US economy growth narrows around AI spending

The risk is not that AI is fading as a theme, but that the market and the economy have become too reliant on a small cluster of winners. Capital spending tied to data centers, chips and cloud infrastructure has lifted industrial activity and supported earnings at companies at the center of the build-out. Nvidia’s latest filings showed revenue more than doubled from a year earlier, while data center sales surged again, underscoring how much of the economy’s most dynamic growth is now tied to AI infrastructure. Microsoft has also seen its share price remain far above long-term trend measures despite recent volatility, reflecting investor conviction that AI can keep driving earnings.

But the breadth problem is visible elsewhere. The Institute for Supply Management’s manufacturing gauge has hovered near stagnation, and job openings have eased from the highs that followed the pandemic reopening, suggesting that the benefits of the expansion are not filtering evenly through the economy. Unemployment remains low at 4.2%, so there is no sign of a recessionary crack yet. The concern is more subtle: when a large part of growth comes from one investment boom, the cycle becomes vulnerable to any slowdown in that spending.

That is why the split between AI winners and losers matters economically. If the largest technology companies keep pouring cash into chips, servers and power capacity, the US can keep growing faster than other advanced economies. If that spending plateaus, the drag would be felt not only by semiconductor and software stocks, but by suppliers of machinery, electricity, construction and industrial metals. Copper demand, in particular, has become tied to the pace of AI build-outs, making the story global rather than just domestic.

For investors, the implication is straightforward: this is no longer just a valuation debate around AI stocks, but a macro concentration risk. The bull case is that productivity gains from AI justify the spending surge and broaden into corporate earnings across sectors. The bear case is that the current boom is front-loaded, with a narrow set of companies generating most of the growth while the rest of the economy remains soft, leaving equity markets vulnerable if capex slows or margins come under pressure.

That tension is likely to dominate the next phase of the cycle. As long as AI investment keeps rising, it can mask weakness elsewhere and keep the US expansion intact. But the more the economy depends on that one engine, the more dangerous any disappointment becomes.