Artificial intelligence (AI) is no longer a futuristic concept in medicine. According to a Korea Health Industry Development Institute survey of 2,125 doctors, conducted in October 2025 and published on February 26, 2026, 47.7% of respondents had experience using medical AI, and 83.3% of those users employed it for medical image interpretation. Yet if such a system produces an incorrect result and a patient is harmed, who answers for it: the physician, the hospital, or the developer?

Most such tools support clinical judgment rather than replace it, but support is only as good as the system behind it. Unrepresentative training data can yield unreliable results for some patient groups, a missed disease can delay treatment, and a false alarm can trigger needless tests.

These risks highlight the importance of independently validating medical AI systems, monitoring their performance across different patient groups, and ensuring that clinicians carefully review AI-generated recommendations.

(출처: 메디팜스투데이 단순 영상판독 시대 끝, 의료AI 다음 단계로…병원업무 흐름 잇는 '연결성' 승부 기사)

The survey suggests why accountability matters: unclear legal responsibility in the event of a medical accident was a major concern for both doctors with AI experience (69.1%) and those without it (76.0%).

South Korea’s Digital Medical Products Act establishes a regulatory framework to ensure the safety, effectiveness, and quality of digital medical products, with approval, certification, and notification requirements. However, this framework does not automatically determine who is liable for every AI-related diagnostic error; responsibility must be assessed according to the circumstances of each case and the applicable laws.

The World Health Organization likewise stresses transparency and external validation. However sophisticated AI becomes, clinical judgment, honest communication with patients, and ethical accountability remain human tasks, and technological progress must advance together with patient safety.

Sources: 
(1) Yonhap News Agency, 26 Feb 2026, https://www.yna.co.kr/view/AKR20260226038200530  
(2) National Law Information Center, Digital Medical Products Act, https://www.law.go.kr/LSW/lsInfoP.do?ancYnChk=0&lsId=014601#0000  
(3) WHO, “WHO outlines considerations for regulation of artificial intelligence for health,” 19 Oct 2023, https://www.who.int/news/item/19-10-2023-who-outlines-considerations-for-regulation-of-artificial-intelligence-for-health

양유진 강남포스트 학생기자  webmaster@ignnews.kr

<저작권자 © 강남포스트, 무단 전재 및 재배포 금지>