The global economy is entering a contradictory phase: artificial-intelligence investment is supporting growth just as war-related energy disruption is reviving inflation. The International Monetary Fund’s latest outlook captures the tension. It projects global growth of 3.1% in 2026 and 3.4% in 2027, while headline inflation rises to 4.7% this year before easing to 4.1% next year. The figures suggest resilience, but not a broad-based recovery.
The central story is a collision between two shocks. On one side, demand for computing infrastructure, semiconductors, data centres and related services is creating a powerful investment cycle. On the other, the war in the Middle East has pushed up energy and food costs, particularly for countries that import fuel. IMF officials describe these forces as a “positive AI demand shock” and a “negative energy supply shock,” pulling economic activity in opposite directions.
Why growth has not collapsed
AI-related spending is concentrated in advanced economies and in Asian and North American technology supply chains, but its effects are spreading through trade, capital expenditure and financial markets. Companies are investing heavily in chips, electricity generation, cooling systems and cloud capacity. That spending has helped offset weaker activity elsewhere and raised hopes that productivity gains could eventually make the expansion more durable.
Yet investment is not the same as productivity. The IMF has warned that markets may be assuming faster returns from AI than businesses can deliver. If projected profits fail to materialise, highly valued technology companies and the suppliers around them could face a sharp correction. A market retreat would not necessarily cause a global recession, but it could reduce corporate investment and expose lenders and pension funds to losses.
“The outlook is becoming increasingly uneven across countries,” the IMF said, highlighting the contrast between economies benefiting from AI investment and those facing higher energy costs.
That unevenness matters because the benefits of the technology boom are not distributed evenly. Countries with semiconductor capacity, deep capital markets, reliable electricity and skilled labour are positioned to capture more value. Lower-income economies may instead bear the cost of pricier fuel and food while receiving fewer gains from the new investment cycle.
Inflation is the immediate constraint
The renewed rise in inflation complicates the response of central banks. Higher energy prices feed directly into transport, heating and manufacturing, then indirectly into wages and services. Policymakers cannot easily produce more fuel by raising interest rates, but they may still need to keep borrowing costs high enough to prevent a temporary energy shock from becoming entrenched.
The IMF says global headline inflation rose from 4.3% in March to 4.9% in May before easing to 4.7% in June. That partial retreat is reassuring, but prices remain well above the levels reached earlier in the year. For households, the key issue is not whether inflation is slowing, but whether food, rent, electricity and transport are becoming affordable again.
Higher rates also intensify pressure on public finances. Governments are being asked to subsidise vulnerable consumers, strengthen energy security, invest in defence and support industrial policy, all while refinancing debt at more expensive yields. The IMF has warned that fiscal and energy buffers are being depleted. Emerging markets face an especially difficult trade-off: supporting households may be necessary, but broad subsidies can worsen deficits and benefit richer consumers who use more energy.
What comes next
The near-term outlook depends heavily on the duration and geographic spread of the war-related disruption. A rapid stabilisation could allow energy prices to fall and inflation to resume its decline. A prolonged conflict, damage to transport routes or further supply interruptions would produce a harsher combination of slower growth and higher prices. In a severe scenario cited by the IMF, global growth could fall to about 2% while inflation rises above 6% by 2027.
Governments therefore face a narrow policy path. Targeted transfers are more defensible than economy-wide fuel subsidies; investment in grids, storage and domestic energy production can reduce vulnerability over time; and credible fiscal plans can limit pressure on interest rates. At the same time, policymakers must avoid allowing AI support to become a disguised transfer to already dominant firms without competition, labour protections or measurable public benefit.
The most plausible outcome is neither an AI-led boom nor a worldwide collapse, but a divided expansion. Technology-intensive economies may continue to grow relatively quickly, while energy importers and highly indebted countries struggle with squeezed households and limited policy space. Whether AI becomes a genuine productivity engine will depend on its diffusion beyond a small group of firms. Until then, the global economy remains resilient in aggregate—and increasingly fragile underneath.