Exploring medical Artificial Intelligence Readiness among Korean medical students: a cross-sectional study

Clinical Snapshot

Key Findings

• P Value: Specific p-values not reported in abstract; AI use frequency was 'significantly associated' with readiness; daily duration was not; gender differences were not significant; medical year students scored significantly higher than pre-medical students on the ability subscale only

• Effect Size: Not reported; no Cohen's d, eta-squared, or correlation coefficients with magnitude provided in the abstract

• Primary Outcome: Overall mean MAIRS-MS score of 4.19 out of 7 (moderate readiness). Subscale means: ethics 4.69, vision 4.44, ability 4.18, cognition 3.92

• Nnt Or Sensitivity: Not applicable to this cross-sectional survey study; no diagnostic, therapeutic, or prognostic metrics reported

• Confidence Interval: Not reported

Clinical Application

The MAIRS-MS instrument is potentially feasible for use as a needs assessment tool in medical schools to identify readiness gaps prior to implementing AI education modules. However, its psychometric properties require local validation before adoption in new institutional contexts. The findings support prioritising cognitive and practical AI competency training over attitudinal or ethical components, where readiness appears higher. Australian medical education is governed by the Medical Deans Australia and New Zealand framework and AMC accreditation standards. The Australian Digital Health Agency and RACGP have both identified digital health literacy — including AI competency — as an emerging priority for the medical workforce. The findings are indirectly relevant to Australian medical schools considering AI curriculum integration, though direct applicability is limited by the Korean single-institution context. The TGA has begun regulating software as a medical device (SaMD), including AI-based clinical decision support tools, making AI literacy increasingly relevant for Australian graduates. No PBS or specific TGA-listed AI tools are implicated in this study. Medical students in structured undergraduate medical programs, particularly those in early-to-mid training years where AI literacy curricula are being developed or revised

Abstract

INTRODUCTION: Artificial intelligence (AI) is rapidly transforming healthcare systems and clinical practice, increasing the need to prepare future physicians to work effectively with AI technologies in clinical contexts. Despite growing interest in integrating AI into medical education, empirical evidence regarding medical students' readiness to use AI remains limited. This study aimed to assess medical AI readiness among Korean medical students and explore factors associated with their readiness. METHODS: A cross-sectional survey was conducted among medical students enrolled in a six-year medical program at a university in Korea. A total of 204 students participated in the study. Medical AI readiness was measured using the Medical Artificial Intelligence Readiness Scale for Medical Students (MAIRS-MS), which assesses four domains: cognition, ability, vision, and ethics. Descriptive statistics, independent t-tests, and Pearson's correlation analyses were performed using SPSS version 27. RESULTS: The overall mean score of medical AI readiness was 4.19 on a 7-point Likert scale, indicating a moderate level of medical AI readiness. Among the subscales, ethics showed the highest mean score (4.69), followed by vision (4.44), ability (4.18), and cognition (3.92). The frequency of AI use was significantly associated with medical AI readiness, whereas the daily duration of AI use was not. No significant gender differences were observed. When students were divided into the low (pre-medical years 1-2) and high (medical years 1-4) groups, the high group showed significantly higher scores only in the ability subscale. CONCLUSIONS: The findings suggest that medical students are not yet sufficiently prepared to utilize AI technologies in clinical practice, particularly in terms of knowledge and practical competencies related to AI. These results highlight the need for structured and longitudinal AI education in medical curricula to better prepare future physicians for AI-integrated healthcare environments.

References

1. 1.Si, J. (2026). Exploring medical Artificial Intelligence Readiness among Korean medical students: a cross-sectional study. Medical Education Online. https://doi.org/10.1080/10872981.2026.2696610