IMF flags 'split global economy' as AI boom lifts chipmakers, squeezes energy importers
Synopsis
The IMF's October 2026 World Economic Outlook lays bare a fracturing global economy: Taiwan is growing at 10.7%, Korea at 3.3%, and Southeast Asia is riding a data centre wave — all powered by the AI chip boom. Meanwhile, energy-importing nations with no stake in the tech supply chain are being squeezed simultaneously by the Middle East war and surging hardware costs. India, notably, escapes both traps.
Key Takeaways
The IMF 's October 2026 World Economic Outlook describes a 'split global economy' driven by the intersection of the AI investment boom and the Middle East conflict .
Taiwan is forecast to grow 10.7% in 2026, up from 8.8% in 2025, on the back of advanced semiconductor exports.
Taiwan and Korea together supply 62% of global net trade in AI-related computing hardware.
China and the United States account for over 75% of net AI hardware imports in 2025 and 57% of new data centres since 2024.
India is excluded from the group of energy-importing emerging economies facing medium-term downgrades, with growth driven by domestic demand and services.
The IMF flagged a potential AI expectation bubble, warning that overoptimistic profit projections could trigger a sharp slowdown in AI infrastructure investment and a 'broad repricing' of equity markets.
The artificial intelligence investment surge is creating stark winners and losers across the global economy, with a handful of semiconductor-producing nations reaping outsized gains while energy-importing countries outside the technology supply chain face mounting pressure from higher import bills and the ongoing Middle East conflict. The International Monetary Fund (IMF) has described this divergence as 'a split global economy' in its October 2026 World Economic Outlook (WEO), released on 13 October 2026.
The Fund said the interplay of geopolitics and technology 'has produced a split global economy: Favorable terms of trade or technology-related exports and investment give some countries a cushion, whereas others face higher import bills and weaker demand.'
The Biggest Winners: Taiwan, Korea, and Southeast Asia
Taiwan stands as the clearest beneficiary. The IMF projected its economy to expand by 10.7% in 2026, accelerating from 8.8% in 2025, before moderating to 4.3% in 2027. The surge reflects Taiwan's dominant position in advanced semiconductor fabrication, which sits at the heart of the global AI buildout.
South Korea is forecast to grow at 3.3% in 2026, up sharply from 1.1% in 2025, 'buoyed by strong external demand for semiconductors, which supersedes the negative impact of the war,' according to the WEO. Vietnam and Malaysia are projected to grow at 8.4% and 5.2% respectively in 2026, with the IMF noting that 'several southeast Asian economies are projected to benefit from data center activity and the upturn in the global technology cycle.'
Notably, the AI hardware boom is highly concentrated. Taiwan and Korea together account for 62% of net global trade in AI-related computing hardware. On the demand side, China and the United States are responsible for more than 75% of net imports of AI-related computing hardware in 2025 and 57% of newly built data centres since 2024.
The Squeeze on Energy-Importing Economies
For countries outside this technology value chain, the outlook is considerably bleaker. The IMF warned that 'energy-importing economies with little connection to the technology value chain face the most difficult configuration: higher import costs and exchange rate pressures, weaker real incomes, more constrained fiscal space, and fewer offsetting demand impulses.' The Fund added that many low-income and vulnerable nations fall into this category, raising the risk that recent shocks 'will widen income gaps across and within countries.'
Net trade in AI hardware has risen 'exponentially' as a share of world GDP, the WEO noted, though it cautioned that this change has been 'driven by price effects rather than volumes' — a distinction that limits how broadly the boom translates into real economic activity.
India: A Domestic-Demand Exception
India emerges as a notable exception to the energy-importer squeeze. The IMF attributed India's growth trajectory to 'resilient domestic demand and robust services activity' rather than chip exports. Crucially, the Fund excluded India when it identified energy-importing emerging economies facing medium-term downgrades — a signal that India's growth model offers a degree of insulation from the current technology-geopolitics shock.
Productivity Doubts and Market Risks
Despite the investment frenzy, the IMF struck a cautious note on AI's productivity payoff. It acknowledged 'a gap between studies showing large productivity gains for specific tasks and the relatively modest effects observed in economy-wide data.' The Fund listed a potential reassessment of AI expectations among the key downside risks to its outlook.
'If expectations of AI-driven profitability and productivity gains prove overly optimistic, investment in AI-related infrastructure and technology could slow sharply,' the WEO warned. Equity markets with high concentrations of AI firms 'could be vulnerable to a broad repricing,' with the report pointing to recent volatility in technology stocks in Korea and the United States as an early signal. The IMF's longer-term prognosis links growth prospects to two pillars: readiness to adopt AI and the ability to reliably meet energy needs — a dual test that much of the developing world is not yet equipped to pass.
Point of View
Because the physical infrastructure of AI is extraordinarily concentrated. What the report quietly signals is that most of the Global South is on the wrong side of both the technology divide and the energy shock simultaneously. India's exclusion from the downgrade list is reassuring, but it reflects the resilience of domestic consumption rather than any meaningful integration into the AI hardware supply chain — a distinction that matters enormously for long-run competitiveness. The IMF's caution about productivity payoffs also deserves more attention than it will likely receive: the gap between task-level AI gains and economy-wide data is a polite way of saying the growth story is not yet proven at scale.
NationPress
13 Oct 2026
Frequently Asked Questions
What does the IMF mean by a 'split global economy' in its October 2026 report?
The IMF uses the phrase to describe a divergence where semiconductor-producing economies benefit from the AI investment boom while energy-importing nations outside the technology supply chain face higher import costs, weaker currencies, and slower growth. The Fund says geopolitics and technology together are driving this widening gap.
Which countries are the biggest winners from the AI boom, according to the IMF?
Taiwan leads with a projected growth rate of 10.7% in 2026, followed by Vietnam at 8.4%, Malaysia at 5.2%, and South Korea at 3.3%. The IMF credits strong external demand for semiconductors and data centre investment for their outperformance.
How does the IMF assess India's position in this split economy?
India is treated as an exception to the energy-importer squeeze. The IMF attributes India's growth to resilient domestic demand and robust services activity rather than chip exports, and explicitly excludes India from the group of energy-importing emerging economies facing medium-term downgrades.
What risks does the IMF flag around the AI investment boom?
The Fund warns that if AI-driven profitability and productivity expectations prove overly optimistic, investment in AI infrastructure could slow sharply. It also cautions that equity markets heavily concentrated in AI firms could face a broad repricing, pointing to recent tech-stock volatility in Korea and the United States as early warning signs.
How concentrated is global AI hardware production and consumption?
Extremely concentrated. Taiwan and Korea together account for 62% of net global trade in AI-related computing hardware. On the demand side, China and the United States represent more than 75% of net AI hardware imports in 2025 and 57% of newly built data centres since 2024.
