World Bank: AI Goods Were 70%+ of Export Growth in 4 Nations

The World Bank’s East Asia and Pacific Economic Update, published October 6, 2026 under the title “Riding the AI Wave”, says the region is expected to grow by 4.5 percent in 2026. It says the 0.3 percentage point upgrade relative to its April forecast “is largely due to the strength of high-tech investments and exports in countries that participate in the global AI value chain”.

The report also says that, by April 2026, AI-related products were “accounting for more than half of total export growth in most countries and more than 70 percent in Malaysia, the Philippines, Thailand, and Viet Nam”. It says that “headline trade growth has been strong, but growth excluding AI-related goods has been weak or negative”.

Business Pill · WHO MADE THE TOTAL?

A one-minute explainer of the idea behind this story: concentration. It teaches the concept, not this story’s figures.

The key insight: The report attributes the 0.3 percentage point upgrade largely to high-tech investments and exports in countries in the global AI value chain, says growth excluding AI-related goods has been “weak or negative,” and says a reversal in global AI activity “would remove a key pillar” holding up growth in trade, investment and financial market valuations. All three are the World Bank’s statements, and this publication draws no conclusion about whether such a reversal will happen.

What the Report Says Drove the Upgrade

The report says: “A global surge in the production of AI-related goods is lifting growth in those countries in the East Asia and Pacific region that are involved in their manufacture, while high energy costs are weighing on growth elsewhere, particularly in Pacific Island economies”. It adds: “The region’s participation in fast-growing global value chains for AI-related goods has supported activity even as other drivers of growth have faded”. It also says “The drag from elevated energy prices has not been as severe as anticipated”.

For the longer run, the report says “Growth in the region is expected to average 4.4 percent per year in 2026–28”. Both growth figures are the World Bank’s forecasts.

Contributions to total export growth in percentage points, as the World Bank's Figure 1.5.F chart data give th
Contributions to total export growth in percentage points, as the World Bank’s Figure 1.5.F chart data give them; AI-related goods follow the WTO (2025) classification of AI-enabling products. 2026 covers exports through April (China: through May) compared with the same period of the previous year.

How the Report Defines AI-Related Goods

The shares depend on what counts as an AI-related good, and the report says so: “There are several methods for delineating the share of “AI-related goods” in total trade, with the WTO, the OECD, and the Federal Reserve all providing alternative definitions”. It adds: “All of these include foundational AI products like chips, semiconductors, and servers, but the OECD also includes upstream inputs such as manufacturing equipment and raw materials, and the WTO includes adjacent goods like telecommunications equipment”.

For its export-growth figure, the report says “AI exports follow the WTO (2025) classification of AI-enabling products”. It also says “AI-related exports are prominent in EAP countries, regardless of the definition used”.

Where AI Goods Carry the Export Growth

The report says: “By April 2026, however, export growth remains robust in the region, except in Indonesia, as increased demand for AI-related products has supported exports in the region, accounting for more than half of total export growth in most countries and more than 70 percent in Malaysia, the Philippines, Thailand, and Viet Nam”.

Figure 1.5.F in the report shows the split. Its note says: “Bars show each component’s contribution to total export growth by country”. For 2026, it says, “growth is calculated using exports through May for China and through April for other economies, compared with the same period in the previous year”. In the chart data file published with the report, the AI-related goods contribution for 2026 is 51.5 percentage points for Malaysia, 31.5 for Viet Nam, 26.5 for Thailand, 17.2 for the Philippines, 7.9 for China and 0.5 for Indonesia. The contribution of all other goods is 19.7, 6.7, 7.9, minus 0.1, 7.7 and 4.6 percentage points, in the same order.

On scale, the report says “China, Indonesia, Malaysia, the Philippines, Thailand, and Viet Nam exported US$1.4 trillion of AI-related goods to the world in the twelve months to April”. It continues: “Much of this went to the United States, which imported US$739 billion of AI-related goods in the 12 months to April 2026, one third of it from East Asia and Pacific”. It also says “Assembled equipment such as computers, servers, and routers accounts for 69 percent of Thailand’s AI-related exports and 60 percent of Viet Nam’s, with a larger share directed to final demand in the United States”.

What the Report Says Could Reverse It

The report says “The region’s dependence on AI-related industrial activity has been a source of strength, but it could become a weakness if global AI activity slows or reverses”. The report says that such a reversal “would remove a key pillar holding up growth in trade, investment, and financial market valuations in many countries”.

On investment, the report says “Gross AI-related capital expenditure has reached 6 percent of U.S. GDP (Armstrong 2025)—similar to the share information-technology investment reached at its peak in 2000, and to the peak share that residential investment reached in 2005”. It adds that “The current episode has risen faster than either previous cycle and is still gaining speed”. It also says: “As in previous episodes, a correction in AI investment would not necessarily indicate that the underlying technology had failed. Instead, it could indicate that investment had run ahead of realized demand”.

