The global economy is proving more resilient than many forecasters expected, but the headline recovery conceals a sharper divide. The International Monetary Fund now projects world growth of 3.1% in 2026 and 3.4% in 2027, lifting this year’s estimate by 0.1 percentage point from its July forecast. Yet growth was 3.5% in 2025, and the IMF says the expansion is increasingly dependent on a narrow group of technology-linked companies, sectors and countries.
That combination—moderate growth, rising inflation and concentrated investment—defines the economic story of the moment. Artificial intelligence is supporting demand for chips, data centers, electricity and capital goods, while war-related energy disruptions are raising costs for households and businesses. The result is not a uniform global slowdown, but a split economy in which the same shocks produce sharply different outcomes.
Two shocks moving in opposite directions
The IMF describes the current environment as the interaction of a negative supply shock and a positive demand shock. Higher energy and commodity prices associated with the war in the Middle East weaken purchasing power and raise production costs. At the same time, the AI investment cycle is increasing trade, construction and corporate spending in economies positioned inside its supply chain.
Global headline inflation is forecast to rise to 4.7% in 2026 before easing to 4.1% in 2027. That path complicates the choices facing central banks. Cutting interest rates could support demand, but doing so while energy prices remain elevated risks prolonging inflation. Keeping rates high may restrain price pressures, but it also increases borrowing costs for governments and companies already carrying substantial debt.
The IMF’s forecast assumes that oil and gas shipments through the Strait of Hormuz resume by March and that AI demand moderates rather than collapses. Those assumptions make the outlook particularly sensitive to events outside conventional economic policy. A prolonged disruption to energy flows would represent a larger supply shock; a sudden reversal in AI valuations could damage investment and financial markets at the same time.
Why the recovery is uneven
Advanced economies are projected to grow 2.1% in 2026, with Taiwan and South Korea benefiting from their positions in the AI supply chain. Countries with technology exports, favorable terms of trade or direct access to investment are receiving a cushion that is unavailable elsewhere.
Energy-importing and lower-income economies face the opposite pressure. They must pay more for fuel and food while receiving less benefit from the technology boom. The IMF expects the Middle East and Central Asia region, directly exposed to the conflict, to contract by 0.3% in 2026. For governments with limited fiscal space, subsidizing energy can protect households temporarily but may deepen debt vulnerabilities.
This division also has a political dimension. Wealthier countries can finance data centers, industrial subsidies and energy support, while poorer governments face trade-offs between development spending, debt service and emergency relief. If the AI boom generates productivity gains broadly, it could eventually raise incomes beyond the firms currently investing. If gains remain concentrated, the technology may instead reinforce existing differences in capital, skills and market access.
The competing interpretations
Optimists see AI as a new productivity cycle. Investment in computing infrastructure could lift efficiency, create new industries and eventually offset the costs of today’s energy shock. The IMF itself identifies faster-than-expected productivity gains as an upside risk.
Skeptics point to the concentration of the boom. A small number of companies and economies are driving a large share of investment, leaving markets vulnerable if expected returns fail to materialize. The IMF warns that an overly optimistic reassessment of AI could trigger financial corrections. Higher sovereign bond yields and renewed trade tensions add further pressure.
Neither view can yet be confirmed by the growth data alone. Investment is clearly supporting activity, but the longer-term productivity gains remain uncertain. The central question is whether AI becomes a broad economic capability or remains an expensive infrastructure race whose benefits accrue mainly to dominant firms and technology exporters.
What comes next
The near-term test is energy security. A sustained conflict-related disruption would push inflation higher and force central banks into a more difficult policy stance. The second test is financial: whether AI investment continues at its current pace without creating a damaging correction. The third is distributional: whether governments can expand access to skills, digital infrastructure and affordable energy.
For policymakers, the IMF’s message is less a forecast of crisis than a warning about narrowing room for error. Fiscal support must be targeted rather than indiscriminate; monetary policy must respond to second-round inflation without unnecessarily damaging demand; and trade restrictions that fragment technology and energy markets could worsen the divide.
The global economy may therefore avoid a synchronized downturn while still becoming more unequal and fragile. Its resilience depends on two uncertain propositions: that the war-related energy shock remains temporary and that AI investment produces real productivity rather than only higher valuations. If both hold, growth can broaden. If either fails, the apparent resilience of 2026 may look less like a stable recovery than a pause before a sharper adjustment.