The AI revolution has become a cornerstone of the US economy, with major players across industries investing heavily in AI infrastructure. But here's where it gets controversial: what happens if this boom slows down?
Data centres, the backbone of AI operations, are now critical to the nation's economic health. Economists are raising concerns about the potential impact of a slowdown in AI investment, especially as traditional economic drivers falter.
AI-related capital expenditure is significant, with hyperscale data centres leading the charge. The top four spenders - Microsoft, Amazon, Alphabet, and Meta - are expected to invest a whopping $344 billion this year, equivalent to 1.1% of US GDP. This investment is estimated to contribute a full percentage point to GDP growth in the first half of 2025, according to Barclays.
The level of dependence on AI infrastructure is causing alarm. Peter Berezin, Chief Global Strategist at BCA Research, believes the economy might already be in recession without the AI boom.
This investment surge has reshaped the US construction landscape. Data centres now make up around 35% of Turner Construction's project backlog, a sharp increase from 5 years ago. These projects require large workforces, with each data centre employing anywhere from 100 to 5,000 people.
The demand has also strained supply chains, with lead times for essential components extending by several months. Ben Kaplan, Managing Director of Turner Construction's advanced technology group, notes that every element of the supply chain is under pressure.
While much of the hardware is imported, especially chips from Nvidia, domestic infrastructure is also seeing major investment. Nvidia expects $65 billion in sales this year, surpassing forecasts.
The demand from AI and data centre operators has led companies like Oracle and CoreWeave to borrow aggressively. Oracle now holds over $100 billion in debt, having issued $18 billion in bonds to fund AI infrastructure. CoreWeave, which rents data centre space and compute power, is also taking on significant debt.
If revenue fails to match borrowing, it could create instability in credit markets.
High stock valuations add another layer of risk. A correction in tech stocks, triggered by concerns over an AI bubble, could have wider economic consequences. Last week, the S&P 500 index fell around 2% as investor confidence wavered.
Rising prices in AI equities have added $180 billion in consumer spending over the past year, according to JPMorgan Chase. A reversal could suppress consumer activity.
Jonathan Millar, Senior US Economist at Barclays, warns that a 20-30% correction in stock prices could cut GDP growth by 1-1.5% over a year. A pause or collapse in AI investment could further trim growth.
As the US economy becomes increasingly reliant on data centre development and AI infrastructure, the risks of over-dependence grow. Without sustained momentum, the effects could ripple across labor markets, consumer spending, credit, and overall growth.
And this is the part most people miss: the potential for a fragile labor market to be further destabilized by a sudden drop in investment. It's a delicate balance, and one that economists are watching closely.
