Generative artificial intelligence (AI) could significantly reshape economies and labor markets across Asia and the Pacific, but widening disparities in readiness mean advanced economies are likely to capture most early gains, according to a report by the Asian Development Bank (ADB) on Friday.
In its latest assessment of AI preparedness, the ADB said the economic impact of generative AI will depend heavily on how widely and effectively it is adopted, with advanced economies better positioned to benefit due to stronger digital infrastructure, computing capacity and technological capabilities.
Countries such as Australia, Hong Kong, Japan, the Republic of Korea, New Zealand and Singapore were found to be clustered near the global frontier in digital infrastructure, scoring an average of 0.19 on the AI Preparedness Index’s infrastructure component.
By contrast, many developing economies — including Cambodia, India, Myanmar, Papua New Guinea and the Philippines — scored below 0.11, reflecting weaker connectivity, limited computing capacity and underdeveloped data infrastructure.
The ADB said these gaps create cascading constraints that slow AI diffusion, as limited computing resources restrict access to cloud services and reduce firms’ ability to train or deploy AI systems. This, in turn, delays integration into production processes and limits productivity gains.
“Without adequate infrastructure, the ability to scale AI adoption remains constrained, particularly in developing economies,” the report said.
Human capital gap slows adoption
Beyond infrastructure, the report highlighted significant differences in human capital, noting that AI’s productivity benefits depend on workers having complementary skills to effectively use the technology.
Developing Asia and the Pacific (DAP) scored an average of 0.13 on human capital and labor market readiness, compared with 0.17 in advanced Asia and the Pacific (AAP), underscoring a persistent skills gap.
Job posting data shows higher and faster-growing demand for AI-related skills in economies such as Singapore and the Republic of Korea, compared with developing economies including India, Malaysia and the Philippines.
The ADB said this reflects differences in firm readiness and organizational capabilities, which influence how quickly productivity gains from AI can be realized.
Innovation ecosystems reinforce divide
The report also pointed to innovation capacity and economic integration as key drivers of divergence.
DAP economies lag AAP peers in innovation and integration, with average scores of 0.11 versus 0.17 on related components of the AI preparedness index.
Leading economies such as China, Japan and South Korea benefit from strong government support and corporate research and development investment, enabling both AI development and local adaptation.
In contrast, many developing economies rely heavily on imported technologies and lack strong domestic research ecosystems, limiting their ability to adapt AI tools to local conditions, languages and markets.
The ADB added that participation in global AI-related value chains — including electronics and computing equipment — is significantly higher in economies such as Singapore and Hong Kong, compared with most developing markets. This integration helps accelerate technology spillovers and adoption.
Institutional quality shapes deployment
Institutional strength was identified as another key determinant of AI uptake, with regulatory clarity, governance frameworks and policy design playing a critical role in shaping investment and deployment decisions.
The report found that DAP economies scored 0.12 on regulatory and ethics frameworks, compared with 0.20 in advanced economies, reflecting weaker institutional readiness.
Uncertainty over regulation, weak enforcement and gaps in data governance and privacy standards were found to discourage investment in AI systems and slow adoption, even where technical capacity exists.
“Credible and transparent governance frameworks reduce uncertainty and support wider uptake across both public and private sectors,” the ADB said.
Economic structure limits gains in developing Asia
The ADB also highlighted structural differences in economies, noting that exposure to AI varies significantly across sectors.
Services such as finance, education, information and communications, and professional services are the most exposed to generative AI due to their reliance on data-driven and cognitive tasks.
By contrast, agriculture, transport and construction are less exposed, as they depend more on physical labor and manual tasks, which are more likely to be affected by robotics than generative AI.
Because developing economies in Asia have larger agricultural sectors and smaller services industries, the report said near-term gains from AI adoption are likely to be more limited.
However, lower exposure may also reduce the risk of rapid labor displacement, allowing a more gradual adjustment of skills and employment structures.
Growth gains skew toward advanced economies
Using a multi-economy, multi-sector dynamic general equilibrium model, the ADB estimated that AI-driven productivity shocks vary significantly across sectors and regions.
Productivity gains from AI are estimated at 0.1 percent to 4.5 percent in agriculture, 0.3 percent to 9.1 percent in industry, and 0.3 percent to 9 percent in services, depending on adoption scenarios.
Overall productivity gains in developing Asia are projected at 0.2 percent to 6.9 percent, lower than in advanced economies due to weaker exposure and readiness.
As a result, advanced Asia and the Pacific economies and the United States are expected to see gross domestic product (GDP) growth increases of 0.6 to 2.1 percentage points by 2030, easing to 0.4 to 1.5 percentage points by 2040.
Developing Asia is projected to see smaller gains of 0.2 to 1.8 percentage points in 2030 and 0.1 to 1.6 percentage points in 2040, though catch-up effects could add up to 0.4 percentage points in some cases.
Within developing Asia, China is expected to record the largest gains, followed by India and other emerging economies.
Investment and trade to shape outcomes
The report said investment would be the main driver of growth gains in developing economies, while exports would play a larger role in advanced economies due to rising global demand for AI-enabling goods such as electronics and computing equipment.
However, developing economies may face short-term pressure from increased imports of capital goods needed for AI-related investment.
Over time, sectoral gains are expected to be led by services, which will benefit most from AI adoption, while industry will gain from increased production of AI-related hardware.
Labor market transition risks
The ADB warned that AI will have uneven impacts on employment, with both job creation and displacement occurring simultaneously across sectors.
In advanced economies, employment gains are expected mainly in services, with limited disruption elsewhere. In developing economies, outcomes will be more mixed, with some short-term gains in agriculture and industry but broader displacement pressures in less-prepared economies.
In the rest of developing Asia, weaker output growth could lead to net employment declines and slower labour market adjustment.
Policy priorities
The ADB said governments must act to strengthen AI readiness through investments in digital infrastructure, education, innovation systems and institutional frameworks.
It also urged policies to support workforce reskilling, expand social protection systems and promote participation in AI-related global value chains.
“Targeted reforms can mitigate short-term displacement risks while enhancing long-term productivity gains,” the report said.
The bank added that services-led transformation is likely to be the dominant channel through which AI reshapes developing economies, while industrial upgrading and skills development will be critical to ensuring inclusive growth.
ADB launches $25M fund to accelerate progress on ASEAN Power Grid