US data centre backlash emerges as factor in midterm election debate
Voter polling points to growing opposition to AI infrastructure, while Democratic gains in major data centre states raise questions about the industry’s political and regulatory future.
Public opposition to the rapid expansion of artificial intelligence infrastructure has emerged as a significant issue in the aftermath of the US midterm elections, with voter polling indicating that data centre construction influenced many electoral decisions. The results have intensified debate over whether concerns about electricity costs, water consumption and local environmental impacts could translate into tighter regulation of the industry.
A congressional voter poll conducted for the Associated Press and a consortium of news organisations found that 56% of surveyed voters considered data centres a crucial or major factor in their vote. The same poll found that 76% would oppose the construction of a new data centre in their community. Opposition was considerably stronger among voters who supported Democratic candidates.
However, the findings do not establish that resistance to data centres determined the election outcome. Government corruption and the cost of living were considerably more prominent concerns, identified as crucial or major factors by 90% and 89% of respondents, respectively. The polling therefore suggests that AI infrastructure has become politically relevant, but it does not establish how many votes were gained or lost because of the issue.
The electoral implications are particularly relevant in states with substantial data centre investment. Texas, Ohio and Iowa, which elected Democratic senators, have 595, 258 and 117 data centres respectively, compared with a median of 55 across the 50 states, according to the BBC’s analysis. Texas also has 31 ongoing data centre construction projects, the highest number of any state.
The political significance extends beyond Senate contests. Texas, Georgia and Ohio elected Democratic senators and governors who have advocated some form of tighter oversight of data centre development. Their proposals include temporary construction restrictions, the removal of tax incentives, stricter environmental safeguards and requirements for developers to finance the additional electricity and water infrastructure their projects require.
Some proposals also involve unionised construction labour, greater transparency and local authority to veto projects. These measures reflect a growing debate over who should bear the costs of the infrastructure needed to support the AI economy.
The BBC’s analysis of state-level polling suggests that states shifting towards Democrats either reflected the national pattern of concern about data centres or appeared less concerned about the issue. It does not establish a direct relationship between opposition to data centre construction and the electoral swing.
The potential regulatory shift matters because investment in data centres has become an increasingly important component of US economic activity. Official data cited by the BBC show that inflation-adjusted monthly spending on data centre construction, excluding the information technology equipment installed inside the facilities, reached $5 billion in August 2026. That was five times the level recorded in 2023.
Research by Columbia Business School suggests that investment in AI infrastructure is now larger as a share of the US economy than spending on new railways was at the height of the railway expansion in the 19th century. The comparison illustrates the scale of the current investment cycle and its significance for construction, energy demand and capital expenditure.
Yet the expansion has created tensions between national economic ambitions and local concerns. Data centres require substantial electricity, land and, depending on their cooling systems, water resources. Their development can place additional demands on power grids and local infrastructure, raising questions about whether households and existing businesses could face higher costs.
These concerns have become politically consequential as communities assess the benefits of investment against its local impact. Construction jobs, tax revenue and associated economic activity provide incentives for local governments to support projects. At the same time, residents are increasingly demanding assurances that the costs of expansion will not be transferred to electricity consumers or local authorities.
Industry analysts cited by the BBC expect the election results to increase pressure on developers, although they differ on how far restrictions could go.
Cori Crider of the Future of Tech Institute said the industry should anticipate that new state leaders could impose conditions on data centre construction or even halt projects they consider excessive. Seema Shah of Principal Asset Management warned that requirements to finance grid improvements, meet tougher water-use standards or surrender tax incentives could increase development costs and slow the arrival of new capacity.
Heather West of Eurasia Group offered a more measured assessment. She said temporary construction bans and new regulations might delay expansion rather than stop it altogether. Projects that have already received permits or entered construction are likely to proceed, while the next wave could face slower approvals, local opposition and greater pressure to finance its own electricity supply.
Oxford Economics analysts have also pointed to the economic incentives facing politicians. Data centres generate tax revenue and construction employment, making policies that severely disrupt development potentially costly for state economies.
Deepak Surana of the Futurum Group consequently expects negotiations between developers and state authorities rather than a comprehensive moratorium. Such an approach could allow projects to continue while imposing additional obligations to protect residents from higher energy and water costs.
The political debate has already produced a response from one of the industry’s major companies. OpenAI, which operates a data centre in Texas and has projects under construction in Ohio, pledged on Thursday, one day after the election results, to ensure that it fully pays for its own electricity.
The commitment comes as concerns about the relationship between AI investment and household energy costs become more prominent. It also illustrates how public pressure and changing political conditions can influence corporate infrastructure policies even before new regulations are formally adopted.
For the wider industry, the central challenge is no longer simply securing land, financing and access to electricity. Developers may increasingly need to demonstrate that their projects will not impose disproportionate costs on local communities.
The midterm results do not establish that a nationwide revolt against data centres produced the Democratic gains. Corruption, living costs and other political issues remained more prominent in voter polling. Nevertheless, the findings indicate that opposition to local AI infrastructure has moved into mainstream electoral debate.
The longer-term implications will depend on how state governments translate campaign commitments into legislation, permitting rules and infrastructure requirements. A negotiated framework could allow the AI investment cycle to continue while shifting more of its costs towards developers. More restrictive policies, however, could delay projects and alter where companies choose to build.
The emerging political test for the AI industry is therefore whether it can sustain rapid infrastructure expansion while demonstrating that the economic benefits do not come at the expense of the communities hosting it.
Sources: BBC Verify — “Did a data centre revolt affect the US midterm election results?” by Ben Chu (6 November 2026); Associated Press — congressional voter polling; Pew Research Center — public attitudes towards data centres; US Census Bureau — construction spending data; Columbia Business School — AI infrastructure investment analysis. Comments from the Future of Tech Institute, Principal Asset Management, Eurasia Group, Oxford Economics and the Futurum Group are reported in the BBC article.