Data center limits could prevent an AI investment bubble, Deutsche Bank economist says

Deutsche Bank economist Brett Ryan told Bloomberg on November 5 that regulatory limits on data-center construction could reduce the risk of an AI investment bubble, after this week's US midterm elections shifted the political landscape in states that hold nearly half of planned US capacity. Goldman Sachs estimates cited by Bloomberg put about 45% of planned data-center capacity in states where newly elected Democrats campaigned on limits, with Texas alone accounting for 25%.

"If some of these moratoriums or regulatory pressures slow some of this buildout, it almost prevents you from blowing too big of a bubble," Ryan said. He added: "This is like a good excuse to kill some projects here that were already looking shaky."

Bloomberg reported that monthly construction spending on data centers has almost doubled over the last 12 months, reaching nearly $100 billion on an annualized basis. Capital expenditures on information processing equipment and software have given the US economy the biggest boost on record over the last year, eclipsing their contribution at the peak of the 1990s dot-com boom.

Where the political pressure is concentrated

Texas Governor-elect Gina Hinojosa, the first Democrat to win the office since 1990, vowed during the campaign to halt new data-center approvals. Keisha Lance Bottoms campaigned on a statewide pause in Georgia, and Amy Acton proposed stricter conditions on new projects in Ohio. Sarah Bianchi, a strategist at Evercore ISI, told Bloomberg she expects state moratoriums to last no more than a year.

Ryan said Deutsche Bank had already been projecting slower investment growth by 2028. After Tuesday's elections, he told Bloomberg, "my knee-jerk reaction would be, OK, I don't have to raise my 2028 forecast anytime soon." Economists do not foresee an immediate pullback, since much of the capacity coming online next year is already under construction. The election results matter more for the 2028 outlook if groundbreaking on new facilities slows.

How the economics split

Anna Wong of Bloomberg Economics wrote that slower data-center construction could deepen financing stress for leveraged developers and neocloud companies, even as tighter computing capacity preserves hyperscalers' pricing power. Aichi Amemiya, an economist at Nomura Securities, told Bloomberg the midterm results may act as a "natural brake on the AI investment boom, which could help prevent the potential bust."

Amazon Web Services CEO Matt Garman told The Information on November 7 that Amazon is "in a pretty good spot" on data center expansion after the election, adding that "others may have some more work to do."

For scale, ING estimated that AI drives about a third of US GDP growth in 2026. Texas, at 25% of planned capacity, represents roughly 56% of the 45% that sits in states where winning Democrats campaigned on limits.

Why this matters for executives and strategy, real estate and construction, and finance

For developers and construction firms, the near-doubling of monthly data-center spending to a nearly $100 billion annualized pace signals how much capital is already committed - and how exposed the pipeline is to a permission shift. Projects not yet breaking ground face a new regulatory gate in states representing 45% of planned capacity. For finance and strategy executives, the signal from Deutsche Bank and Nomura is that supply constraints arriving through regulation, rather than chips or power, may function as a circuit breaker on overinvestment. The constraint has rotated from what can be built to what will be permitted.