Artificial intelligence investment is helping to keep the world economy growing above 3 per cent, but the benefits are becoming concentrated in fewer countries as war, high energy costs and rising debt threaten the outlook, the International Monetary Fund has warned.
In its latest World Economic Outlook, released on Tuesday during the IMF and World Bank Annual Meetings in Bangkok, the Fund projected global economic growth of 3.1 per cent in 2026. That is slightly higher than its 3.0 per cent estimate in July, although still below the 3.5 per cent expansion recorded in 2025.
Growth is forecast to improve to 3.4 per cent in 2027, unchanged from the IMF’s July projection. The relatively resilient headline numbers, however, conceal a widening gap between economies linked to the AI investment boom and those facing the full force of expensive energy, weak demand and heavier borrowing costs.
AI infrastructure becomes an economic engine
The IMF said sustained spending on data centres, computing hardware and other AI infrastructure was a major reason the global forecast had improved. Economies building AI platforms, as well as countries supplying chips, electronics and other equipment, have received some of the clearest upgrades.
South Korea, Malaysia, Mexico and Thailand were among the beneficiaries identified by the Fund. The United States is expected to grow by 2.3 per cent in both 2026 and 2027, supported by household spending, tax cuts and AI investment. India is forecast to expand by 7.0 per cent in each year, maintaining its position as the fastest growing major economy.
The IMF also raised its 2026 forecasts for the euro area and Britain, partly because of stronger investment connected to AI and, in Europe, increased defence spending. China is expected to grow by 4.5 per cent this year and 4.0 per cent in 2027, slightly below the Fund’s July estimates as higher oil prices deepen pressure on already weak domestic demand.
IMF chief economist Silvana Tenreyro said: “The global economy has been resilient, but that resilience is under strain, with risks to growth tilted to the downside.”
The warning matters because the AI boom is no longer only a technology sector story. Spending on servers, power systems, data centres and advanced manufacturing is now large enough to influence national growth forecasts, trade flows and inflation. It is also creating a new division between countries able to supply or adopt the technology and those largely outside its value chain.
Fewer countries are driving global expansion
Only 32 countries now account for 95 per cent of global growth, according to the IMF, compared with 54 countries in its April assessment. The Fund lowered its forecasts for 47 per cent of economies over the same period, with energy importers outside the AI supply chain among the most exposed.
That concentration is an important signal for developing economies. Countries without reliable electricity, affordable finance, digital infrastructure and specialised skills may struggle to capture the investment now supporting growth elsewhere. They could also face higher import bills and debt servicing costs at the same time.
For African policymakers and businesses, the outlook strengthens the case for investment in power, broadband, technical education, cloud capacity and regional digital markets. It also underlines a harder question: whether African economies will mainly consume imported AI services or build companies and infrastructure that allow them to participate in the new value chain.
War and oil prices threaten the forecast
The IMF’s outlook assumes an average oil price of $96 a barrel in 2026 and $87 in 2027. It also assumes that the Strait of Hormuz will reopen in March 2027 and that shipping conditions will broadly return to their pre-war pattern by October next year.
A more severe scenario would be substantially worse. If a deepening Middle East conflict pushed oil prices to $130 a barrel through the first quarter of 2027, the Fund estimates that global growth would be reduced by 0.2 percentage point in 2026 and 0.5 percentage point in 2027. Global inflation next year would rise by a further 1.1 percentage points to 5.2 per cent.
The Gulf economies have already suffered some of the sharpest downgrades. Saudi Arabia is now forecast to contract by 3.3 per cent in 2026, compared with growth of 1.7 per cent projected in July. The IMF also expects deep contractions in Iran, Iraq, Kuwait, Qatar and the United Arab Emirates.
Global headline inflation is forecast at 4.7 per cent in 2026 and 4.1 per cent in 2027. That combination of weak growth in vulnerable economies and persistent price pressure leaves central banks and governments with limited room to respond.
The AI boom carries its own risks
The Fund said faster adoption and investment could lift global output by a further 0.6 per cent by 2029, with the United States and China receiving most of the gains. But the reverse is also possible.
If expected productivity improvements fail to appear and companies sharply cut AI spending, the IMF estimates that United States output could fall directly by 0.8 per cent. The shock would spill into Asian technology exporters and financial markets. When weaker asset prices, tighter financing conditions and reduced consumption are included, global output could be 1.1 per cent lower by 2028.
Debt adds another layer of vulnerability. The IMF warned that elevated public borrowing, particularly in advanced economies, could drive financing costs higher and weaken growth. It urged governments to rebuild fiscal buffers and set out credible medium term plans for sustainable debt.
The central message is that AI has become a powerful support for the global economy, but not a guarantee of durable prosperity. Its benefits remain uneven, its profitability is still being tested and its infrastructure depends on vast amounts of capital and energy.
As finance leaders continue their meetings in Bangkok, governments and businesses should watch not only the headline growth rate, but also who is producing it and how exposed that growth is to war, debt and a sudden change in confidence around AI.
What do you think? Can African economies secure a meaningful place in the AI value chain, or will infrastructure and financing gaps widen the divide? Share your view with Podium News.
