AI infrastructure has changed from a “technological narrative” to a “vote poison”! Wall Street warns in unison: AI deals face political risks in the midterm elections

The Zhitong Finance App learned that the physical framework of artificial intelligence, the data center, is changing from the grand blueprint of Silicon Valley to a “giant electricity bill” at the doorstep of American voters. Barclays Bank's stock tactical strategy team officially included this risk in its market outlook on Tuesday, warning investors not to assume “rapid AI application growth and a relaxed political environment can coexist for a long time.” Evercore ISI and BCA Research have also previously warned that a populist backlash against AI could cause trouble for the stock market. Bank of America also stated bluntly that if the Democratic Party wins the Senate and the post of governor of Texas in the midterm elections, US stocks may fall by more than 10% next year.

“Data center construction is turning AI from an abstract tech story to a specific cost of living issue,” Barclays strategists Jenny Yang and Alex Altmann wrote in the report. Even voters who have never used ChatGPT can't escape the consequences of data centers — higher electricity bills, increasing pressure on water resources, and rising industrial facilities within the community.

While Wall Street is still debating the valuation and capital expenditure of AI chip stocks, a political storm from the American hinterland is quietly changing the fundamental logic of AI investment. From Pennsylvania to Texas, and from New York to Florida, the boom in AI data center construction is facing an unprecedented bipartisan boycott — and this boycott is becoming the most lethal political issue in the 2026 midterm elections.

This AI-induced “Neighbor Avoidance Campaign” (NIMBY) is becoming an “invisible bomb” that will tear the two parties apart in the 2026 midterm elections.

From “Neighborhood Avoidance Effect” to “Voter Poison”: The Political Nature of AI Infrastructure Has Completely Changed

Wall Street's concerns did not come out of nowhere; poll data clearly revealed a sharp shift in voters' sentiment. According to a survey released by Gallup in May, about 71% of Americans oppose building AI data centers in their own regions. A Fox News poll in July showed that 70% of voters opposed the construction of data centers to support AI in their constituencies, while 78% were in favor of slowing down the pace of construction. Opposition crossed party lines — 69% for Democrats, 54% for Republicans, and 53% for non-parties. In the first quarter of 2026 alone, 75 data center projects in the US with a total investment of about 130 billion US dollars faced varying degrees of community opposition.

The core cause of voters' opposition is rising electricity bills. Research data shows that the construction of data centers will cause local residents' electricity bills to rise by an average of about 18%. Due to the huge electricity consumption of AI data centers, the highest wholesale electricity price in the US has risen nearly 2.7 times. Against the backdrop of continuing inflationary pressure, this is tantamount to adding fuel to the fire of ordinary households.

This anger is rapidly organizing. In July of this year, opponents of data centers held 142 protests in 42 states across the US. From the east coast to the west coast, the “heat” of data centers is shifting from the tech sector to voter forums.

Barclays strategists Jenny Yang and Alex Altmann broke the essence in their report to clients: “Data center construction is turning AI from an abstract tech story to a specific cost of living issue.” “Even voters with limited access to AI will be affected by higher electricity bills, water pressure, and the construction of community industrial facilities.”

The July report pointed out that opposition to AI data centers has become “one of the few topics capable of mobilizing voters across the political spectrum,” and politicians from both parties are trying to adapt to this new reality.

States Regulate “Avalanche”: From New York Ban to Pennsylvania's “Strictest” Rules

Opposition at the political level quickly translated into substantial regulatory tightening.

In mid-July, New York became the first state in the US to put the “brakes” on hyperscale data centers at the state level — Governor Hochul signed an executive order suspending state-level environmental approvals for large data centers above 50 megawatts. This executive order is widely regarded as the first substantial restriction on AI infrastructure in the US.

Florida passed legislation explicitly requiring that additional grid costs for large data centers not be passed on to ordinary residents and small businesses.

Michigan Republican Senate candidate Mike Rogers publicly called for a one-year suspension of data center construction.

Ohio has suspended data center tax relief. Senate Democrats also proposed a draft to end the federal tax incentives currently enjoyed by data centers.

Meanwhile, the AI data center controversy is deeply embedded in the political landscape of the midterm elections. Texas and Pennsylvania became two key battlegrounds in this game.

On August 18, Pennsylvania introduced the nation's “strictest” data center regulations. Governor Shapiro signed an executive order requiring data center developers to resolve electricity supply on their own, bear related electricity costs, and gradually increase the proportion of clean energy used; at the same time, they must also “respect local communities, be transparent with the community, and prioritize local recruitment.” Developers must sign a “legally binding consent order,” and failure to comply will face strict penalties.

Texas Governor Abbott ordered an audit of all data center projects applying for grid access this month. Jefferies analysts characterized the move as a “chilling signal” for electricity stocks. Abbott also publicly advocated that large data centers should bear their own related infrastructure costs, increase electricity supply, and recycle water resources.

Texas is one of the states with the largest number of operating and planned data centers in the US, and is also the traditional ticket office of the Republican Party. Bank of America's chief investment strategist Michael Hartnett's team placed the Texas gubernatorial election at the center of market predictions. The election campaign between current Republican Governor Greg Abbott and Democratic challenger Gina Hinojosa is viewed as a referendum: on the one hand, there are demands from tech giants to expand AI infrastructure, and on the other, voters' deep concerns about rising electricity costs and the impact on the community.

The Hartnett team warned that a substantial policy tightening on data center construction in Texas would send a clear signal that anti-AI infrastructure sentiment has transcended party boundaries and become a cross-party political consensus. Abbott ordered an audit of all data center projects applying for grid access this month, a move characterized by Jefferies analysts as a “chilling signal” for power stocks.

Wall Street's “political pricing”: Barclays, Bank of America, and Evercore collectively sound the alarm

Wall Street is incorporating this political risk into investment frameworks at an unprecedented rate.

Barclays Bank is the latest institution to join the warning. Its strategy team notes that the bank's customized AI data center index — which includes more than 40 stocks such as Ultra Micro Computer (SMCI.US), Arista Networks (ANET.US), and Microsoft (MSFT.US) — has begun to reflect this risk. Barclays believes that regardless of the midterm election results, AI trading lacks a new upward catalyst.

Hartnett's team placed the Texas gubernatorial election at the heart of their market predictions. Hartnett warned that if the Democratic Party simultaneously controls the Senate and takes the position of governor of Texas, US stocks may fall by more than 10% next year, reaching the definition of a technical correction.

Evercore ISI and BCA Research have also previously warned that a populist backlash against AI could cause trouble for the stock market. The BCA's core conclusion is even more far-reaching: “A populist backlash against AI may lead to joint bipartisan push for regulatory legislation in 2027, but it is particularly likely to trigger large-scale tax increases after 2029.”

Trump's “Headwind”: When the President's AI Enthusiasm Encounters Collective Voters' Boycott

President Trump is a staunch supporter of AI and data centers, but his stance is becoming a political burden for Republican candidates.

Trump said in a radio interview last week that the community that rejected the data center “is making a mistake.” He has repeatedly emphasized that data centers bring “huge employment opportunities and tax revenue,” and framed the AI competition as a national security priority for competition against China.

However, this position is tearing apart within the Republican Party. According to Bloomberg's July report, the Republican candidate is distancing himself from Trump on the data center issue to improve the prospects for the midterm elections. The Senate Republican campaign has issued an internal memo to top AI companies warning that negative views on data centers are killing the party's chances of holding a key seat in Ohio.

Henrietta Treyz of investment advisory firm Veda Partners points out that “AI superleaders are facing huge risks” and that states' AI regulatory policies will provide a blueprint for federal-level legislation.