South Korea

South Korea is aging faster than any country in recorded history. It crossed the "super-aged" threshold — more than 20% of the population aged 65 or older — in 2024, a transition that took France nearly four decades and South Korea roughly seven years. A striking share of older Koreans live alone: about 37.8% in 2024, according to analysis by the Carnegie Endowment (Carnegie Endowment). That combination of speed, scale, and isolation is what makes the country a real-world laboratory for caregiving technology.

The most distinctive response is the AI wellness-check call. Naver's Clova CareCall, SK Telecom's NUGU CareCall, and KT's AI Care service place automated, conversational phone calls to seniors living alone — asking about meals, sleep, and mood, and remembering earlier conversations. What matters most is the back end: if a senior doesn't answer or sounds unwell, the system flags a local welfare officer to follow up. By August 2024, Naver reported Clova CareCall running in roughly 128 municipalities with about 30,000 users (Telecompaper). It is, in effect, a national experiment in pairing cheap automated outreach with a human safety net.

Robots and devices fill out the picture. The Hyodol companion doll — a soft, AI-voiced "robot grandchild" — has been distributed mainly through government and welfare channels, with reporting citing over 12,000 units in homes by late 2025 (Rest of World). Seoul installs IoT sensors that watch for movement and temperature in vulnerable seniors' homes and alert a caseworker when something looks wrong, while SK Telecom's NUGU speaker lets a senior summon emergency help by voice. Heavier physical-assist robotics remain at the pilot stage, with national plans to develop "Physical AI" care robots from 2028.

Underpinning all of this is a formal care system that predates the technology. South Korea introduced mandatory Long-Term Care Insurance in 2008 — an early move among Asian nations to socialize eldercare. It delivers mostly in-kind services rather than cash, and has grown from 214,000 beneficiaries in 2008 to about 1.1 million — roughly 11% of the elderly — by the end of 2023 (Korean Long-Term Care System: 2024 Update).

But the system is straining. Professional care is unaffordable for many families, an estimated shortfall of around 116,000 care workers looms by 2028, and the country has begun opening foreign-worker visa pathways to fill the gap. The insurance model leans on paid services rather than supporting family caregivers directly, and many families end up cutting work hours or quitting to provide care themselves.

Two things have sharpened since we first wrote this entry. The first is scale: Hyodol has now reached roughly 14,500 homes, and domestic makers Angel Robotics and Cosmorobotics both listed on the KOSDAQ at steep premiums, so the sector is drawing real capital rather than pilot grants. Korea has also moved the coordination up a level, with two ministries writing a full-cycle AI care-technology strategy. The same companion device that is a boutique purchase in the United States is, in Korea, closer to municipal equipment — which is the whole difference between a market and an infrastructure.

The second is that the evidence underneath it is thinner than the deployment. In our own reporting on Hyodol, the strongest study found reduced depression and better quality of life at three months and no significant effect by six, and a separate trial helped users with mild cognitive impairment while doing nothing for moderate. Public funding buys distribution at a scale no consumer market can match. It does not buy a device that keeps working.

And the bill is coming due. A peer-reviewed fiscal projection expects the long-term-care fund's accumulated reserves to be exhausted by 2030, with annual deficits climbing steeply through the 2030s (fiscal projections, 2023–2042). That is the sentence every other country should read twice, because Korea built its system two decades early and is still running out of money first.

The honest frame is the contrast. South Korea is genuinely ahead of the world on deploying AI and robots in eldercare — and at the same time carries the OECD's highest elderly poverty rate (about 39.8% in 2023) and highest elderly suicide rate, several times the OECD average. The technology demonstrably helps with loneliness in early studies, but it is being asked to stand in for income security and human care that aren't fully there. For caregivers watching from elsewhere, the lesson is encouraging and sobering at once: a well-designed automated check-in can extend a thin human safety net — but it cannot replace one.