The changing political winds around data centers could buffet the entire US economy, reduce investment in artificial intelligence and slow growth, according to Wall Street analysts.
The historic surge of money poured into AI, from data centers and frontier models to cutting-edge chips and network equipment, has boosted business investment and financial markets in recent years. Federal Reserve researchers recently estimated that spending on data centers, power plants and other AI-related tech equipment accounted for about 20% of US gross domestic product in the previous year.
In 2026, spending by “hyperscalers” such as Alphabet, Amazon, Meta and Microsoft – the companies racing to build data centers – should double to around $800 billion and is expected to reach $1 trillion in 2027, according to Capital Economics. Although the extension of this expansion could serve the public interest, given the widespread opposition to data centers, the crimping of that flow of dollars will have economic effects.
“Our economy is being run on this idea that AI is going to make us rich,” economist Allison Schrager with the Manhattan Institute, a nonpartisan think tank, told CBS News. “There are risks to slowing it down – it will mean slower growth.”
“Bad Effects”
Economists and investors now expect the Democrats’ big midterm election gains to both slow and increase the cost of data center development across the United States.
As a result, investment in data centers is likely to continue in the near term, both for projects underway and for facilities that have received approvals.
For companies investing in AI, the bigger issue will be continuing such capital spending at the same pace, even if the productivity gains and profits that AI is expected to deliver remain largely unrealized for now.
“The high-level calculations necessary to justify that level of CapEx, if sustainable, we are talking about $2 trillion in revenue for large AI companies by the early 2030s, which seems very unlikely,” Capital Economics Chief North America Economist Stephen Brown told CBS News. “Our view is that this will probably cause some delay in data center construction. We’re not saying it will stop entirely, but the level of investment will probably ease. And that would have some detrimental effects on the US economy.”
Impact on stocks and consumer spending
Moody’s Analytics chief economist Mark Zandi thinks investors will be willing to continue pouring money into AI for years to come, but said the long-term outlook is less clear.
“There are a lot of questions about whether that capital will continue to fill that hole and, if not, AI companies will have to raise prices to cover costs. Then you won’t see the same kind of investment,” he said.
In 2025, investment in AI will add about 0.3 percentage points to real GDP growth, according to Capital Economics, which expects investment in the sector to boost growth by about the same amount this year.
If AI investment were to fall in 2028, “It could become an immediate drag on GDP,” Capital Economics chief economic adviser Vicky Redwood said in a research note on Friday, estimating that a 10% annual decline in AI spending from current levels would knock about 0.1% of GDP growth.
If investment in AI slows, the biggest boost to growth could come from reduced consumption, Capital Economics’ Brown told CBS News. High-income Americans who invested in AI stocks saw their wealth rise sharply. With the S&P 500 up nearly 100% over the past three years, such investors have driven increased spending.
“A large part of the wealth effect comes from the strong rise in stock prices, and if data center construction slows, the possible weakness we will see in the stock market is a sign of things to come,” said Brown, who estimates that this “wealth effect” will increase economic growth by about 0.5% this year.
If the S&P 500 lost 20% of its value, that boost would translate into a 0.5% drop in GDP growth, Brown said. “That would be a 1% hit to GDP,” he said. “We would be talking about a sharp slowdown in GDP growth versus a recession.”