This page was translated automatically from Turkish. Read the original

This page is an AI-assisted Turkish summary of the study presented at the 26th World Congress of Psychiatry. The study belongs to the presenting authors; please refer to the original source when citing it.

EV0518

From knowledge to action: a new role-based competency framework for artificial intelligence in medical education

FROM KNOWLEDGE TO ACTION: A NEW ROLE-BASED COMPETENCY FRAMEWORK FOR ARTIFICIAL INTELLIGENCE IN MEDICAL EDUCATION

A role-based competency framework for medical education is proposed that translates AI literacy into observable professional behaviors across eight core competency domains and three clinical roles.

The poster notes that although artificial intelligence is increasingly embedded in diagnosis, prediction, treatment, management, and communication, medical students and physicians feel only partially prepared for its responsible clinical use. It therefore aims to develop an updated competency framework that translates AI literacy into the observable professional behaviors required for safe and accountable clinical care. The study presents no empirical data; rather, it proposes a conceptual educational framework based on the observation that existing frameworks increasingly recognize digital competencies, yet AI-specific competencies remain inconsistent and predominantly knowledge-oriented, formulated as learning objectives and knowledge domains. The proposed framework comprises eight integrated competency domains: foundational AI literacy and core concepts; data literacy, bias, and statistical reasoning; critical appraisal of AI tools in clinical contexts; human–AI collaboration and professional accountability; ethical, legal, and social responsibility; patient communication and shared decision-making; interdisciplinary and system-level collaboration; and curriculum integration, assessment, and lifelong development. Alongside these domains, three clinical roles are defined: the critical appraiser, who evaluates clinical appropriateness, intended use, limitations, and external validity; the interpreter and communicator, who interprets probabilistic outputs, integrates them with patient-specific context, communicates uncertainty, and documents AI-supported decisions; and the safety steward, who detects automation bias and overreliance, recognizes system-level risks and recurring failures, and initiates review and escalation processes. The poster emphasizes that implementing the framework requires faculty development and adequate training for educators, longitudinal curriculum integration alongside postgraduate and continuing education, and institutional readiness encompassing real AI tools or simulations, governance structures, data protection, and accountability processes. It also highlights the need to address patient concerns in sensitive fields such as mental health, where the therapeutic alliance, digital phenotyping, and AI-mediated risk prediction raise distinctive ethical issues. In conclusion, it argues that physicians must evaluate, interpret, communicate, override when necessary, justify, and escalate AI-supported decisions in real-world clinical and sociotechnical settings. The poster reports no sample, statistical analysis, or numerical results.

All WCP 2026 summaries