of AI-Related Stress and Anxiety

The rapid diffusion of generative artificial intelligence (AI) technologies has raised growing concerns about their psychological impact, particularly in relation to stress and anxiety. While research on AI anxiety is expanding, existing measures are limited by theory-driven item construction, narrow samples, and the failure to distinguish anxiety from stress. The present research introduces the Artificial Intelligence Stress and Anxiety Scale (AISAS), a psychometrically validated instrument designed to assess both AI-related anxiety and AI-related stress in the general population. Scale development followed best-practice guidelines and comprised two preregistered studies conducted on representative samples of adults from the United Kingdom. An initial pool of 62 items was generated through an extensive literature review and refined using expert ratings to compute a Content Validity Index. Exploratory factor analysis in Study 1 (N = 301) supported a robust four-factor structure capturing job-related concerns, AI-related stress, privacy concerns, and existential/consciousness-related concerns, explaining 72% of the variance. The scale demonstrated excellent internal consistency (α = .92), full measurement invariance across gender, and strong convergent, divergent, and predictive validity with established measures of anxiety, stress, technology readiness, AI literacy, and AI usage behavior. Criterion validity analyses showed that specific AISAS dimensions differentially predicted AI usage frequency and behavioral intention. Study 2 (N = 324) provided additional evidence for the stability of the AISAS factor structure across independent samples, supporting metric invariance. Descriptive analyses show that concerns regarding privacy implications of AI are widespread, while AI stress is infrequent. Overall, the AISAS offers a reliable, multidimensional, and methodologically robust tool for assessing AI-related stress and anxiety.