Democratic Midterm Sweep Puts AI Data Centers Under New Scrutiny

Democrats won majorities in both the House and Senate in the midterm elections, and the shift could reshape how Washington and the states oversee artificial intelligence over the next two years.

The stakes are large. AI companies are in the middle of a massive buildout of data centers, and that construction has already drawn political scrutiny on several fronts: the electricity consumed by data centers, the impact of power demand on consumers and local communities, cybersecurity, and the technology's effects on the labor force.

President Donald Trump has made AI development a priority of his administration. He has rebranded AI as "superintelligence" in official White House policies. But Trump's authority has limits. He can veto legislation passed by a Democratic Congress, and he retains control of regulatory agencies in the executive branch — yet Democratic gains at the federal and state level mean companies building AI infrastructure will face a different oversight environment than they planned for.

The takeaway: divided power in Washington, combined with new Democratic power in the states, is the real story for AI investors and operators.

What are investors saying about the midterm results?

Strategists at State Street Investment Management were blunt about the pressure Democratic gains could create. Chief macro policy strategist Elliot Hentov and investment strategist Christine Norton wrote that "Democratic gains could create additional headwinds for… and increase scrutiny of AI providers and data center development, particularly the impact of power demand on local communities."

They pointed to the governor's races as a particular concern. "These concerns are heightened by some of the state gubernatorial results," Hentov and Norton added.

But they drew a clear line between state-level friction and federal action. "Yet at a federal level, congressional control alone would not be enough to curtail AI-related data center construction in the face of a presidential veto," they wrote.

The two strategists also pushed back on the idea that tighter scrutiny kills the investment cycle. "AI also remains a strategic priority for both parties, suggesting that greater policy scrutiny should not automatically be equated with an end to the AI capital spending cycle," they wrote. "AI investment will likely continue to be robust in 2027, albeit tempered by strains on the supply chain and tighter credit."

In plain terms: expect pressure, not a shutdown.

What does the Texas flip mean for data centers?

The most concrete numbers in the story come from UBS. Strategists at the bank flagged potential pauses on data center construction in states like Texas, where Democrats flipped the governorship and the state House, and said those pauses could affect the outlook for AI infrastructure and cloud computing companies.

The scale is significant:

• Data center pauses will slow the pace of buildout into 2028
• Roughly 30% of expected 2028 additions are now at risk of being impacted
• Texas alone accounts for 18 GW of that at-risk capacity

"Data center pauses will slow the pace of buildout into 2028 with ~30% of expected 2028 additions now at risk of being impacted, and Texas alone is 18 GW of that," the UBS strategists wrote. "Congress is less likely to go beyond the states in our view with data centers now more likely to pay for grid upgrades – which has the potential to raise costs."

That last point matters for the economics of the buildout. If data centers have to foot the bill for grid upgrades, the cost of AI infrastructure rises — and those costs eventually show up somewhere, whether in cloud prices or in corporate capital budgets.

How will semiconductor and AI infrastructure stocks be affected?

UBS offered a mixed picture for the companies that supply the AI buildout. "The slowdown is a negative for Semis and AI infrastructure, but a lot is already priced into Semis with valuations nearing multi-year lows," the strategists wrote. "Additionally, Semis positioning has been cut sharply."

Translation: the bad news is real, but investors have already sold a lot of it. Semiconductor valuations are near multi-year lows, and positioning — the bets investors hold — has already been reduced sharply, which limits how much further damage a slowdown can do to share prices.

What should AI companies expect from regulators?

Dallas Dolen, technology, media and telecom lead at PwC, told FOX Business that Democratic control of Congress brings a specific agenda of oversight topics. "We're likely to see greater scrutiny around AI safety, bias, labor, copyright and transparency, but more oversight doesn't necessarily mean more certainty," Dolen said.

That distinction between oversight and certainty is the heart of the problem for companies. More hearings and more rules don't automatically produce clearer rules.

Dolen said companies face a patchwork problem. "Companies will still be navigating an evolving patchwork of state requirements, and they can't put major AI investments on hold while the rules take shape," Dolen said. "They need a strong governance baseline that allows them to keep investing, building and scaling AI responsibly."

His prescription is practical: build internal governance now, so the same AI project can pass muster in different states with different rules.

Why the White House still holds most of the cards

For all the attention on Congress, the structural reality of the next two years limits what Democrats can do at the federal level. Trump sets AI policy through the executive branch, controls the regulatory agencies, and can veto any AI legislation that reaches his desk.

That's why State Street's Hentov and Norton argued that congressional control alone "would not be enough to curtail AI-related data center construction in the face of a presidential veto." It's also why the sharpest near-term changes are likely to come from governors and state legislatures — where Texas's flip puts 18 GW of potential capacity in question — rather than from Capitol Hill.

The likely equilibrium, according to the analysts quoted here: federal gridlock, state-level friction, and a national AI investment cycle that keeps running but gets more expensive as data centers shoulder grid upgrade costs and tighter credit weighs on 2027 spending.