The biggest threat to AI isn’t coming from Silicon Valley. It’s coming from voters.

The 2026 midterm elections delivered a decisive Democratic sweep, sending immediate shockwaves through the technology sector. As Democrats reclaimed Congress and secured major gubernatorial upsets, the massive companies powering the AI boom watched their stock prices stumble.

To understand why a political shift is putting so much pressure on AI stocks, look past the software and examine the physical foundation of artificial intelligence.

AI may live in the cloud. But its problems are firmly on Earth.

AI does not just live in the cloud. It requires massive, warehouse-sized data centres filled with tens of thousands of advanced computer chips. These facilities devour enormous amounts of electricity. Data centre electricity demand in the United States currently sits at around 176 terawatt-hours annually. To put that into perspective, that is roughly the amount of electricity the entire state of Texas consumes in a full year.

The AI revolution is running into a very human problem

Across the country, citizens are pushing back against the tech industry. Many cite noise pollution—caused by massive industrial cooling fans running at all hours—as well as land grabs and the severe strain on local utility grids. However, a significant source of the opposition is also the uncertainty AI creates.

This is not just a fringe movement. Surveys show that a wide majority of U.S. residents now oppose the construction of new AI data centres in their neighbourhoods.

Local communities are organising to block massive tech developments, and the resistance crosses traditional political lines. In Minnesota towns like Pine Island and Hermantown, political newcomers recently unseated incumbents by running almost entirely on anti-data centre platforms. Residents argue that data centres bring massive infrastructure headaches without the economic upside. While a new manufacturing plant might employ thousands of locals, a billion-dollar data centre often employs fewer than fifty full-time technicians once construction is complete.

The Texas rebellion that could derail AI’s expansion

This local frustration boiled over into the midterms, culminating in Texas. The state has been the epicentre of this infrastructure boom, attracting companies with cheap land and a historically deregulated energy market. Over a quarter of all planned AI infrastructure is based in Texas.
However, Texas voters are acutely sensitive to power grid stability. The state operates on its own independent power grid, which has repeatedly struggled to meet demand during extreme heat waves and severe winter storms. Residents have faced calls to turn up their thermostats to prevent rolling blackouts. Consequently, adding dozens of massive AI data centres—each consuming as much power as a small city—became a highly volatile political issue.

The election upended the tech industry’s expansion plans. Democrat Gina Hinojosa defeated Republican Governor Greg Abbott, marking the first time a Democrat will lead the state in over three decades. A core pillar of Hinojosa’s campaign was a promise to hit the pause button on new data centre construction. Her administration plans to freeze approvals until the state can implement stricter rules to protect the electricity grid and secure local water supplies in drought-prone counties.

Similar concerns were echoed in Georgia, where Governor-elect Keisha Lance Bottoms also won on a platform that included reassessing unchecked data centre expansion. For the AI industry, these proposed building moratoriums threaten the massive physical expansion required to build smarter, faster models.

When voters push back, AI stocks feel the pressure

Faced with the reality of building moratoriums, delayed approvals, and stricter environmental rules, investors backed away from AI infrastructure. The stock market reaction was swift. The fear is simple: If the tech industry cannot build the physical facilities required to train its next-generation algorithms, the entire AI boom stalls.

When the election results rolled in, companies tied to the physical build-out of artificial intelligence saw an immediate sell-off:

• CoreWeave, a specialised cloud provider for AI, fell roughly 7%.
• Oracle, deeply invested in cloud infrastructure and expanding its physical footprint, dropped 4.3%.
• Tesla and SpaceX slid between 5% and 6%.

Investors are now recognising a new reality.

The biggest hurdle for artificial intelligence is no longer software engineering or microchip design. It is local zoning laws, environmental permitting, and intense community resistance.

When money can’t buy the outcome Silicon Valley wants

The election results delivered a harsh blow to tech leaders who spent heavily to protect the AI boom. Over the past few years, powerful billionaires have tried to shape local and national politics to ensure a favourable, deregulated environment for their mega-projects.

Elon Musk was the most prominent figure in this group. According to campaign finance disclosures, Musk donated more than $158 million just in September to political action committees backing Republican candidates. His goal was to protect business-friendly policies in states like Texas, where he had deliberately relocated his core companies. Musk moved his operations to Texas precisely to take advantage of the state’s historically loose regulatory environment, low taxes, and independent power grid.

For the tech sector’s wealthiest advocates, the strategy was to spend heavily to keep zoning laws relaxed and grid access open. With Democrats reclaiming Congress and securing major state-level upsets, that strategy failed.

All is not lost, however. With his preferred candidates defeated and stricter regulations looming over his earthly infrastructure, Musk took to his social media platform, X, to highlight a possible advantage: “Data centres in space looking pretty appealing right now.”