On financing, the report says “of the US$2.9 trillion of AI capital expenditure planned for 2025–28, US$800 billion is expected to be financed by private credit”. It says “AI-related lending has risen to 34 percent of private credit activity in 2025, from an average of 18 percent over the previous five years”, citing the FSB. On concentration in equity markets, it says “At the end of April 2026, the two Korean semiconductor majors accounted for 43 percent of the KOSPI market valuation”.

The report says “A 1-percentage-point slowdown in U.S. growth lowers growth in other emerging market and developing economies by an estimated 0.6 percentage point after one year, with an investment effect roughly twice as large”. It adds: “A slowdown concentrated in AI would be material for East Asia because of the region’s prominence in the AI supply chain”.

The Structural Read

The report’s argument has two sides. On the export side, it says AI-related products were accounting for more than half of total export growth in most countries and more than 70 percent in four, and that growth excluding AI-related goods has been weak or negative. On the investment side, it says gross AI-related capital expenditure has reached 6 percent of U.S. GDP and that US$800 billion of the US$2.9 trillion planned for 2025–28 is expected to be financed by private credit.

The report places the link between the two sides in the AI supply chain. It says the six economies it names exported US$1.4 trillion of AI-related goods in the twelve months to April, that the United States imported US$739 billion of such goods, one third of it from East Asia and Pacific, and that “A slowdown concentrated in AI would be material for East Asia because of the region’s prominence in the AI supply chain.”

The report keeps the reversal conditional. It describes it as a risk, says a correction would not necessarily mean the underlying technology had failed, and says it could instead indicate that investment had run ahead of realized demand. This publication does not assess how likely a reversal is.

The World Bank report

“Such a reversal would remove a key pillar holding up growth in trade, investment, and financial market valuations in many countries.”

Three Implications

DEFINITIONS MOVE THE SHARE The report says the WTO, the OECD and the Federal Reserve define AI-related goods differently, and its export-growth figure follows the WTO classification. This publication did not test the more-than-70-percent figure under another definition.

TWO STATEMENTS, KEPT APART The report says trade growth excluding AI-related goods has been “weak or negative” and that domestic activity has been weak in many countries. These are the Bank’s statements, and this publication does not say how long they will last.

A CONDITIONAL RISK, NOT AN EVENT The report describes a reversal in AI activity as something that “would” remove a key pillar, and says a 1-percentage-point slowdown in U.S. growth lowers growth in other emerging market and developing economies by an estimated 0.6 percentage point after one year. These are estimates under stated conditions, and the passages read do not say a reversal has happened.

What Is Not Established

Everything above is the World Bank’s statement in its own report. The 4.5 percent and 4.4 percent growth figures are forecasts. The investment, financing and spillover figures are the report’s estimates or its citations of other sources, and the report describes the reversal as a risk, not an event that has happened. The export shares depend on the WTO classification of AI-enabling products, which the report says is one of several definitions.

The six economy figures for Figure 1.5.F come from the chart data file published with the report, rounded to one decimal place, and the “2026” bars cover exports through April or May compared with the same period of the previous year. This publication read the report’s executive summary, the summary and AI-related passages of Chapter 1 (pages 3 to 8 and 20 to 23) and that chart data file.

It did not read Chapter 2, the boxes or the annexes, did not contact the World Bank, and verified none of the report’s claims independently.

Business Engineer Framework

The Map of AI — Where Chips, Servers and Data Centers Sit in the Stack

The Map of AI places more than 200 companies across nine layers of the stack, from silicon to application. The goods the report calls AI-related, including chips, semiconductors and servers, sit in the lower layers, and the Map shows who occupies the layers around them.

Read the Map of AI →

The Bottom Line

The World Bank’s October 6, 2026 East Asia and Pacific Economic Update says the region is expected to grow by 4.5 percent in 2026, a 0.3 percentage point upgrade it attributes largely to high-tech investments and exports in countries in the AI value chain, and that AI-related products accounted for more than 70 percent of export growth in Malaysia, the Philippines, Thailand and Viet Nam by April 2026.

The same report says growth excluding AI-related goods has been weak or negative, and that a reversal in global AI activity would remove a key pillar supporting growth. All of these are the World Bank’s statements, forecasts and estimates, not findings of this publication.

91,000+ executives read Business Engineer for the AI strategy frameworks cited by ChatGPT, Claude, and Perplexity.

A note on sourcing. We drew this piece from the World Bank’s East Asia and Pacific Economic Update, October 2026 (“Riding the AI Wave”), published October 6, 2026, reading the executive summary, parts of Chapter 1, and the Figure 1.5.F chart data. We did not contact the World Bank, and we have not checked its claims independently. Nothing here is a forecast or investment advice.

Sources: World Bank Open Knowledge Repository record · full report (PDF) · chart data workbook published with the report (Figure 1.5.F)

Scroll to Top

Discover more from FourWeekMBA

Subscribe now to keep reading and get access to the full archive.

Continue reading

FourWeekMBA