United States stocks plunged on Thursday, 5 November, as investors dumped AI stocks, and market watchers said the AI concerns behind the selloff go beyond a single bad day. The Dow Jones Industrial Average fell 1,833 points, or 3.44 percent, while the Nasdaq Composite plunged 5.52 percent in its worst session since April 2025. The S&P 500 dropped 4.36 percent.
Behind the rout are growing AI concerns about the massive spending and borrowing that fund the AI boom, and whether those investments will pay off. Tech giants have poured billions of dollars into the infrastructure that powers artificial intelligence, and much of that buildout is financed with debt taken on at rising interest rates.
“Debt,” said Paul Mueller, an economist with the American Institute for Economic Research, when asked what Wall Street is most worried about when it comes to AI. “Wall Street is worried about the enormous amounts of debt obligations being issued by tech companies to build data centers, both because interest rates have been rising and because there is greater political uncertainty around how much opposition data center construction will face,” Mueller told Fox Business.
The debt behind the AI concerns
The borrowing sits at the center of the AI concerns. Hyperscale cloud providers have committed to data center projects worth hundreds of billions of dollars, and some of the smaller firms in the supply chain, known as neoclouds, have borrowed heavily to buy chips and rent computing capacity. Amazon’s trailing free cash flow has turned negative, a figure that has been cited repeatedly in recent market commentary. JPMorgan estimates global AI-related spending at about $900 billion this year, with cumulative spending reaching about $5.5 trillion by 2030.
Investors had been willing to fund that spending while share prices kept rising, and the AI concerns stayed in the background. The Nasdaq Composite had set a record close on 6 October, and chip stocks had delivered large gains through the first half of the year.
Adam Patti, chief executive of VistaShares, said investors are confronting three concerns at once: political uncertainty, return on investment, and a crowded AI trade. “The first is political: the midterm results, including wins in several key data center states, raised fears that the buildout will face more regulation, slower permitting and pushback over power bills,” Patti said. “The second is return on investment, with investors asking how long hyperscalers can keep spending at this pace while free cash flow goes negative. The third is positioning. Semiconductors were the most crowded trade in the market, and crowded trades unwind violently.”
Power and permits feed the AI concerns
Power is another strand of the AI concerns. The electricity needs of AI data centers raise questions about grid capacity, utility costs and how quickly new facilities can come online. “Power is the real bottleneck in AI today, more than chips, and it deserves investor focus,” Patti said.
The politics sharpened after Tuesday’s midterm elections. Democrats won control of Congress and governorships in Texas, Ohio and Georgia, states where large numbers of new AI data center projects are planned. Gina Hinojosa, the Democratic winner of the Texas governor’s race, pledged an immediate moratorium on approvals for new data center construction during the campaign. She said data centers are draining residents’ drinking water and driving up electricity bills.
Voters also weighed in directly. Residents voted against data center development in 40 of 41 referendums held across 10 states, according to Ballotpedia data cited by Politico and CNBC. Goldman Sachs estimated that about 45 percent of planned United States data center expansion capacity could be affected by the Democratic election victory.
Meta president Dina Powell McCormick told a Wall Street Journal technology conference on Wednesday night that building data centers is “going to be harder” after the elections, CNBC reported.
What happens next with the AI concerns
The mix of the election result and disappointing tech earnings provided the spark.
Jakub Rochlitz, a market analyst at eToro, said the AI concerns mean the market now expects the buildout to slow. “The market’s interpretation is that the AI infrastructure buildout will become slower and more expensive, which is rewriting growth expectations across the whole supply chain,” Rochlitz said. He noted that rising electricity prices had turned the rapid construction of power-hungry data centers into a political issue during the campaign, with some Democrats calling for restrictions or pauses on new projects.
The AI concerns have not yet become a verdict on the technology. Demand for AI computing remains strong, and the largest cloud providers say customers still want more capacity. The question investors are asking is narrower: how much of the planned spending will survive higher borrowing costs, slower permits and a new political climate, and what returns it will produce.