Examining the association between AI-enhanced education and medical students' self-directed learning using an integrated TAM-UTAUT2 model.

Background: Artificial intelligence (AI)-assisted education has become an increasingly important instructional model in medical education and raises questions about how it shapes students' self-directed learning processes and technology adoption. Objectives and methods: This study examined factors associated with self-directed learning and AI-related behavioral intention among medical students using a cross-sectional survey design. A model integrating self-directed learning dimensions with TAM- and UTAUT2-related constructs was tested with 600 valid questionnaires, and data were analyzed using partial least squares structural equation modeling (PLS-SEM) in SmartPLS 4. Results: The results supported 21 of the 24 hypothesized paths. Motivation emerged as the strongest predictor in the model and was significantly associated with attitude and all three self-directed learning dimensions. Attitude was also significantly associated with self-planning, self-management, and self-monitoring. Self-planning was positively associated with self-management, and self-management was positively associated with self-monitoring. In the technology acceptance pathway, perceived ease of use and perceived usefulness were associated with behavioral intention, and behavioral intention was associated with actual behavior. Facilitating conditions and social influence were also associated with behavioral intention. The model explained substantial variance across key constructs, ranging from 47.6% in actual behavior to 69.9% in self-management. Conclusion: These findings suggest that motivational support, structured self-directed planning activities, and adequate digital infrastructure may be relevant considerations for AI integration in health sciences education. The study provides preliminary evidence that a model integrating self-directed learning dimensions with TAM and UTAUT2 related constructs may help explain AI-assisted learning behavior in this population and highlights the need for longitudinal research to clarify the directionality of these associations.