Acceptance of generative artificial intelligence among adults: A cross-sectional study of UTAUT-based dimensions and group differences
Generative artificial intelligence (AI) is rapidly transforming important areas of human activity, yet adults’ acceptance of generative AI applications remains insufficiently understood. This cross-sectional study examined the acceptance of generative AI applications among 460 adults recruited through online convenience sampling. Data were collected using the generative artificial intelligence acceptance scale (GAIAS), a validated instrument based on dimensions derived from the unified theory of acceptance and use of technology (UTAUT). The scale comprised four dimensions: performance expectancy, effort expectancy, facilitating conditions, and social influence. Participants reported moderate to high overall acceptance of generative AI. Effort expectancy, reflecting perceived ease of use (PEOU), received the highest mean score, followed by performance expectancy, reflecting perceived usefulness (PU), whereas social influence received the lowest mean score. This descriptive pattern indicates that perceived usability and usefulness were rated more positively than perceived social influence within the present sample. After Holm-Bonferroni correction across the 25 primary comparisons, statistically significant group differences remained for age and educational level in overall generative AI acceptance, performance expectancy, and effort expectancy. No multiplicity-adjusted differences were identified according to gender, personal computer ownership, or previous computer training. The age- and education-related findings should not be interpreted as evidence of linear relationships, independent effects, or causation. Within this digitally connected convenience sample, the findings indicate that PEOU and PU were more prominent dimensions of generative AI acceptance than social influence. Age- and education-related group differences were also observed. However, the cross-sectional design, online convenience sampling, and marked subgroup imbalances limit causal interpretation and generalizability to the wider adult population. The findings may nevertheless inform the development of accessible generative AI tools and targeted digital-literacy initiatives.