Wall Street watches governor races that will help set rules for $2.5 trillion of AI infrastructure
Investors are tracking governor races, and voters will decide dozens of local measures that could delay or block data centers before they break ground.
By James Mitchell ·
Why it matters: Restrictions on unbuilt data centers could delay projects and put loans or bonds at risk, with smaller AI players more exposed than hyperscalers.
For information only, not investment advice. Prices delayed.
Tuesday's midterms carry a line item most campaign ads skip: the AI build-out. Investors are closely watching a few key races in states where data centers are planned but construction hasn't started, Axios reported, and Wall Street firms are tracking them down to the makeup of local utility commission boards.
The governor races matter most, Axios reported, citing Morgan Stanley's Salvatore, because state leaders have the most explicit and tangible power over building. Ten of those races will help set the rules for 67.4 gigawatts of planned data centers, representing $2.5 trillion of AI infrastructure investment, according to an October analysis by research firm Cleanview that Axios cited.
Texas first
Texas, home to 25% of planned data center build, is the state to watch, according to Axios. Democrat Gina Hinojosa, who is challenging Republican Gov. Greg Abbott, supports a moratorium on new data centers until tougher rules are in place. Abbott has imposed a limited pause on permits, though critics have documented permits still moving forward, Axios reported.
Then the ballot measures
Below the statehouse, voters in 38 local jurisdictions across 10 states will consider at least 41 measures that would delay, prohibit or add requirements for data centers, Reuters reported. In Sunbury, near Columbus, a measure banning data centers that use more than 25 megawatts could derail Amazon's ($AMZN) proposed $2 billion development, according to Reuters. An Amazon spokesperson declined to tell Reuters whether the measures would affect its expansion plans.
The measures could create headaches for hyperscalers like Meta Platforms ($META) and developers such as Applied Digital ($APLD), Bloomberg reported. In Piqua, where Meta is trying to build a data center, voters will also decide whether to recall the mayor and two city commissioners who have supported AI infrastructure there. Local resistance is already measurable: at least 45 projects worth $68 billion were disrupted by local opposition in the second quarter, according to a Data Center Watch report cited by Bloomberg, and a Quinnipiac poll in September found 72% of voters would oppose a new data center in their community, up from 65% in March.
What changes for your portfolio
Morgan Stanley projects that a Democratic midterm sweep, especially alongside wins for anti-data-center challengers at the state and local level, would be the most negative outcome for AI investor sentiment, Axios reported. Even that wouldn't necessarily mean a broad stock sell-off or an end to AI spending, Axios added.
The risks are project by project, according to Axios: new restrictions on a site that hasn't broken ground can bring delays that put loans or bonds at risk. Smaller AI players, particularly neoclouds already seeing higher costs and investor pullback, are more exposed than the hyperscalers. Morgan Stanley also notes slower construction could reduce borrowing needs and help credit markets.
There's a counterweight, too. If Ohio voters approve restrictions, developers are likely to steer projects elsewhere rather than abandon them, Baron Fung of Dell'Oro Group told Reuters. US data center capacity under construction rose 48% from December 2025 to June 2026, according to Cushman & Wakefield's John McWilliams, Reuters reported. The votes are Tuesday, and so far the only results anyone has are projections.
Sources
- Why Wall Street is watching governor races, Axios
- AI data center boom faces backlash at US ballot box, Reuters (via Investing.com)
- Data center restrictions on the ballot in dozens of U.S. cities, Bloomberg (via The Seattle Times)
Tags: elections · ai · big-tech · energy