Factors associated with generative AI chatbot acceptance among medical students with prior academic use: a cross-sectional study based on an extended technology acceptance model

Generative artificial intelligence (GenAI) chatbots are becoming increasingly embedded in medical education, providing students with on-demand support for learning, clinical reasoning, and academic tasks. However, the psychological and technological factors that shape medical students’ acceptance of these tools remain insufficiently understood. This study examined acceptance among medical students with prior academic use of GenAI chatbots by extending the Technology Acceptance Model (TAM) to include AI literacy (AIL) and perceived security and privacy (PSP), and by examining their associations with perceived ease of use (PEOU), perceived usefulness (PU), attitude toward using GenAI (ATG), and behavioral intention to use GenAI (BIG). A cross-sectional online survey was conducted among medical students at Hormozgan University of Medical Sciences (HUMS), Iran, between May and June 2026. Eligibility required prior academic use of a GenAI chatbot. Data were collected using a structured, self-administered questionnaire and analyzed using partial least squares structural equation modeling (PLS-SEM). The measurement model was assessed for reliability and validity, while the structural model was evaluated using path coefficients (β), coefficients of determination ( R ²), and effect sizes ( f ²). Of the 292 students invited, 267 completed the survey, and 252 participants with prior academic use of GenAI chatbots were included in the final analysis (mean age, 22.64 ± 2.28 years; 147 [58.3%] female). The measurement model demonstrated adequate reliability and validity. AIL was positively associated with PEOU (β = 0.479, p < 0.001), while PSP was positively associated with PU (β = 0.157, p = 0.005). PEOU and PU were positively associated with ATG, and ATG showed the strongest association with BIG (β = 0.524, p < 0.001). The model explained 38.7% of the variance in BIG, with R ² values ranging from 0.229 to 0.387 across the endogenous constructs. The findings indicate that medical students’ acceptance of GenAI chatbots is associated with AIL and PSP alongside the core TAM constructs of PEOU and PU, with ATG showing the strongest association with BIG. Strengthening AIL and establishing clear privacy and security policies may support the informed and responsible integration of GenAI chatbots into medical education.