Objective To explore the internal mechanism of the influence of artificial intelligence (AI) on medical students' creativity.

Methods Based on the stimulus-organic-response theoretical model, a cross-sectional survey of medical students in 6 undergraduate colleges in Anhui province was conducted using general data questionnaire, perceived AI system characteristics scale, perceived AI providing support scale and perceived AI influencing creativity scale.

Results Cronbach's α coefficient of all scales was more than 0.8, average variance extracted was more than 0.5, composite reliability was more than 0.7, factor loading of each question was more than 0.6, variance inflation factor was less than 5, and average variance extracted of each question was more than its square correlation with all other questions. The scores of medical students' perceived AI system characteristics scale, perceived AI providing support scale and perceived AI influencing creativity scale were all above 3 points. The accuracy, intelligence and interpretability of perceived system characteristics, information support, heuristic support and evaluation support of perceived providing support, and perceived AI influencing creativity were positively correlated with each other (P ＜ 0.01). The structural equation model verified the 7-factor model with the fitting indexes as follows: χ² = 259.111, χ²/df = 1.713, RMSEA = 0.049, GFI = 0.959, AGFI = 0.971, CFI = 0.982, NFI = 0.970, IFI = 0.976, and produced six mediating paths of accuracy → information support → creativity, accuracy → evaluation support → creativity, intelligence → information support → creativity, intelligence → heuristic support → creativity, intelligence → evaluation support → creativity and interpretability → evaluation support → creativity which passed the mediation test.

Conclusions This study verifies the internal mechanism of the influence of AI technology use on the creativity of medical students in school, which has practical significance for medical students to use AI to assist learning and scientific research.