Generative AI Use and Educational Needs among Korean Medicine Students: An Exploratory Study on Perception Differences by Academic Year

This study applies Kolb’s Experiential Learning Theory (ELT) and Davis’s Technology Acceptance Model (TAM) to characterize generative AI use among students in a College of Korean Medicine (KM) and to analyze perception differences by clinical practicum experience. A cross-sectional, mixed-methods survey was conducted with 126 students: 72 pre-clerkship Year 1–3 students (before university hospital practicum) and 54 clerkship Year 4 students (during practicum). The questionnaire covered five domains: (1) experience, level, and purposes of generative AI use; (2) expectations, concerns, and solutions for AI in KM clinical and educational contexts; (3) trust and validation strategies; (4) legal and ethical issues; and (5) overall concerns, mitigation strategies, and anticipated benefits. Quantitative items were analyzed with frequency and descriptive statistics, and free-text responses with thematic analysis. Most students regarded generative AI as an information and learning support tool and articulated the need for structured AI education. Technology-acceptance attitudes differed by practicum status: pre-clerkship students expected support for clinical reasoning and evidence-based decision-making, whereas clerkship students emphasized patient education and clinical communication. Major concerns across cohorts included reliability and accuracy, algorithmic bias, legal/ethical accountability, and the risk that overreliance could weaken self-directed learning and critical thinking. Overall, the findings demonstrate that practicum experience shapes acceptance and envisioned uses of generative AI and support a staged, ELT-based integration model: critical appraisal and creative reconstruction in pre-clerkship years, and safe, patient-facing applications with explicit validation and governance in clerkship. These implications can guide curriculum design and faculty development for responsible AI adoption in KM education.