Cross-cultural determinants of AI acceptance in medical education among medical students in China and Pakistan

BACKGROUND: Although artificial intelligence (AI) is increasingly introduced into medical education, little is known about the psychological mechanisms through which social influences shape AI acceptance across different cultural contexts. This study addresses this gap by examining whether perceived usefulness and attitude mediate the relationship between subjective norms and AI acceptance, sequentially, among medical students in China and Pakistan. METHODS: Between March and April 2025, a cross-sectional survey was administered to medical students enrolled in accredited universities across two culturally distinct regions. A stratified sampling approach was employed to ensure representativeness across academic years. Data from 1,232 students (Pakistani, n = 551, and Chinese, n = 681) were analyzed using structural equation modeling (SEM) and multi-group analysis. The hypothesized model assessed the direct and indirect effects of subjective norms on AI acceptance, with perceived usefulness and attitude as mediators. Measurement invariance was tested to validate cross-cultural comparisons. RESULTS: In the Chinese sample, subjective norms had a significant direct effect on AI acceptance (β = 0.231, p < 0.01), while this effect was not significant in the Pakistani sample (β = 0.113, p = 0.125). In both samples, subjective norms significantly influenced perceived usefulness and attitude, which, in turn, mediated the relationship between these variables and AI acceptance. The sequential mediation path was significant in both samples. Multi-group analysis confirmed full measurement and structural invariance, supporting the model's cross-cultural validity. CONCLUSION: The findings underscore the importance of culturally responsive strategies when promoting AI acceptance in medical education. While the sequential mediation of perceived usefulness and attitude appears robust across contexts, the influence of social norms on behavioral intention to use AI varies across cultural and institutional environments, underscoring the need for tailored engagement and implementation approaches.