Key points are not available for this paper at this time.
Introduction This study aimed to explore the perceived knowledge, attitude, and practice (KAP) regarding artificial intelligence (AI) of medical students in Guangxi, China. Methods A cross-sectional survey was conducted from October to November 2024 at two universities in Guangxi, China. The survey assessed students’ KAP regarding AI using a 5-point Likert scale. Quantile regression models ( τ = 0.5) were used to identify factors associated with KAP scores. Results A total of 894 undergraduate medical students were enrolled. Participants demonstrated a moderate level of perceived AI knowledge (mean score 13.36 ± 3.26) and a moderately positive attitude (mean score 38.49 ± 5.69). However, AI practice was low (mean score 15.40 ± 4.57), with the highest frequency of AI practice for exam preparation (mean score 2.40 ± 0.84). Gender, academic year, major, hometown, visiting a science museum or exhibition in the past year, and studying AI during undergraduate education were associated with perceived AI knowledge. Gender, visiting a science museum or exhibition in the past year, and perceived AI knowledge were associated with attitude towards AI. Gender, academic year, studying AI during undergraduate education, perceived AI knowledge, and attitude towards AI were associated with AI practice. Significant barriers included limited practical opportunities, lack of specialized textbooks and courses, and insufficient professional guidance. Conclusion Medical students in Guangxi have moderate perceived AI knowledge and positive attitudes, but structural barriers (limited practical opportunities, lack of specialized textbooks/courses, and insufficient professional guidance) hinder AI integration into medical education. Based on these findings, we propose: integrating elective AI modules with hands-on workshops using open-source tools; developing open-access, low-cost learning materials and faculty training for rural-serving institutions; and fostering cross-disciplinary collaborations to apply AI to clinical cases. Future research should evaluate such interventions and address structural inequalities in AI learning.
Chen et al. (Tue,) studied this question.