East Asia Rides the AI Investment Wave as Uneven Growth Exposes a Wider Divide

East Asia and Pacific is projected to grow by 4.5% in 2026, with several economies outperforming earlier expectations as high-tech manufacturing and exports benefit from the global surge in AI-related investment, according to the World Bank’s latest bi-annual regional Economic Update. However, the headline resilience conceals widening differences across economies, while limited AI adoption, uneven workforce readiness and weak domestic demand in some markets are shaping a more complicated outlook.

East Asia and Pacific is heading into 2026 with stronger momentum than many parts of the global economy, but the region's 4.5% growth projection conceals a widening split in how countries are experiencing the technology cycle now reshaping trade and investment. Several economies are riding a surge in demand for the high-tech goods behind artificial intelligence, while others are being held back by weaker domestic demand, exposure to energy costs and limited buffers against external shocks.

The World Bank's latest bi-annual Economic Update presents a region that remains resilient but increasingly uneven. Vietnam, Malaysia and Thailand have all received upward growth revisions, while Pacific Island economies are expected to expand more slowly than previously projected. China, still the region's largest economy, is growing at a more moderate pace as a soft labour market and continuing adjustments in the property sector weigh on domestic demand.

The divide is not only between faster- and slower-growing economies. It also runs through the region's engagement with artificial intelligence itself. Some countries are already benefiting from supplying the hardware and manufactured goods behind the global AI investment surge, but the use of AI across firms, workplaces and sectors remains far less advanced.

The first AI dividend is arriving through factories, not offices

The clearest economic gains from the current AI cycle are appearing in trade and manufacturing rather than broad-based adoption. High-tech production and exports have helped several East Asian economies grow faster than anticipated in 2026, placing them in a favourable position within the supply chains supporting rising global investment in AI.

Vietnam's growth forecast has been raised by 1.1 percentage point to 7.4%, the strongest upward revision cited in the World Bank update. Malaysia's forecast has increased by 0.7 percentage point to 5.1%, while Thailand's projection has also been lifted by 0.7 percentage point to 2.0%.

The pattern highlights an important feature of the current technology cycle: economies can benefit from AI before the technology is deeply embedded across their own businesses. Producing the components and goods required by rising global investment can generate export demand even while domestic adoption remains constrained.

Such gains, however, are concentrated rather than universal. Pacific Island countries are projected to grow by 2.2%, half a percentage point below earlier expectations. Their vulnerability to high energy prices and limited capacity to absorb external shocks underscores how differently the same global environment can affect economies within a single region.

China presents another contrast. Growth of 4.4% still makes it a major contributor to regional activity, but domestic demand is restrained by labour-market softness and the continuing adjustment of the property sector. The combination reinforces the sense that regional resilience is increasingly being assembled from very different national growth stories.

AI is spreading, but the gap between access and use remains wide

The next stage of the region's AI story will depend less on whether the technology is available and more on whether firms can use it effectively. Adoption is increasing among individuals and businesses, yet it still trails advanced economies, with cost, expertise, security and privacy acting as persistent constraints.

High prices can limit the ability of firms to experiment with new tools, while shortages of relevant skills can make adoption difficult even when technology is accessible. Security and privacy concerns add another layer of hesitation, particularly for businesses handling sensitive information or operating without strong internal technical capabilities.

The World Bank sees the most immediate opportunity in what it calls "Small AI", adapting and applying existing technologies rather than focusing primarily on the development of frontier systems. The argument reflects the economic structure of much of the region, where the near-term productivity opportunity lies in making usable tools more accessible across ordinary businesses and sectors.

Local adaptation becomes especially important in this context. Affordable AI tools designed for local languages and practical sector needs could extend adoption beyond the technology industry and into areas capable of employing large numbers of workers, including tourism and agribusiness.

Such a shift would broaden the geography of AI gains as well as their sectoral reach. Economies currently benefiting through manufacturing exports would continue to capture value from the global technology cycle, while wider business adoption could determine whether productivity benefits spread more deeply through domestic economies.

Jobs are being reshaped through skills before they are displaced

The employment debate surrounding AI often centres on automation, but the World Bank's assessment points to a different near-term dynamic in East Asia and Pacific. AI has not yet had a significant impact on automatable jobs, even as employers begin changing the combinations of skills they demand.

Firms are increasingly looking for AI expertise alongside analytical capabilities and social skills. The emerging labour-market effect is therefore less about immediate job replacement and more about changing what workers need to contribute inside existing or evolving roles.

A deeper constraint sits beneath this transition. AI is most effective in supporting jobs that require complex thinking and judgment, yet only 13% of jobs in the region fall into this category, compared with 39% in advanced economies. The difference limits how widely the technology can complement workers under current employment structures.

Therefore, skills policy becomes inseparable from technology policy. Access to AI tools alone cannot produce broad labour-market gains if workers lack the capabilities required to use them productively or if economies generate too few jobs in which judgment, analysis and problem-solving play a central role.

The composition of employment may ultimately determine how inclusive the region's AI transition becomes. Workers already positioned in higher-complexity roles could benefit more quickly from complementary technologies, while others may depend on how effectively governments and firms adapt AI for industries that employ people across a wider range of skill levels.

The policy challenge is no longer whether to adopt AI, but how widely its gains can spread

The World Bank's recommendations place governments at the centre of the next phase. Stronger business conditions, reliable energy and digital infrastructure, access to finance and workforce skills are all identified as foundations for wider adoption.

Infrastructure is particularly significant because AI adoption depends on more than software. Firms require dependable digital systems, energy access and financing before new technologies can move from isolated experimentation into regular business operations.

Governments also face a dual responsibility as both users and regulators of AI. Public authorities can deploy the technology to improve services, while simultaneously building digital foundations, setting rules and supporting regional cooperation.

The balance will not be straightforward. Faster adoption can create new opportunities, but businesses still face legitimate concerns over security, privacy, expertise and cost. Expanding access without addressing those constraints risks producing a technology landscape in which adoption grows numerically without becoming economically deep.

Regional cooperation could become increasingly important because digital markets, business networks and technology supply chains extend across borders. The World Bank identifies cooperation as part of the policy response, alongside stronger regulation and public-sector use, although the release does not specify how those arrangements would operate.

Future growth will hinge on whether governments and businesses can narrow the gap between economies that manufacture for the AI boom and firms and workers that actually use AI productively. The 2026 outlook shows that East Asia and Pacific remains resilient, but the next phase will be judged by how broadly the gains from technology are distributed across countries, sectors and workers.