The global economy has proved more resilient than many forecasters expected, but the International Monetary Fund’s latest outlook offers little reason for complacency. The IMF raised its 2026 growth forecast to 3.1%, from 3.0% in July, while projecting a slowdown from 3.5% in 2025. It expects growth to improve to 3.4% in 2027, but the headline figures conceal a widening gap between economies benefiting from artificial-intelligence investment and those absorbing higher energy and borrowing costs.
The central force behind the upgrade is an investment cycle linked to AI. Spending on computing infrastructure, semiconductors and data centres has created what the IMF describes as a demand impulse, helping activity withstand trade tensions and the economic effects of conflict in the Middle East. Economies integrated into the technology supply chain—including Taiwan and South Korea—are expected to outperform many peers. The United States and other advanced economies are also receiving support from strong investment and comparatively resilient domestic demand.
That resilience has changed the immediate policy debate. Instead of preparing for a synchronized global downturn, governments and central banks must manage a more complicated combination of moderate growth, renewed inflationary pressure and concentrated financial enthusiasm. The IMF forecasts global inflation at 4.7% in 2026, before easing to 4.1% in 2027. Rising oil prices, linked to the war involving Iran and its regional effects, are adding to the pressure.
A recovery with different winners
The benefits of the current expansion are not evenly distributed. Advanced economies are forecast to grow 2.1% in 2026, while emerging-market and developing economies are expected to slow to 3.7%, down from 4.5% in 2025. Countries that import large quantities of oil and gas face a direct squeeze on household incomes, public finances and trade balances. The Middle East and Central Asia region is projected to contract by 0.3% in 2026, according to the IMF.
China illustrates another constraint. Its economy is expected to grow 4.5% this year, but weak domestic demand remains a drag. That matters beyond China: slower consumption in the world’s second-largest economy can weigh on exporters, commodity producers and manufacturing supply chains even when technology investment elsewhere is accelerating.
Supporters of the AI boom argue that the investment could eventually lift productivity, wages and living standards. If businesses use automation to produce more with fewer resources, the current capital spending surge could become a durable source of growth rather than a temporary market phenomenon. The technology also offers governments a potential way to offset ageing populations and labour shortages.
Critics point to the concentration of the gains. A small group of companies and countries controls much of the advanced chip, cloud-computing and data-centre ecosystem. The IMF warns that the boom may be vulnerable if productivity gains fail to materialize. In that scenario, US equities could fall by 20%, while global growth could be 1.1 percentage points below its baseline by 2028; the effect on Asia could be larger because of the region’s exposure to the technology supply chain.
Why energy and debt matter
The energy shock creates a second test. The IMF’s forecasts assume oil prices averaging about $96 a barrel in 2026, nearly 41% higher than the previous year. Higher energy costs can spread quickly through transport, food production and industrial goods, forcing central banks to choose between tolerating inflation and keeping interest rates high enough to restrain demand.
That choice is especially difficult for heavily indebted governments. Higher rates increase debt-servicing costs just as states face pressure to subsidize energy, protect households and invest in defence or strategic industries. Poorer countries have less fiscal room to respond, and the IMF has warned that the apparent global resilience may be nearing its limits.
There is a more optimistic interpretation: if conflict eases, energy prices retreat and AI productivity becomes visible in business data, the current expansion could broaden. The IMF’s 2027 forecast assumes some improvement, and falling inflation would give central banks greater freedom to reduce rates.
What comes next
The next phase will depend less on the headline growth rate than on whether investment produces widely shared gains. Policymakers face three priorities: preserving competition in AI markets, directing infrastructure and skills spending toward economies outside the leading technology hubs, and protecting vulnerable households without permanently subsidizing high energy consumption.
Investors will watch whether corporate earnings justify the valuation of AI-linked companies. Central banks will track oil-driven inflation against signs of weaker demand. Governments, meanwhile, must decide whether to treat AI as a private-sector boom or as strategic infrastructure requiring public oversight.
The IMF’s message is therefore neither a declaration of recovery nor a forecast of imminent crisis. It is a warning about dependence: global growth is holding up partly because one powerful investment cycle is compensating for war, energy shocks and weak demand elsewhere. If that cycle delivers productivity, the economy may broaden. If it disappoints, the world will discover how narrow its resilience has become.