The Silver Economy’s AI Payoff Depends on What Regions Build Around It

A region can develop AI patents, firms and digital networks while older residents still struggle to find dependable care. The distance between technological capacity and services people can use is the key issue examined in the study "The Impact of Artificial Intelligence on the Sustainable Development of the Silver Economy: Empirical Evidence from China," by He Zhang, Mengda Lu and Yingyan Wu, published in Sustainability.

Based on data from 30 Chinese provincial-level regions between 2014 and 2023, the researchers find that stronger AI development is generally associated with stronger conditions for a sustainable silver economy. However, the relationship changes with local circumstances. It is stronger where financial development is greater and, in the study's analysis, turns negative below a threshold in regional marketization.

The aging economy needs more than growing demand

The silver economy is described in terms of the products and services older people might buy. That account leaves out a harder problem: whether care organizations have the staff, facilities and funding to meet varied needs reliably. A larger older population may increase demand while placing greater pressure on families, workers and public services.

The researchers measure more than the size of the market. Their silver-economy index includes care facilities and personnel, pension provision, the share of older residents and indicators of regional environmental conditions. It describes a province's broader capacity to sustain aging-related activity, rather than the revenue of care businesses or the quality of any one service.

AI could help providers identify needs, coordinate resources and reduce time spent on routine tasks. Those possibilities are especially relevant where services must respond to people with different health conditions, incomes and levels of digital confidence. But a province's ability to develop AI does not show that its care providers have adopted useful tools or that older residents have benefited from them.

The provincial pattern is positive, but conditional

Across the study period, provinces with greater AI development tended to score higher on the silver-economy index. The relationship remained positive when the authors changed aspects of the sample and analysis, including removing the pandemic years of 2020–2022 and examining AI development from the preceding year. A further statistical check using AI pilot-zone policy produced a similar positive estimate.

The research also points to two possible ways the relationship operates. Higher AI development was associated with greater service-sector output per worker and with higher research and development spending relative to regional GDP. Better organization of services could help providers use scarce resources; sustained innovation spending could help adapt technology to aging-related needs.

Those measures cover entire regional economies, however. Service-sector productivity does not tell us whether home care became more accessible, and overall R&D spending does not identify investment in elder care. The findings make these channels plausible without demonstrating that a particular AI application improved an older person's experience.

The clearest warning against a uniform AI strategy comes from the study's comparison of local conditions. The positive association was stronger in provinces with greater financial development. When the researchers divided observations by a marketization-index threshold of 7.039, the estimated AI relationship was negative at or below that level and positive above it. That number is specific to the study's measure and sample, not a target that governments elsewhere should try to reach.

Technology spending has to meet service capacity

AI investment and service-system development need to move together. Care providers require workers who can use new tools, organizations able to act on the information they produce, and financing to maintain them. Without those supports, more regional AI activity need not translate into better aging-related provision.

For governments, the immediate question is where technology could relieve an identifiable service bottleneck. Tools intended to improve scheduling, for example, should be assessed against whether people receive timely help and whether workers can use the system effectively. Public procurement and investment decisions need evidence about service outcomes, not only evidence that a region has a growing AI industry.

The implications extend beyond China, particularly for developing countries managing aging alongside uneven digital access and constrained care budgets. The study suggests that adopting AI without examining local finance, workforce capacity and market conditions could produce sharply different results from those expected.

There is also a distributional question. A service may become more efficient on average while remaining difficult to reach for people with low incomes, limited connectivity or complex care needs. Businesses, public agencies and civil society groups will need measures of affordability, accessibility and care quality to determine whether technological gains reach the people an aging economy is meant to serve.

What the evidence can, and cannot, settle

Ten years of provincial data allow the authors to compare changes within regions, while their additional analyses test whether the main relationship persists under different assumptions. Examining finance and market institutions brings the conditions for using AI into the assessment.

Its scale also sets limits. The 300 province-year observations cannot show what happens inside care homes, clinics or households. The AI index combines patents and firms with infrastructure, industrial robots and human-capital measures, so it captures a regional technology ecosystem rather than direct AI use in older people's services. Likewise, the silver-economy index describes broad regional capacity, not individual wellbeing.

The authors caution that their statistical approach does not remove every alternative explanation. Provinces designated as AI pilot zones, for instance, may also receive other forms of innovation and development support. The results consequently support a robust regional association, but they do not prove that AI caused the observed gains or losses.

The next evidence gap is concrete: which tools improve care, for whom, at what cost and under which local conditions? Research using data from providers and older residents could test changes in access, service quality and affordability after specific AI applications are introduced.