Opposition to data centers becomes a sensitive issue in the US midterm elections, Pennsylvania introduces the nation's "strictest" data center regulations

Grassroots opposition triggered by AI infrastructure construction is rapidly evolving into a core topic for the U.S. midterm elections, and Wall Street analysts have begun factoring this political risk into their stock market forecasting models.

Bank of America strategists pointed out that if the Republican Party performs strongly in the November midterm elections, U.S. stocks will rise. If the Democrats win the Senate and secure the Texas gubernatorial seat, U.S. stocks could drop by more than 10% next year.

Previously, Evercore ISI and BCA Research also warned that populist pushback from voters against AI could put pressure on the stock market.

This judgment is based on a substantial shift in state governments’ attitudes toward data center construction—moving from tax incentives to audit requirements and outright construction bans; a tightening regulatory trend is spreading across the United States.

On Tuesday, Pennsylvania introduced the nation’s “strictest” data center regulation, requiring developers to secure their own electricity supply, bear related electricity costs, and gradually increase their use of clean energy; meanwhile, they must “respect local communities, maintain transparency, and prioritize local hiring.”

At the market level, this policy uncertainty is already reshaping individual stock performance. Caterpillar’s stock price recently came under pressure due to expectations of slowing AI data center construction; meanwhile, power equipment companies GE Vernova and Bloom Energy have found new growth narratives due to rising self-supplied electricity demand for data centers.

The Philadelphia Semiconductor Index (SOX) closed down 5% on Tuesday, extending recent bouts of volatility, though the index is still up 72% year-to-date.

A team at Bank of America led by strategist Michael Hartnett has placed the Texas gubernatorial election at the core of its market forecasts.

In their analytical framework, the race between incumbent Greg Abbott and Democratic State Representative Gina Hinojosa is essentially a referendum—on one side is the tech sector’s drive for expanded AI infrastructure, and on the other is voter anxiety over energy prices, inflation pressure, and the impact of data centers on communities.

Texas is one of the states with the highest number of operational and planned data centers nationwide, and is a longstanding Republican stronghold. Hartnett’s team notes that if this Republican stronghold sees substantive policy tightening on data center construction, it would send a clear signal: anti-AI infrastructure sentiment has crossed party lines and become a bipartisan political consensus.

Henrietta Treyz, co-founder of consulting firm Veda Partners, said, “AI supermajors are facing major risks.”

She also pointed out that state-level AI regulation will serve as a blueprint for federal legislation. Although it is unlikely that the next Congress will pass broad legislation before the midterms, lawmakers are expected to take up the issue once campaigns are over, which could shift investor expectations regarding the trajectory of future regulation.

Pennsylvania Introduces Nation’s “Strictest” Data Center Regulatory Rules

Pennsylvania Governor Josh Shapiro signed an executive order requiring data center developers to meet strict criteria before proceeding with projects, including gaining community approval and banning nondisclosure agreements.

Shapiro stated that the executive order mandates developers to sign a “legally binding consent order that locks them into these GRID requirements and imposes strict penalties for non-compliance.”

The GRID requirements stipulate that developers must secure their own power supply, bear related electricity costs, gradually increase the proportion of clean energy used; they must also “respect local communities, maintain transparency with the community, prioritize local hiring, and provide a comprehensive community benefits agreement.”

Abbott ordered audits this month for all data center projects applying for grid access, a move that quickly triggered concerns about tighter regulation in the market.

Jefferies analyst Julien Dumoulin-Smith described the move in a client note as an “ominous signal” for power stocks, and noted that the audit order is the latest evidence of “escalating anti-data center sentiment in Texas and across the nation.”

He also noted that regulators are explicitly focusing on requiring new data centers to be self-powered (“BYOG”—Bring Your Own Generation) rather than relying on the public grid.

This policy shift has pressured some regulated utilities, and the market is already debating the outlook for companies such as American Electric Power Company Inc. and NRG Energy Inc.

New York State has gone even further, imposing a moratorium on new large data centers, directly impacting AI-boom-linked Caterpillar—whose power generation equipment previously saw soaring demand with the rush to build new data centers.

In contrast, a Morgan Stanley research team led by Michelle Weaver pointed out in a report this week that as demands for self-supplied power at data centers increase, GE Vernova and Bloom Energy are poised to benefit.

The team wrote, “As community resistance to data centers intensifies, policy action is gathering momentum, and political visibility of this issue is expected to rise further as the midterm elections approach.”

While policy uncertainty has increased market volatility in the short term, some market participants believe that a clear regulatory framework will ultimately provide a more solid foundation for long-term AI investment.

Aniket Shah, Jefferies’ global head of Washington, Sustainability, and Transformation Strategy, said that the AI infrastructure sector may continue to experience short-term price swings, but once the federal government intervenes to harmonize the current patchwork of state regulations, it will ultimately be positive for investors.

“It will become a regulated technology, just like any other mature technology,” he said. “That’s good for long-term AI investment, because if you want a future for a technology, it has to be regulated.”

As for the overall market, Wall Street has not turned bearish on AI trades. The S&P 500 index remains near record highs, and AI value-chain beneficiaries continue to lead market performance.

The Philadelphia Semiconductor Index is up 72% year-to-date; if sustained through the year, it would mark the best annual performance since 2003, with the capital spending uptrend to chipmakers showing no sign of a fundamental reversal yet.