Korea's medical AI advances, but clinical adoption remains slow

While artificial intelligence (AI) holds great potential to transform health care, its clinical adoption in Korea remains slow due to inadequate reimbursement and limited access to medical data, experts said.

The remarks came during a symposium hosted by Eulji University Medical Center to mark its 70th anniversary, under the theme "Future Healthcare with AI," in Seoul.

AI has already been transforming Korea's health care system. It has significantly improved public access to medical information, while helping physicians at primary and secondary health care institutions stay up to date with the latest medical research and developments.

Despite these advances, AI adoption in clinical settings remains slow.

While AI-based medical devices have gained regulatory approval in Korea, few are covered by national health insurance. Based on Ministry of Food and Drug Safety data, 549 AI medical devices have received regulatory approval until the first half of this year, but only 49 are covered by national health insurance.

"This is a global challenge. Regulatory approval does not necessarily mean a technology is needed in clinical practice," said Seo Joon-beom, professor at Asan Medical Center. "More importantly, Korea has yet to address its longstanding reimbursement issue — whether hospitals should be paid more for technologies."

Park Jung-hwan, healthy ministry's Medical AI and Data Policy Division, also pointed to regional disparities as one of challenges Korea is facing.

"Health care disparities between the Seoul metropolitan area and other regions remain significant, while underserved areas continue to struggle with essential services such as emergency, pediatric and maternity care," Park said. "New AI technologies are also concentrated in the capital region, leaving regional public hospitals behind in AI transformation."

To address these challenges, the government aims to make Korea the first country to fully integrate AI into health care, ensuring that all citizens benefit from AI-driven medical innovation.

The Presidential Council on National Artificial Intelligence Strategy established a national AI health care strategy in August, with the goal of building a digital health care infrastructure within the current administration's term.

"AI-powered health care services can overcome geographical barriers and narrow regional health care gaps," Park said. "AI could analyze patient transfer and emergency room data to ensure timely treatment during the critical golden hour. It could also analyze personal medical records to prevent the same tests when patients switch hospitals."

Going beyond the government's strategy, Seo stressed the need for Korea to develop its own healthcare foundation models rather than rely on foreign technologies.

He called for greater access to research data, citing Epic's database covering 100 million patients as an example. He also emphasized the need to integrate public health insurance data with private medical data and establish a public AI health care platform to support clinical adoption.

Kim Hyeon-chang, director of the Yonsei Institute for Digital Health, said the ultimate goal of medical AI is to advance future health care through early disease prediction, timely intervention, effective treatment and personalized health management.

"The goal is not simply to develop high-performance AI, but to improve disease prevention, medical treatment and health management," Kim said.

Korea Advanced Institute of Science and Technology professor Jung Jae-seung, known for brain science, agreed that AI could shift health care from treatment-focused to prediction and prevention, enabling continuous health management in everyday life.

Amid this shift, experts highlighted Korea's advantages, given its unified national health insurance system, extensive medical records and nationwide health screening data.

"University hospitals and national medical institutions must strengthen their ability to connect data across institutions, translate research findings into clinical value and integrate validated AI technologies into patient care and hospital operations," Kim further suggested.