Data Centre Backlash Grows In Us After Midterm Results, Putting AI Expansion Under Pressure

The November 3 US midterm elections could mark the beginning of a tougher political battle over data centres, as voters in several states pushed back against the rapid expansion of infrastructure needed to power artificial intelligence.
Democrats emerged from the elections with control of Congress while Republicans retained the White House, creating the prospect of a deeply divided government. The election results also showed that opposition to data centres is becoming an increasingly important local political issue.
Latest and Popular Mobiles
Data centres house the computing infrastructure behind AI services, but their rapid expansion has raised concerns over electricity consumption, environmental impact, water use, land requirements and the effect on local utility bills.
An NBC exit poll found that more than half of voters believed AI would do more harm than good. Nearly 80% opposed having data centres built in their neighbourhoods, while close to 60% said the issue was a major or crucial factor in their voting decision.
Related Articles
Nadine Smith, president and CEO of racial justice organisation Color of Change, described the election as a referendum on who benefits from the changing economy. She pointed to communities such as South Memphis, Tennessee, where xAI operates a large data centre powered partly by gas turbines.
The backlash was particularly visible in Ohio, where OpenAI has considered a massive data centre project reportedly valued at $500 billion, with more than $100 billion in financing backed by Nvidia. In one Ohio county where Meta had planned a 700,000-square-foot data centre, voters approved a ban on data centres. Other communities in the state also rejected proposals related to further data centre development.
Analysts at Morgan Stanley have suggested that Ohio governor-elect Amy Acton could create additional uncertainty for the state’s data centre expansion. Candidates who campaigned against rapid data centre construction also won races in states including Texas, Maine, Michigan, Nevada, Georgia and South Carolina.
Anshi Bhatt, a technology and policy researcher, said the results showed that large campaign spending by the technology industry did not necessarily translate into electoral victories. She also noted that presenting data centres as a local problem can become an effective campaign strategy.
The political response could now lead to more construction moratoriums and tighter rules around new facilities. One proposal gaining attention is requiring data centre operators to pay for electricity and grid upgrades needed to support their facilities rather than shifting those costs to surrounding communities.
Goldman Sachs has predicted that Democratic gains could contribute to delays in data centre construction in states such as Georgia and Texas.
Democratic control of Congress could also give greater momentum to legislation aimed at slowing AI infrastructure expansion. Earlier this year, Senator Bernie Sanders and Representative Alexandria Ocasio-Cortez introduced a proposal calling for a nationwide pause on new data centre construction until Congress establishes rules covering electricity costs, environmental impacts and community approval. However, such legislation could still face a presidential veto.
Another proposal, the GRID Savings Act from New Mexico Senator Martin Heinrich, seeks to prevent the costs of grid infrastructure associated with large AI data centres from being passed on to local communities.
Environmental concerns are also becoming more prominent. Gas-powered generation at xAI’s Memphis facility and similar arrangements at data centres in Virginia have drawn attention from environmental and community groups.
Maia Woluchem, programme director at research organisation Data & Society, said changes involving energy regulations, land rights and environmental rules could have an impact on future data centre projects. She also expects state and local governments to remain active regardless of what happens at the federal level.
In Texas, governor-elect Gina Hinojosa has supported a statewide moratorium on data centres, while Governor Greg Abbott has also called for certain restrictions. The debate is particularly significant in Texas because its electricity network has faced reliability challenges in recent years.
However, even among Democrats, there is disagreement over a complete freeze on data centre construction. Some lawmakers and technology supporters argue that restrictions could slow the development of AI and weaken US competitiveness against China.
Nadine Smith also expects technology companies and AI lobbyists to increase their efforts in Washington as the new Congress takes shape. She said Democrats campaigned on making major technology companies pay more and argued that voters will expect lawmakers to follow through.
The opposition is not necessarily about data centres themselves. A private equity executive told the publication that public frustration is increasingly directed at AI, with data centres becoming a physical symbol of broader concerns about the technology and the power of major technology companies.
Woluchem similarly said the election results point to growing frustration with technology power, describing the issue as one that crosses traditional political boundaries.
Another source of anger is the way some projects have been approved. In Gilroy, California, Amazon reportedly built a data centre with limited public awareness by relying on an old zoning provision. Concerns have also grown over local officials being asked to sign nondisclosure agreements that restrict what they can say about planned projects.
Tax incentives have added to the controversy, with data centre operators receiving substantial benefits from governments while promoting job creation and community investments as reasons to support new facilities.
For many residents, however, the central issue is whether they have a meaningful say over projects that can reshape their communities and affect electricity prices and other living costs.
Smith argued that if AI infrastructure is genuinely a national priority, the companies building it should cover the costs of the power and infrastructure they require, comply with environmental rules and provide greater transparency to nearby residents.







