World Bank raises this year’s East Asia-Pacific growth forecast to 4.5% on AI

[Financial News] The World Bank (WB) sharply raised its economic growth forecast for the East Asia and Pacific (EAP) region this year, reflecting strong exports of semiconductor- and artificial intelligence (AI)-related products. However, it issued a strong warning that excessive reliance on the AI boom could become a major liability if global IT investment cools in the future.
A report released by the WB on the 6th (local time) projected that the East Asia and Pacific economy, comprising 23 countries including China, Vietnam, Indonesia, Malaysia and Thailand, would grow by 4.5% this year. That is 0.3 percentage points higher than the previous forecast issued last April. Growth is expected to gradually ease to 4.4% next year and 4.3% in 2028. Among major economies, Vietnam saw the largest upward revision, with its growth forecast rising 1.1 percentage points to 7.4%.
The World Bank identified AI-related manufacturing and exports as the main drivers of the growth forecast upgrade. However, it assessed trade growth in products other than AI goods as “weak or negative.”
Indeed, AI-related goods accounted for at least half of export growth in most East Asian countries. In Malaysia, the Philippines, Thailand and Vietnam, at least 70% of total export growth came from AI-related goods. According to the report, the value of AI-related products shipped by six countries—China, Indonesia, Malaysia, the Philippines, Thailand and Vietnam—over the 12 months through April this year alone reached $1.4 trillion (about 1,882 trillion won).
South Korea, too, has seen an overwhelming contribution to AI exports, led by semiconductors. According to recently released official data, South Korean exports in September surged 83.5% year on year to a record $120.9 billion, with semiconductors accounting for half of total exports. The World Bank noted that just two semiconductor companies, Samsung Electronics Co., Ltd. and SK hynix, account for 43% of the total market capitalization of the KOSPI (as of the end of April), highlighting the South Korean market’s concentration in specific high-tech industries.
The World Bank raised serious concerns about the pace of AI investment and how it is financed. AI-related capital expenditure (CAPEX), currently equivalent to about 6% of U.S. gross domestic product (GDP), is on a par with the peak in IT investment during the 2000 dot-com bubble. The report noted, “This AI investment cycle is rising faster than any previous one and is still accelerating.” The Bank for International Settlements (BIS) had also warned in its annual report last June that “the scale and speed of the current AI boom are reminiscent of the dot-com frenzy of the 1990s and past speculative manias.”
A particular risk identified was the lack of transparency around the funding going into AI investment. Of the $2.9 trillion in AI investment planned for 2025 to 2028, $800 billion is expected to come from private credit. AI-related loans’ share of all private credit surged from a five-year average of 18% to 34% in 2025. Meanwhile, private-credit portfolios have recently seen a series of asset write-downs, outflows and defaults. The World Bank expressed concern that “the private-credit market is less transparent and has not yet been tested by a severe recession.”
A shift in the direction of global monetary policy is another headwind. The U.S. central bank, the Federal Reserve System (Fed), raised interest rates for the first time in more than three years last month and signaled another hike within the year. As major central banks have begun raising rates again for the first time since 2023, the AI investment boom, which has relied on abundant liquidity, could slow, analysts say.
The World Bank said that an adjustment in AI investment would not necessarily mean “the end of the AI supercycle,” but that current investment is “running far ahead of actual market demand.”
If U.S. growth were to fall by 1 percentage point, growth in other emerging economies would decline by about 0.6 percentage points, while the impact on investment would be estimated to be at least twice as large. The World Bank stressed that “given East Asia’s central role in the AI supply chain, the impact would be substantial if an AI-focused slowdown materialized.” Countries with relatively high foreign-currency debt in the banking sector, including Malaysia (29.2% of GDP) and the Philippines (20.7%), are also highly exposed to risk.
Taiwan’s statistics bureau also sharply raised its growth forecast for this year from 9.6% to 11%, reflecting recent AI demand, while warning that “if high-tech industries run into difficulties, the negative impact on the regional economy could be much greater than expected.”
[email protected] Yoon Jae-jun Reporter