World Bank: East Asia's AI exports lack broad benefit

East Asia's AI Hardware Export Growth

East Asian economies have experienced a substantial increase in exports of artificial intelligence (AI) hardware. The World Bank reported that China, Indonesia, Malaysia, the Philippines, Thailand, and Vietnam collectively exported US$1.4 trillion worth of AI-related goods in the twelve months leading up to April.

This surge is also reflected in import trends, with US imports of AI-enabling goods rising by almost 60% in early 2026, even as its other imports saw a nearly quarter-point decline, according to the World Bank. Notably, AI-related exports from Vietnam and Malaysia have at least doubled when compared to their average between 2020 and 2022. This export performance is a primary factor behind the World Bank's upgraded 2026 growth forecast for the region.

AI Infrastructure's Modest Employment Impact

Despite the robust growth in AI hardware exports, the World Bank indicates that this trend does not significantly accelerate job creation in the same way traditional manufacturing once did. The institution notes that the historical role of manufacturing as an 'escalator' moving workers into higher-paid, more productive roles has slowed.

In Malaysia, for instance, data centres require an investment of RM4.6 million to create a single job, a figure substantially higher than the less than RM1 million needed for each job in the electronics and automotive sectors, as reported by the World Bank.

When assessing sectors capable of generating employment at scale, the World Bank identifies tourism and agribusiness as more significant contributors than AI.

Disparity in Value Capture

A closer examination of the types of AI-related goods being exported reveals a disparity in value capture within the global supply chain. For Thailand, assembled computers, servers, and routers constitute 69% of its AI-related exports, while for Vietnam, this figure stands at 60%, according to the World Bank.

China's AI-related exports, which totalled US$785 billion, are primarily intermediate components, many of which are shipped to ASEAN countries for final assembly. The World Bank highlights that the critical elements of design, software, intellectual property, and data, which confer the highest value, are largely developed and retained in other economies.

This pattern reflects a familiar challenge within global value chains, where the tasks a country performs determine its economic benefit more than the final products it ships.

Services as a Route to Economic Advancement

The World Bank's analysis also points to the growing importance of services in capturing value and generating employment. Services already account for approximately 26–33% of the value embedded in global manufactured exports.

Crucially, the World Bank's current AI trade figures do not fully encompass AI-related services, such as software development, business process outsourcing, and data work. Countries like the Philippines and Vietnam are actively building new capabilities in these service areas.

Vietnam's experience following its accession to the World Trade Organisation, which involved opening its telecommunications, finance, and other service sectors, led to productivity gains not only within those sectors but also in the manufacturing industries that rely on them.

Vietnam, while still lagging some East Asian peers in services openness, is pursuing an export-led growth path through enhanced services capabilities.

Policy Focus on Standards and Value

The economic terms governing the AI economy, including rules on cross-border data flows, digitally delivered services, standards, and export controls, are being established primarily through unilateral decisions by importing markets or via bilateral and mini-plurilateral agreements, rather than through multilateral forums.

Most developing countries currently have limited input into these rule-setting processes. The World Bank cautions that the current AI export boom, significantly bolstered by US hyperscaler investments exceeding US$1 trillion over 2025 and 2026, could reverse if this investment outpaces actual demand.

For policymakers and investors in Asia, securing a voice in global AI standards is a critical trade policy demand to ensure regional economies can capture greater value from the AI sector. The direction of capital flows and the sustainability of this investment cycle will be key areas to monitor.