Asia’s growth model is under intense strain, and rapid population ageing will only make matters worse. But Lee Jong-Wha thinks the key to maintaining its momentum will be adopting AI, building human capital, and a framework of enabling institutions

Two decades after globalisation fueled a global economic boom, growth has shifted onto a more subdued path, where it is likely to remain for the foreseeable future. Beyond the immediate shock of fragmenting trade and investment ties – a result of rising geopolitical tensions, particularly between the United States and China – lie structural headwinds, including population aging, stagnant productivity, and the growing costs of inequality and natural disaster. These challenges strike at the heart of Asia’s growth model.

Not only is fragmentation causing Asia’s export-oriented growth engine to sputter; rapid population aging is tightening the labour supply and compounding fiscal pressures across the region. The share of people aged 65 and above in Asia will nearly double by mid-century, from 9.8% in 2023 to 18.6% by 2050. Japan and South Korea are already “super-aged” societies, with more than 20% of the population aged 65 or older, and China’s population has begun to decline. India and some Southeast Asian economies still enjoy a demographic dividend, but that window is closing quickly.

But demographics need not be destiny. Using new digital technologies – and, in particular, AI – Asia can invigorate productivity growth, ease labour shortages, and extend people’s working lives. After all, like electricity and the internet, AI is a general-purpose technology with the potential to transform production, services, and innovation.

But, while early evidence indicates that AI can improve worker productivity, enable the automation of a growing range of tasks, and spur the creation of new products and services, the technology’s broader macroeconomic impact remains uncertain. Some analyses suggest that AI could raise global productivity growth by 0.8-1.3 percentage points annually over the next decade, while others offer more conservative estimates.

Moreover, gains from AI are likely to be concentrated among a few leading firms and sectors at first. Broader productivity gains often take time to emerge, and productivity might even decline in the short term, as firms bear the high costs of adoption, training, and integration – a pattern known as the “productivity J-curve.”

As with previous technological revolutions, AI’s impact will ultimately depend on countries’ capacity to absorb and apply it effectively. For now, this capacity varies widely across Asia. According to the IMF’s AI Preparedness Index, advanced economies, such as Australia, Japan, and Singapore, rank above their peers elsewhere, based on indicators like digital infrastructure, economic integration, regulation, and labour-market policies. The region’s large emerging economies, including China, India, and Indonesia, also perform better than their peers. But low-income economies, like Bangladesh and Cambodia, are lagging.

Given the speed and scale of the transition ahead, and the gathering growth headwinds, all Asian economies should be working to strengthen their ability to harness AI. To understand what must be done, it is worth revisiting the Nobel Prize-winning research of Philippe Aghion, Peter Howitt, and Joel Mokyr. Aghion and Howitt formalised Joseph Schumpeter’s idea of creative destruction – the process by which new technologies replace old ones – while Mokyr, drawing on history, found that lasting progress also requires openness to new ideas and a commitment to scientific inquiry. Together, they show that sustained prosperity depends on continuous technological progress supported by an enabling culture and institutional framework.

For Asian countries, the lesson is two-fold. First, they must foster the skills people need to make use of AI. Investment in digital literacy and STEM education (science, technology, engineering, and mathematics) can help to prepare people for the transformation ahead, while lifelong-learning and reskilling programs can smooth the adjustment for those already in the workforce. With AI enabling the automation of a growing range of routine and complex tasks, complementary technical, digital, and social skills are essential.

Such programs should be broadly accessible, and complemented by strengthened social-protection systems and structural reforms that reduce labour-market dualism, thereby ensuring that good, secure jobs remain widely available. Otherwise, job polarisation and income disparities could widen, and growth will become more imbalanced and less resilient.

Beyond fostering the skills people need to make use of AI, Asian governments must make it possible for people to apply those skills creatively. This means investing in research and development, building reliable and broadly accessible digital infrastructure and secure data systems, and designing regulatory frameworks that ensure ethical AI use. Open competition, cross-border research collaboration, and greater access to finance for startups can then ensure that innovation is not confined to a few firms or countries. Such inclusion is vital: progress demands that new ideas and young firms have the tools and opportunities to challenge incumbents.

The stakes are high. If AI adoption remains limited to a few frontier economies or sectors, technology gaps would widen, with far-reaching economic, social, and political consequences. But if Asia harnesses human creativity and AI tools effectively, it can build resilient, innovative, and inclusive economies capable of overcoming current challenges and unlocking new sources of growth.

Lee Jong-Wha, Professor of Economics at Korea University, is a former chief economist at the Asian Development Bank and a former senior adviser for international economic affairs to the president of South Korea. (c) 2025 Project Syndicate. Used here with permission.