The global economy is entering a contradictory phase: growth is holding up better than expected, yet the foundations of that resilience are becoming narrower and more vulnerable. Artificial-intelligence investment is powering business spending and helping offset the shock from higher energy prices, while households continue to face persistent inflation and governments confront widening fiscal and geopolitical risks.
A Reuters poll of more than 500 economists found that respondents raised their 2027 inflation forecasts for 31 of 50 economies. The same survey projected global growth of 3% this year and 3.2% next year, an improvement on its previous estimates but still below the pace forecast by the International Monetary Fund. The message is not one of collapse; it is of expansion increasingly dependent on investment that has yet to deliver its full productivity payoff.
Why AI matters so much
The United States is at the center of this story. The Reuters poll forecasts US growth of 2.3% this year and 2.1% in 2027, with AI-related capital expenditure described as a principal driver of domestic activity. Spending on data centers, chips, networking equipment and electricity infrastructure has created demand well beyond the technology sector.
That investment has also helped economies absorb energy and security shocks. Companies are still ordering equipment, building computing capacity and positioning themselves for expected gains in automation and software. Morgan Stanley global chief economist Seth Carpenter told Reuters that the AI-led investment cycle and stronger-than-expected regional activity were supporting resilience despite higher energy prices.
But investment is not the same as productivity. The economic case for the boom depends on AI eventually raising output enough to justify its enormous costs. Until then, growth may reflect a construction and procurement cycle rather than a permanent improvement in living standards.
The inflation contradiction
Higher energy prices are complicating the transition. Reuters reported that wars involving Iran, Ukraine and Yemen have pushed oil and natural-gas prices higher as European ministers prepare for winter. Energy costs spread through transport, food, manufacturing and household heating, creating pressure that central banks cannot easily neutralize without weakening demand.
Recent US indicators illustrate the dilemma. S&P Global’s October flash survey showed composite business activity slowing to 57.6 from 58.4 in September, while new orders reached their highest level since March 2022. Companies reported sufficient demand to raise prices. That combination—solid activity alongside renewed pricing power—makes it harder for central banks to declare victory over inflation.
For policymakers, the choices are uncomfortable. Cutting interest rates too quickly could allow energy and services inflation to become entrenched. Keeping rates high for longer could damage housing, business investment and lower-income households, even as AI-heavy firms continue to attract capital.
The correction risk
The most serious threat is that investors begin to doubt whether AI revenues will match the scale of spending. The IMF has warned that a sharp fall in AI investment, caused by disappointment over productivity or profitability, could reduce global GDP by 1.1% by 2028 and US output by 1.8%.
That scenario would work through several channels: falling technology valuations, weaker corporate investment, reduced wealth and tighter financing conditions. The impact would not be limited to Silicon Valley. Suppliers, construction firms, utilities and regions dependent on data-center development could all be affected.
Supporters of the current spending wave argue that previous infrastructure booms were also criticized before their benefits became visible. The internet and mobile networks required years of investment before transforming commerce and productivity. On this view, temporary overcapacity would not invalidate the technology’s long-term value.
Skeptics counter that the present cycle is unusually concentrated and capital-intensive. If adoption lags behind expectations, the economy could be left with expensive infrastructure, high debt and fewer policy options because inflation limits central banks’ ability to respond.
What comes next
Near-term attention will turn to US employment and earnings. Reuters reported that September payroll growth was only 29,000, while unemployment stood at 4.2%; upcoming data will help determine whether the Federal Reserve sees weakness that warrants support or inflation that requires restraint.
The wider test is whether AI investment spreads into broad productivity gains before energy shocks and interest-rate pressure undermine demand. Governments can help by expanding electricity grids, training workers and ensuring competition, but they cannot guarantee commercial returns. They also face the political task of protecting households from high bills without permanently subsidizing inefficient energy use.
The likely outcome is neither an immediate global downturn nor a risk-free boom. Growth may remain resilient while becoming more unequal across sectors and countries. If AI delivers measurable productivity gains, today’s spending could prove transformational. If returns disappoint, the same investment that currently cushions the economy may become the mechanism that amplifies its next slowdown.