AI alone won’t fix Korea’s eldercare crisis, experts say < Policy < Article

“Korea should not follow AI trends from other countries.” 

Korea’s push to integrate AI into eldercare is running into fundamental roadblocks. Speaking at a National Assembly forum on Thursday, Kim Hyun-jeong, Chairperson of the Korean Society of Digital Health, argued that the country needs to focus on its own urgent healthcare needs.

The reality, she said, is that Korea’s aging population has shifted the conversation from future potential to immediate demands.

By 2030, Korea is projected to become the world’s oldest country, surpassing Japan, with seniors aged 65 or older expected to account for 32 percent of the population by 2040.

A higher elderly ratio inevitably means a surge in medical needs. “The burden isn’t just increasing. It’s compounding,” Hwang Hee, CEO of Kakao Healthcare, said during the forum, which focused on AI’s role in eldercare and policy barriers. “Chronic disease management, long-term care, and hospital visits escalate exponentially in an aging society. We’re already seeing it.”

Hwang underscored the need for an equity-based approach to healthcare resource distribution. “Rather than distributing resources equally across all regions, more resources should be allocated where they are needed the most,” he said, adding that the medical needs of individuals over 65 differ greatly from those of younger populations.

Yet Korea’s healthcare and long-term care systems operate in silos. Medical services are covered by the National Health Insurance system, while long-term care falls under a separate insurance framework, and welfare services rely on local government funding. That fragmentation leads to inefficiencies, service gaps, and difficult choices for patients, Kim said. When individuals with disabilities turn 65, for instance, they must decide whether to stay within the disability welfare system or transition to long-term care insurance—often losing key services in the process.

Hwang emphasized that the ultimate goal of a welfare-based healthcare system “needs to move toward justice, not equality or equity,” ensuring fair and equitable access to medical services by addressing systemic barriers rather than just redistributing resources.

“The infrastructure is there,” Kim said, referring to Korea’s long-term care insurance, one of the most advanced globally, alongside Germany and Japan. “Even the U.S. doesn’t have a comparable structure. But digitization has to go beyond record-keeping. It has to enable actual care delivery.”

That’s where AI could help, but only if the right framework is in place. Professor Lee Ho-young of Seoul National University Bundang Hospital’s Department of Nuclear Medicine pointed to the U.S., where government-led legal and institutional reforms have shifted healthcare digitization from simple system implementation to value creation through data. Korea, in contrast, has relied on individual hospitals and private companies to drive digital adoption, often without interoperability or standardization.

“Since 2000, 95 percent of Korea’s medical institutions have adopted electronic medical records (EMR),” said Lee. “But standardization is still an afterthought. We’re only now recognizing how crucial it is for AI-driven eldercare.” Without standardized data, he warned, AI systems will remain fragmented, limiting their real-world impact.

Cheon Ha-ram, a member of the New Reform Party, acknowledged that Korea lags behind the U.S. and China in AI investment but argued that vertical AI applications, particularly in healthcare, represent a space where Korea can lead. “We don’t need to compete in general AI development,” he said. “We need to own specific domains.”

Kakao Healthcare is already working on one such domain: AI-driven chronic disease management. Its AI-powered blood glucose monitoring app, PASTA, has around 100,000 users, integrating real-time data with lifestyle recommendations. 

The company announced a partnership with senior care provider Caring to deploy AI-enabled health management solutions in Caring-operated senior housing facilities three days earlier. As part of the collaboration, PASTA will be integrated into these settings to enhance disease management and preventive care for aging populations.

Professor Kang Jae-woo of Korea University’s Department of Computer Science and Engineering added that Korea’s AI-driven healthcare push cannot move forward without better data accessibility for researchers. “Access to large-scale data is crucial. Researchers must be able to utilize extensive datasets,” Kang said.

He also proposed piloting AI eldercare technologies in nursing homes. “To properly test AI eldercare solutions, we need to designate select nursing homes as pilot sites where these technologies can be implemented without regulatory constraints,” he suggested. “While AI projects are being initiated, there has been no commitment to easing regulations, which remains a significant barrier.”

“Digitization is binary. It either works or it doesn’t,” Kim added. “Most digital healthcare solutions fail when they hit real-world caregiving. AI can’t just be a tech layer on top of broken systems. It has to make care more inclusive, efficient, and sustainable.”