How South Korea Uses A.I. To Check on Its Elderly—and Fight Dementia
SEOUL—Every morning at 7 a.m., 83-year-old Lee Hye-sook receives a phone call that sounds almost human. A warm, slightly modulated voice greets her by name, asks how she slept, and reminds her to take her blood-pressure medication. If she doesn’t answer, the call loops back three times. If she still doesn’t pick up, a human social worker is dispatched to her apartment in Seoul’s Mapo district within the hour.
The caller isn’t a neighbor or a relative—it’s an artificial-intelligence system developed by South Korea’s Ministry of Health, and Welfare. Known as “AI Care Call,” the program has become a cornerstone of the country’s strategy to support its rapidly aging population, which is now the fastest-aging in the world. With nearly 18% of South Koreans aged 65 or older—a figure projected to rise to 40% by 2050—officials are turning to technology to fill gaps in care, reduce isolation, and detect early signs of cognitive decline.
“We’re not replacing human caregivers,” says Dr. Kim Ji-yeon, director of the Digital Health Bureau at the Ministry of Health and Welfare. “We’re using AI to make sure no one falls through the cracks.”
The Scale of the Challenge
South Korea’s demographic crisis is stark. The country’s fertility rate—0.72 births per woman in 2025—is the lowest in the world, whereas life expectancy has climbed to 84.1 years. The result is a shrinking workforce supporting a ballooning elderly population, many of whom live alone. Government data shows that more than 1.5 million South Koreans over 65 live in single-person households, a number that has doubled in the past decade.
Loneliness and cognitive decline are among the most pressing concerns. A 2024 study by the Korea Institute for Health and Social Affairs found that nearly 30% of elderly South Koreans living alone showed signs of depression, while dementia cases are expected to triple by 2050, reaching 3.1 million. The economic burden is equally daunting: dementia-related care costs are projected to exceed $30 billion annually by 2030, straining both public healthcare systems and family caregivers.
How AI Care Call Works
Launched as a pilot in 2021 and expanded nationwide in 2023, AI Care Call is a cloud-based system that uses natural language processing (NLP) to conduct daily check-ins with elderly individuals. The program is integrated with local government databases, allowing it to access medical histories, medication schedules, and emergency contacts. Here’s how it operates:
- Daily Check-Ins: The AI calls registered users at a scheduled time, typically between 7 a.m. And 9 a.m. The conversation is designed to mimic human interaction, with the AI asking open-ended questions like, “How did you sleep last night?” or “Did you take your medication today?”
- Voice Analysis: The system analyzes the user’s responses for signs of distress, confusion, or cognitive decline. Changes in speech patterns—such as slower responses, slurred words, or repetitive phrases—trigger alerts to human operators.
- Emergency Escalation: If the AI detects a potential issue (e.g., the user doesn’t answer, sounds disoriented, or mentions a fall), it escalates the call to a human social worker, who can dispatch emergency services or visit the home in person.
- Data Integration: The system logs each interaction, creating a longitudinal record of the user’s health and behavior. This data is shared with healthcare providers to track trends and intervene early if cognitive decline is detected.
“The AI isn’t just a chatbot—it’s a lifeline,” says Park Min-joo, a social worker in Busan who oversees a team of 12 human operators supporting the AI Care Call system. “Last month, the AI flagged a user who kept saying, ‘I think I fell.’ We sent someone over, and they found her on the floor. She’d been there for six hours.”
Beyond Check-Ins: AI for Early Dementia Detection
While AI Care Call focuses on daily monitoring, South Korea is also deploying AI to detect dementia earlier. In 2024, the Seoul Metropolitan Government launched a pilot program called “Dementia AI Watch,” which uses machine learning to analyze gait, speech, and facial expressions during routine medical check-ups.
