Readiness and Perception Toward Artificial Intelligence Among Undergraduate Medical Students at a Tertiary Care Teaching Institute: A Cross-Sectional Study
TL;DR
Overall, undergraduate medical students demonstrated an average level of AI readiness and generally mixed-to-positive perceptions toward artificial intelligence, with a considerable proportion of students remaining neutral across several items.
Abstract
Background The growing integration of artificial intelligence (AI) into healthcare and medical education has created an urgent need to evaluate how prepared undergraduate students are to engage with these technologies. Medical graduates will increasingly encounter AI-driven tools across clinical and educational settings, yet systematic assessment of their readiness and perceptions remains limited, particularly in India. This study aimed to assess AI readiness and perceptions among undergraduate medical students and to examine how readiness varied in relation to sociodemographic characteristics, prior AI training, and patterns of AI tool utilization. Methods This cross-sectional study enrolled 310 undergraduate Bachelor of Medicine, Bachelor of Surgery (MBBS) students at a tertiary care teaching institution in Surendranagar, Gujarat, between September and November 2025. AI readiness and perception were assessed using the Medical Artificial Intelligence Readiness Scale for Medical Students (MAIRS-MS) and a separately developed, validated 10-item questionnaire, respectively. Descriptive statistics were summarized as means, standard deviations, frequencies, and percentages. Participants were categorized as having poor (≤66), average (67-78), or good (≥79) readiness using cut-offs derived from the 33.33rd and 66.67th percentiles of the observed MAIRS-MS score distribution. The same percentile-based approach was applied to each MAIRS-MS domain score. Pearson's chi-square test or Fisher's exact test, as appropriate, was used to examine associations between categorical variables, while Spearman's rank correlation coefficient was used to assess relationships between readiness scores and selected variables. A two-sided p-value of less than 0.05 was considered statistically significant. Results The mean age of the study participants was 20.03 ± 1.65 years, and 182 (58.71%) were female subjects. Using the 33.33rd and 66.67th percentiles of the observed MAIRS-MS score distribution, 169 of 238 participants (71.0%) were classified as having average AI readiness. Previous exposure to AI training (χ² =6.33, p=0.042) was significantly associated with the ethics domain of AI readiness. Total AI readiness showed extremely weak positive correlations with age (r=0.087, p=0.181) and academic year of study (r=0.057, p=0.381), with neither relationship reaching statistical significance. Among all 310 participants, more than half of the students perceived AI as useful for several educational purposes, including teaching (n=165, 53.23%), assignment preparation (n=162, 52.26%), self-learning (n=166, 53.55%), understanding complex concepts (n=170, 54.84%), and clinical case scenarios (n=161, 51.93%). Substantial proportions also expressed concerns regarding misleading information (n=151, 48.70%), potential effects on clinical skills and critical thinking (n=151, 48.71%), and data privacy (n=133, 42.90%). Conclusion Overall, undergraduate medical students demonstrated an average level of AI readiness and generally mixed-to-positive perceptions toward artificial intelligence, with a considerable proportion of students remaining neutral across several items. Previous AI training or exposure was significantly associated with the ethics domain of AI readiness. An extremely weak positive correlation was observed between the total readiness score and both age and academic year of study.