Artificial intelligence safety and ethics perception scale

This study aimed to develop the Artificial Intelligence Safety and Ethics Perception Scale (AISEPS), a multidimensional instrument designed to assess individuals’ perceptions of artificial intelligence in terms of safety and ethics, and to examine its psychometric properties. The research was conducted as a mixed-methods scale development study with a sample of 558 university students. Based on international ethical frameworks and recent empirical literature, an initial pool of items was created, and content validity was established through expert evaluations. Construct validity was assessed using Exploratory Factor Analysis and Confirmatory Factor Analysis, while reliability was evaluated through Cronbach’s alpha coefficients, item–total correlations, and 27% upper–lower group comparisons. The findings confirmed a five-factor structure consisting of 40 items. The scale includes the dimensions of Safety Awareness (7 items, α = .942), Ethical Values (10 items, α = .953), Responsibility and Accountability (8 items, α = .952), Social Impact (8 items, α = .966), and Positive Expectations (7 items, α = .966). Each subscale demonstrated a unidimensional structure, with explained variance ranging from 82.7% to 93.7%. CFA results indicated acceptable to good model fit across all dimensions. Demographic analyses revealed no significant differences according to academic field, whereas significant differences were found with respect to age and educational level. Overall, the results indicate that the AISEPS is a valid and reliable multidimensional instrument for measuring individuals’ perceptions of artificial intelligence regarding safety and ethics.