Association of generative artificial intelligence use with perceived academic benefit and reported performance among medical students

Generative artificial intelligence (AI) tools are widely adopted by medical students. Although student attitudes toward AI have been surveyed extensively, far less is known about whether students who use these tools more often also report better academic outcomes. This study asked whether AI usage frequency is associated with perceived academic benefit and with self-reported academic performance among medical students at a single institution. A cross-sectional survey was administered to medical students at a single U.S. allopathic medical school. Students reported AI usage frequency, perceived academic benefit (composite of three Likert-scale items; Cronbach’s α = 0.84), self-reported overall exam performance, and self-rated academic standing. Spearman correlations with 95% confidence intervals and Kruskal-Wallis tests were used to examine associations. Of 103 respondents, students who used AI more frequently reported significantly greater perceived academic benefit (ρ = 0.69, 95% CI [0.57, 0.78], P < 0.001). However, AI usage frequency was not correlated with self-reported exam performance (ρ = −0.02, P = 0.81) or self-rated academic standing (ρ = −0.01, P = 0.93). Over half of students (53.4%) reported using AI Often or Always for studying. In this exploratory single-institution study, students who used AI more frequently perceived greater academic benefit, but this perception was not reflected in coarse, self-reported performance measures. This discrepancy should be interpreted cautiously and is at least as likely to reflect a measurement artifact, arising from the insensitivity of a single global performance item, recall bias, and the inability to distinguish performance before and after AI adoption, as a true gap between perception and attainment. Findings underscore the importance of helping medical students critically evaluate their AI-assisted study practices and identify priorities for more rigorous future investigation using objective, granular performance outcomes.