The rapid expansion of AI infrastructure, and governments’ ambitions to remain competitive in the field, are facing growing resistance through opposition to data-centre development. Richard Gardiner examines the growing prominence of the issue in North America and Europe, and assesses the commercial implications of a landscape in which major opportunities are increasingly set against growing risks.
The rapid build-out of artificial intelligence (AI) infrastructure has become one of the defining commercial and strategic contests of recent years. Hyperscale data centres, which provide much of the storage and computing capacity underpinning AI systems, now form part of the physical backbone of a technology increasingly linked to economic growth, national security and data sovereignty.
However, the speed and scale of this expansion are generating growing public scrutiny, particularly in North America and Europe. Data centres have become a focal point for concerns about AI’s impact on employment, environmental sustainability and the concentration of economic power. As a result, the race to expand AI infrastructure is increasingly colliding with local political and social concerns, creating a more complex operating environment for developers, investors and organisations reliant on digital infrastructure.
Local concerns, wider anxieties
Public opposition to data-centre development is typically driven by a combination of local concerns and wider political grievances. At community level, campaign groups often focus on the tangible effects of new facilities: pressure on electricity and water supplies, land use, noise, and the perception that data centres generate few long-term jobs relative to their scale and resource demands.
These concerns increasingly intersect with broader debates about affordability, climate policy and environmental justice. Europe’s hot and dry summer of 2026, for example, has intensified scrutiny of water-intensive cooling systems used by some facilities in areas experiencing drought conditions or restrictions on household water use. In the US, communities have also questioned whether tax incentives for data-centre investment disproportionately benefit large technology firms, while host locations bear the costs of additional power demand, and potential increases in utility bills. Technology companies have pushed back, arguing that efficiency improvements can reduce the energy intensity of data centres, but this does not necessarily translate into lower overall energy demand given that the volume of AI computing is expected to grow rapidly.
Data centres provide campaign groups with a visible and fixed target to challenge a technology that is becoming increasingly influential in daily life, but which is otherwise difficult for the public to contest directly.''
At the same time, more ideological campaigns have connected local objections to broader criticism of AI. These include concerns about the political influence of major technology companies, the concentration of computing capacity and ownership, AI’s impact on employment, and its potential effects on privacy, surveillance and society more broadly. Opposition to data centres therefore rarely takes the form of a single, cohesive movement, but rather a broad collation of disgruntled or concerned actors.
The opposition takes shape
Data centres provide campaign groups with a visible and fixed target to challenge a technology that is becoming increasingly influential in daily life, but which is otherwise difficult for the public to contest directly. To date, opposition to the development of data-centres across North America and Europe has largely remained peaceful and focused on legal and political channels.
Campaign groups have achieved some success through planning objections, petitions, public-awareness campaigns and electoral pressure, helping to delay projects or secure tighter development conditions. Early-stage proposals are particularly vulnerable, as local stakeholders can influence decisions before construction begins. According to Data Center Watch – a research project that tracks local data-centre activity – opposition and associated regulatory uncertainty delayed or blocked at least 75 US data-centre projects worth approximately USD 130 billion in the first quarter of 2026. The reported disruption was the highest quarterly level recorded by the tracker and roughly matched the number of projects it identified as delayed or blocked during the whole of 2025.
Domestic opposition meets national priorities
The largely domestic and highly localised opposition to data centres is emerging just as geopolitical competition around AI is pushing national governments to prioritise infrastructure that can strengthen data and AI sovereignty. In the US, maintaining an advantage over China in defence, intelligence and cybersecurity has become a strategic priority. Across Europe, governments and European Union institutions are seeking to reduce dependence on US and Chinese providers and retain the capacity to store data, develop AI models and support critical technologies domestically.
This has created a gap between national policymakers focused on strategic and economic goals and state or local representatives wary of alienating constituents. The Austrian energy secretary’s response to opposition to a hyperscale data centre in Kronstorf captured this tension, arguing that “We cannot call for greater European AI sovereignty while at the same time putting up ever-new obstacles that hinder the very infrastructure on which it is based.”
In the US, with the 2026 midterm elections approaching, data-centre development has become increasingly politically sensitive in states including Virginia, New York and Texas. Candidates and state officials are being forced to balance local concerns, including calls for moratoriums, and tighter regulation, against the wider economic and strategic case for AI investment. New York has temporarily paused certain permits for large data centres, while Texas has also paused new data-centre approvals.
While most opposition is likely to remain non-violent, the risk of more disruptive direct action could increase if public concerns surrounding AI become more politically salient. The risk of mobilisation may also be greater where a project is connected to politically polarising companies or high-profile technology executives that have supported or advocated for certain political parties. This potential for escalation is already visible in online threats. The Soufan Center, a US-based non-profit organisation, has identified a rise over the past year in threats directed at developers, policymakers and technology companies, with physical sabotage of proposed and operational facilities appearing as a recurring theme in online forums. In sporadic cases, this rhetoric has also physically manifested. In November 2025, for example, an insurrectionary anarchist group claimed responsibility for an arson attack on equipment at the construction site of an AI Campus in Meudon, France.
The commercial stakes
Growing opposition is turning data-centre development into an increasingly political issue with clear commercial consequences. As scrutiny increases, the timelines and profitability of data centre developments may prove less predictable than investors had initially expected. In Ohio, for example, a US company had purchased just over 50 acres for a proposed data centre when Monroeville Village Council approved a one-year moratorium on new developments, placing the project on hold while officials reviewed regulation and zoning requirements. Developers and investors also face greater reputational exposure as projects come under public scrutiny. The industry has drawn criticism over a perceived lack of transparency, particularly following reports that major technology companies and industry bodies lobbied EU policymakers to keep site-level data on individual data centres’ energy use, water consumption and environmental performance confidential.
The implications extend beyond developers and investors. Organisations dependent on AI infrastructure and cloud services could face indirect disruption if opposition delays new capacity, particularly in markets where demand for computing power is already high. In more extreme scenarios, sabotage or physical attacks on data centres and supporting infrastructure could cause sudden and significant outages. Although there have not yet been comparable incidents in the US or Europe, recent examples elsewhere illustrate the potential impact. Iranian drone strikes on data centres in the United Arab Emirates and Bahrain in March 2026 caused damage and service disruption across several sectors, including online banking. Additionally, in Nepal, during the 2025 Gen Z protests, fires and vandalism damaged data-centre infrastructure, contributing to localised internet and cloud-service disruption.
Although such cases are still considered to be low-probability scenarios in America and Europe, the possibility of serious damage to a data centre due to sabotage is still likely to be on the minds of key players in the sector, particularly if anti-AI sentiment deepens and opposition becomes more openly confrontational.
Scaling up or pushing back?
The balance between commercial opportunity and risk in data-centre development is likely to become more finely poised as competition for AI leadership intensifies and the debate moves further into the political domain. Governments will continue to pursue the infrastructure required to support AI growth, even as concerns mount over the speed, concentration and potential consequences of the technology’s development.
This tension is unlikely to reduce demand for new digital infrastructure. But developers, investors and end users should be prepared for potential obstacles: the sector’s future will be shaped not by technical capacity alone, but by how effectively it responds to public concern, regulatory scrutiny and the wider political debate over AI’s role in society.
