AivexaNewsSearch
AI news for builders and product teamsChecked every hour

The backlash about data centers might really just be about electricity

Collected Nov 9, 2026

Reported from Ashburn, Virginia, in Loudoun County, this MIT Technology Review piece argues that the growing backlash against data centers is driven more by electricity than by the broader cultural anxieties the industry tends to cite. The reporter visited what is described as the world's largest data center hub, where concrete buildings line the major roads, during the week of the US midterms. The trip followed comments from OpenAI's chief global affairs officer, Chris Lehane, who wrote two days after the midterms that data centers have become a proxy for public unease about affordability, corporate power and people losing control of their futures.

The central claim is that the industry cannot escape the physical consequences of its power appetite. There are broadly two ways to supply an AI-scale facility: connect to the grid or generate electricity onsite. In Loudoun County, where buildout has been steadier than elsewhere, most facilities draw from the grid. One Equinix facility the reporter entered keeps large diesel generators for backup but buys power from Dominion Energy, much of whose electricity comes from natural gas. The average data center in this part of Virginia has a capacity of 54 megawatts, more than the power draw of Denver International Airport.

Elena Schlossberg, a prominent Virginia activist, rejects the culture-war framing, saying the fight was always about power, meaning electricity. The reporter allows that some protesters are animated by anti-tech populism, but argues the grid consequences have shaped the backlash most. Recent protests produced new state rules requiring data centers to pay in full for new transmission lines rather than spreading the cost across other ratepayers. Developers say they are preparing for rules that would cut them off from grid power at peak demand. Schlossberg also worries that a strained grid will push operators toward more diesel generator use, or push Dominion toward more fossil generation instead of renewables.

Most AI data centers being built elsewhere look different. The desired scale pushes companies into rural areas with little grid infrastructure, so they bring their own power, known as behind-the-meter generation, typically natural-gas turbines plus diesel backup. Anthropic has reportedly agreed to pay $3.4 billion for gas turbines, described as the largest power generation purchase by an AI company. Grids would struggle to serve the demand anyway: the smallest facilities attractive for AI training are 300 megawatts, and the ideal is closer to 6,000 megawatts, roughly the estimated draw of Chicago.

Onsite generation may look like a way to spare the grid, but at this scale it affects surrounding communities, which have traditionally expected a say when utilities build infrastructure of comparable size. Gas turbines emit air pollutants and diesel backup is worse. The piece cites EPA findings that Elon Musk violated air quality laws when installing dozens of turbines to power a Memphis data center. The reporter argues that treating the backlash as a culture war obscures a simple point: a small city of computers is hard to live beside, even with better tax deals, local hiring and community engagement.

Electricity is not the only motive. Job losses, existential risk, concentrated corporate power and misinformation, including overstated water use, all play a part, and heavy electricity consumption is not inherently bad. Still, one developer said industry insiders see ambition constrained less by public opposition than by a lack of electricity. Looking ahead, the piece suggests siting may shift toward places with few people able to resist, such as the Permian Basin with its gas pipeline network, or abroad: Google announced the day after the elections that it will spend $15 billion on data center projects in Malaysia. Musk posted hours after results came in that data centers in space look appealing.

Why it matters: Developers and infrastructure teams planning AI capacity now face electricity as the binding constraint, not just land, chips or permits. In practice, siting, power procurement and community relations are becoming core product decisions, and behind-the-meter generation brings air-quality and permitting exposure that grid connections avoided. A likely inference is that power availability, not public opinion alone, will decide where the next generation of training clusters gets built.

Read at MIT Technology Review · AI

Based on reporting from the original publisher. Visit the source for full context and later updates.

Publisher excerpt

This story originally appeared in The Algorithm, our weekly newsletter on AI. To get stories like this in your inbox first, sign up here. As voters went to the polls across the US last week, I was in Ashburn, Virginia, part of what’s known as Data Center Alley—the world’s largest data center hub. It’s also where…