The global economy is entering a more fragile phase: growth is holding up, but the forces sustaining it are unusually concentrated and increasingly vulnerable to shocks. The International Monetary Fund projects global output to expand by 3.1% in 2026 and 3.4% in 2027, while worldwide headline inflation rises to 4.7% this year before easing to 4.1% next year. Those figures suggest neither a global recession nor a clean recovery. They describe an economy being pulled in opposite directions by artificial-intelligence investment, war-related energy disruption and renewed trade friction.
The IMF’s October outlook identifies the central tension. A technology cycle led by AI is generating demand, investment and productivity expectations, while the war in the Middle East is producing a negative energy shock. Higher energy and food prices are lifting inflation just as governments and companies face more expensive borrowing. IMF Managing Director Kristalina Georgieva has warned that economic resilience is reaching its limits, with fiscal and energy buffers being depleted.
Why growth has not collapsed
AI investment has become a powerful source of global demand. Spending on data centers, semiconductors, software and electricity infrastructure is supporting activity in economies positioned to supply those systems. The IMF says countries benefiting from the AI boom or favorable terms of trade are performing relatively well, while many energy-importing and lower-income economies face a more difficult environment.
This helps explain why the current slowdown looks uneven rather than synchronized. Advanced technology producers and capital-rich economies can absorb higher financing costs more easily. Import-dependent countries, by contrast, are exposed to energy prices, currency pressures and tighter global credit. The apparent stability of the aggregate growth number therefore conceals a widening gap between winners and losers.
There is also a plausible optimistic interpretation. If AI tools raise productivity faster than expected, companies could produce more with the same labor and capital, improving wages, profits and public revenues. Faster productivity growth would make today’s investment boom more economically durable and could help governments manage high debt burdens.
The concentration problem
The downside is that the expansion is becoming dependent on a relatively small group of firms, industries and countries. The IMF cautions that technology-related investment is concentrating risks. That creates a familiar financial vulnerability: expectations can rise faster than underlying revenues, leaving markets exposed if projected returns fail to materialize.
Investors are already questioning whether AI companies’ future income will be sufficient to service the debt financing the construction of new infrastructure. Deloitte reports rising concern over AI-related bonds and warns that some specialized cloud providers may struggle to operate independently if capital expenditure greatly exceeds returns. These are not proof that an AI crash is imminent, but they show how a productive technology can also become a source of leverage and contagion.
“Risks surrounding AI are two-sided,” IMF research official Deniz Igan said, noting both the possibility of a market correction and the upside if productivity gains arrive sooner than expected.
The distinction matters. A correction in speculative valuations would not necessarily destroy the technology’s long-term value. But a disorderly retreat could hit semiconductor orders, data-center construction, commercial property, lenders and public finances at the same time. The more investment decisions rely on a narrow set of optimistic assumptions, the greater the risk that disappointment spreads beyond technology markets.
Trade and policy pressures
Trade tensions add another layer of uncertainty. The IMF expects world trade growth to slow sharply from 5% in 2025 to 3.5% in 2026 before recovering to 4.3% in 2027. Tariffs and the reorganization of supply chains may improve resilience for individual countries, but they also raise costs and reduce the efficiency that supported decades of globalization.
Governments face a difficult policy balance. Cutting interest rates too quickly could revive inflation, weaken currencies and encourage further asset-price excesses. Keeping rates high can stabilize prices but intensifies debt-service pressures, particularly for heavily indebted states and firms. Fiscal support may protect vulnerable households from energy shocks, yet broad subsidies risk worsening deficits and sustaining demand when inflation remains elevated.
There is no single solution for every economy. Technology-exporting countries can focus on competition, infrastructure and workforce training. Energy importers need diversified supplies and targeted assistance rather than indefinite blanket subsidies. Lower-income countries require access to affordable financing, because higher global yields can turn a temporary energy shock into a prolonged debt crisis.
What comes next
The next phase will be determined by whether AI productivity catches up with AI expectations, whether energy disruptions persist and whether trade disputes escalate. A benign outcome would combine falling inflation, continued technology investment and a gradual recovery in trade. A harsher one would involve prolonged energy inflation, higher bond yields and a sharp repricing of AI assets.
Policymakers cannot eliminate those risks, but they can reduce their reach. Stronger financial supervision, transparent corporate borrowing, targeted social protection and credible medium-term fiscal plans would make the economy less dependent on optimism. The central lesson of the IMF forecast is that resilience is real—but it is no longer a sufficient strategy. The world economy needs broader sources of growth before the next shock tests the narrow ones now carrying it.