KECERDASAN BUATAN DALAM PENDIDIKAN KEDOKTERAN: KAJIAN NARATIF TENTANG ASESMEN, UMPAN BALIK, DAN PENALARAN KLINIS

Abstract

Artificial intelligence (AI) is increasingly shaping medical education, particularly in assessment, feedback, and clinical reasoning. This narrative review synthesizes recent evidence on the use of large language models, conversational agents, and AI-enhanced simulation in these three areas. The review highlights that AI can support educators by generating assessment materials, assisting rubric-based scoring, providing formative feedback, simulating patient encounters, and reviewing clinical reasoning documentation. These applications may help address persistent challenges in medical education, including limited faculty time, difficulty assessing higher-order reasoning, and the resource demands of high-quality simulation. However, the benefits of AI are accompanied by important risks. Concerns related to validity, bias, privacy, academic integrity, local relevance, and professional formation remain central, especially when tools developed in well-resourced and English-dominant contexts are applied in different educational and clinical settings. This review argues that AI should not be understood as a replacement for educational judgment, clinical supervision, or human relationships in medical training. Rather, AI is best positioned as a supervised educational tool that can extend the reach of good teaching when embedded in clear curricula, locally validated assessment systems, transparent governance, and faculty development programs. The central challenge for medical educators is therefore not whether AI should be used, but how it can be used responsibly while preserving the ethical, relational, and patient-centered foundations of medical education.