The global economy is still expanding, but its apparent resilience increasingly depends on a narrow and potentially unstable foundation: the artificial-intelligence investment boom. The International Monetary Fund expects world output to grow 3.1% in 2026 and 3.4% in 2027, revising this year’s forecast slightly upward from July. Yet the improvement comes as war-driven energy costs, record public debt and high interest rates leave governments with fewer tools to absorb another shock. The IMF’s October World Economic Outlook describes an economy that remains resilient but increasingly uneven.
The central tension is visible in the forces pulling growth in opposite directions. The Middle East war has created what IMF Managing Director Kristalina Georgieva called a negative energy-supply shock, raising oil and other commodity prices while weakening production. At the same time, spending on data centres, semiconductors and related infrastructure is generating a positive demand shock. The IMF says AI investment and trade helped activity hold up during the first half of the year, offsetting some of the damage from conflict and continuing trade tensions. IMF briefing transcript
Growth with a narrow base
That combination makes the headline forecast less reassuring than it first appears. AI investment is concentrated in a limited group of firms and economies, while higher energy prices affect households and businesses worldwide. Countries that produce energy or host technology supply chains may benefit from the current cycle; energy-importing and lower-income economies face a more difficult trade-off between supporting growth, protecting consumers and preserving fiscal stability.
The distributional consequences matter. Large technology companies can finance capital-intensive projects, while smaller firms may struggle with higher borrowing costs and uncertain access to computing capacity. Consumers could eventually benefit from productivity gains and cheaper services, but those gains depend on adoption, competition and the ability of workers to move into new roles. The IMF has estimated that AI, if deployed effectively, could add roughly half a percentage point to annual global growth, but that outcome is not automatic. Reuters
The bubble question
The most immediate financial risk is not that AI has no economic value. It is that investor expectations may outrun measurable profits. The IMF warns that disappointing productivity or profitability could trigger a sharp reversal in AI-related investment and asset prices. In one downside scenario, a steep fall in AI investment would reduce global GDP by 1.1% by 2028, including a 1.8% decline in US output. Reuters
Supporters of the boom would argue that this comparison understates the possible gains. AI infrastructure is already stimulating semiconductor production, construction, electricity demand and cross-border trade. If productivity improvements spread beyond the technology sector, stronger potential growth could make today’s investment look less speculative. The IMF’s forecast of 3.4% global growth in 2027 reflects some confidence that the expansion will broaden.
Sceptics counter that much of the spending is being justified by projections rather than established returns. A correction would not necessarily resemble the collapse of the entire technology sector, but it could expose highly leveraged companies, weaken equity markets and reduce tax revenues just as governments face rising debt-service costs.
Why policy space is shrinking
Debt magnifies the danger. The IMF says global public debt is at its highest level since the Second World War and is projected to exceed 100% of world GDP before 2030. Bond yields in advanced economies have risen to multi-decade highs, reflecting persistent inflation, heavy borrowing needs and expectations that AI could lift future growth. Higher yields raise refinancing costs and constrain public investment, particularly in countries without deep domestic capital markets.
Central banks therefore face a difficult balance. Cutting rates too quickly could reignite inflation if energy prices remain elevated or AI demand continues to push up investment and wages. Keeping rates high for too long could expose indebted governments, businesses and households. Fiscal authorities face a similar dilemma: broad energy subsidies can soften the immediate hit but worsen deficits, while abrupt withdrawal can intensify hardship.
What comes next
The next phase will be judged less by the global growth average than by whether investment produces broad productivity gains. Governments can reduce the risk of a disorderly correction by improving financial disclosure, monitoring leverage outside banks, coordinating energy policy and directing public support toward skills, grids and research rather than simply subsidising valuations.
The outlook also depends on the war’s duration and the path of oil prices. A prolonged conflict would reinforce inflation and further divide winners from losers. A de-escalation could ease the supply shock, but it would not remove the structural problems of debt, weak fiscal buffers and concentrated technology gains.
The IMF’s message is therefore neither that AI is a bubble nor that it will rescue the world economy. It is that the current expansion has become unusually dependent on an investment cycle whose benefits remain uncertain. If productivity follows spending, the boom could broaden growth. If expectations break first, the same concentration that now supports the economy could become its most exposed fault line.