The system, developed in collaboration with Samsung Electronics and Seoul National University Hospital, works as follows:
- Gait Analysis: Cameras in clinics capture how patients walk, measuring stride length, speed, and balance. Research shows that changes in gait—such as shuffling or hesitation—can appear years before cognitive symptoms.
- Speech Patterns: The AI records and analyzes speech during cognitive tests, looking for pauses, word-finding difficulties, or repetitive phrases.
- Facial Expressions: Subtle changes in facial expressions, such as reduced emotional responsiveness, can indicate early-stage dementia.
“Early detection is critical because interventions—like cognitive therapy or medication—are most effective in the early stages,” says Dr. Choi Eun-kyung, a neurologist at Seoul National University Hospital. “AI gives us a way to screen thousands of people quickly and consistently, without relying solely on subjective assessments.”
The program has already shown promise. In a 2025 study published in the Journal of Alzheimer’s Disease, the AI system detected early signs of cognitive impairment with 89% accuracy, outperforming traditional screening methods. The government plans to expand the program to all public health centers by 2027.
Privacy and Ethical Concerns
Despite its benefits, AI-driven elderly care has sparked debate in South Korea. Privacy advocates worry about the collection and storage of sensitive health data, while critics argue that AI cannot replace human connection.
“Technology should supplement care, not replace it,” says Lee Ji-hoon, a researcher at the Korea Institute for Health and Social Affairs. “There’s a risk that policymakers will witness AI as a cheap fix, rather than investing in human caregivers and community support.”
The government has attempted to address these concerns. AI Care Call data is encrypted and stored on secure servers, with access restricted to authorized personnel. Users must opt in to the program, and they can withdraw at any time. The Ministry of Health and Welfare has pledged to increase funding for in-home caregivers alongside its AI initiatives, though critics say the pace of hiring has been slow.
Global Implications: A Model for Aging Societies?
South Korea’s approach is being closely watched by other countries facing similar demographic challenges. Japan, which has the world’s oldest population, has experimented with AI-driven care robots and GPS tracking for dementia patients, while the European Union is funding projects like SHAPES, which uses AI to monitor elderly health.
“South Korea’s system is one of the most comprehensive we’ve seen,” says Dr. Alistair Burns, a professor of old-age psychiatry at the University of Manchester. “The combination of daily check-ins, voice analysis, and early detection is unique. But the real test will be whether it can scale without losing its human touch.”
For now, the results are encouraging. Since AI Care Call’s nationwide rollout, emergency dispatches to elderly individuals have decreased by 15%, while early dementia diagnoses have increased by 22% in participating regions. For users like Lee Hye-sook, the daily calls are more than a safety net—they’re a source of comfort.
“At first, I thought it was strange to talk to a machine,” she says. “But now, I seem forward to it. It’s nice to know someone—or something—is checking on me.”
What’s Next?
The South Korean government has ambitious plans to expand its AI-driven elderly care initiatives. By 2026, it aims to:
- Extend AI Care Call to 3 million elderly individuals, covering nearly half of the country’s senior population.
- Integrate AI Care Call with smart home devices, such as motion sensors and fall detectors, to provide real-time monitoring.
- Launch a pilot program for AI-powered mental health support, using chatbots to provide cognitive behavioral therapy (CBT) for elderly individuals with depression.
For policymakers, the stakes couldn’t be higher. With the number of elderly South Koreans expected to double by 2050, the country’s ability to care for its aging population will shape its economic and social future. AI, for all its limitations, offers a glimmer of hope—a way to extend independence, improve quality of life, and buy time as the country grapples with one of the most profound challenges of the 21st century.
The next official update on AI Care Call’s expansion is expected in June 2026, when the Ministry of Health and Welfare releases its mid-year progress report. Until then, systems like the one calling Lee Hye-sook each morning will continue to bridge the gap between technology and human care.
What do you think about using AI to care for the elderly? Should other countries adopt similar systems? Share your thoughts in the comments below, and don’t forget to share this article with anyone interested in the future of aging and technology.