AI NIMBYism: Communities push back against energy-hungry data centres. What’s the solution?
Residents impacted by AI infrastructure projects have cause for concern over resource consumption and impact on local jobs and quality of life. Find out which UK and EU cities are hosting AI data centres
Behind all the Trump-related news feeds, a new form of community resistance is emerging across the United States and beyond. The phenomenon, dubbed “AI Not-In-My-Back-Yard” by industry observers, sees local communities increasingly pushing back against the construction of massive data centres required to power AI systems, citing concerns over energy consumption, water usage, and the impact on their quality of life.
The scale of opposition is staggering. According to Data Center Watch research, over $64 billion worth of data centre projects have been blocked or delayed between March 2024 and May 2025 due to local opposition. This represents a significant escalation “Not In My Backyard” (NIMBY) sentiment, traditionally associated with infrastructure projects such as power plants and airports.
In Europe alone, according to real estate advisor Savills, 94 new data centre projects are due to be delivered in the next four years, totalling approximately 2,800 MW in capacity.
UK and EU cities where AI data centres are being built
- Northumberland (Blyth): Blackstone is developing a £10 billion AI-ready facility, projected to be Europe’s largest AI data centre
- Hertfordshire: The £3.75 billion DC01UK project has received planning approval, aiming to become one of Europe’s largest AI and cloud hubs
- Cardiff, Wales: Latos Data Centres plans a 90 MW Tier III hyperscale data centre
- Stockport, Greater Manchester: Kao Data is constructing a £350 million data centre
- Oxfordshire: CloudHQ is developing a £1.9 billion data centre campus
- Loughton, Essex: Nscale is investing £2.5 billion in a new facility
- Nationwide UK Investments:
- Amazon Web Services (AWS) announced an £8 billion investment in UK data centres
- CyrusOne, ServiceNow, CloudHQ, and CoreWeave have collectively committed £6.3 billion
- NVIDIA is collaborating with NVIDIA Cloud Partners Nebius and Nscale to deploy 14,000 NVIDIA Blackwell GPUs
EU AI Data Centres
The European Union is also a hub for AI data centre development:
Germany:
- Bischofsheim: Yondr Group launched the first 20 MW phase of its 40 MW data centre project
- Lübbenau: Schwarz Group is developing a new data centre campus
- A consortium including SAP, Deutsche Telekom, Ionos, and Schwarz Group is discussing building a major AI data processing centre
- NVIDIA is establishing an AI Factory in Germany
France:
NVIDIA supports regional technology and cloud providers Domyn, Mistral AI, and Nebius to deploy NVIDIA Blackwell Systems.
Italy:
- Milan (Cornaredo cluster): Virtus Data Centres is entering the Italian market with a 70 MW facility.
- Near Rome: Gruppo Tim is investing approximately $136 million to develop a new data centre.
- NVIDIA is establishing and expanding AI technology centres in Italy
Spain:
- Aragon: Blackstone is developing a €7.5 billion, 300 MW data centre campus
- The NVIDIA AI centre in Spain will expand to include a new AI factory with the Barcelona Supercomputing Centre
Netherlands: Microsoft has hyperscale data centres
Belgium (Saint-Ghislain, Hainaut): Google has unveiled plans to build a €1 billion data centre
Sweden: Borlänge: EcoDataCentre is converting a former paper mill plant into a data centre mega campus
NVIDIA is also building AI technology centres in Sweden
Finland: Mäntsälä: Verne announced plans for a 70 MW data centre campus
Pori: CompassForge Ventures unveiled plans for the Ecogrid Energy Park
NVIDIA is expanding AI technology centres in Finland
Switzerland (Lupfig): Microsoft has broken ground on its Zurich West campus. Swisscom is launching new AI services
Denmark (Ølgod): atNorth plans to build its largest data centre.
Norway: Telenor is expanding its sovereign AI infrastructure
Data centres do bring job opportunities, including:
- Commissioning and QA/QC roles
- MEP project management
- Data centre design engineering
- Health & safety professionals with DC experience
Source: AA Euro
Local towns drained of natural resources?
The rapid expansion of AI capabilities has created unprecedented demand for computing power, with cascading effects on local resources. By 2028, Lawrence Berkeley National Laboratory forecasts that US data centres alone could consume as much as 12% of the nation’s electricity, driven primarily by generative AI applications.
Water consumption is a more pressing challenge. Bloomberg’s analysis reveals that data centres are increasingly being built in areas where an “abundant supply of water” is scarce, threatening local water supplies, agriculture, and energy production. The Stanford report notes that these facilities, now growing to “brobdingnagian sizes,” require significant amounts of water for cooling systems.
The human impact is becoming increasingly evident. One Virginia resident described living near data centre infrastructure as “like living in hell,” highlighting the noise and disruption these facilities can bring to communities.
Quality of life near AI data centres
According to a Semafor report, 71-year-old Elizabeth Martorana voiced her concerns at a Prince William County meeting in Virginia about how the retirement community where she lives sits within a few miles of more than 20 proposed data centre parks under various stages of development. Her home is adjacent to a proposed AI campus expected to be the largest in the world once operational.
Not exactly what she and other community members were sold when they signed up for their relaxing days of retirement.
Back in the UK, communities are also voicing concerns about AI data centre impacts. However, those calls for help won’t likely be answered by government intervention.
In fact, the reclassification of data centres as “Critical National Infrastructure” by the UK government may allow developers to override local objections. There have been concerns about new data centres being built on green belt land.
In East London (Havering Council), Digital Reef has plans to build a £5.3 billion data centre campus which has faced significant opposition from the Campaign to Protect Rural England (CPRE). Campaign members have concerns about massive environmental damage and the unsuitability of the proposed site, which is a Site of Interest for Nature Conservation.
The project has plans to construct a 12-building data centre campus, with 300 acres of adjoining land given over to a public ecology park. However, the CPRE is organising opposition, according to a Techerati report.
“We are astonished and appalled the council has agreed to sink valuable time into developing a proposal which will cause massive environmental damage and is highly unlikely to gain permission,” said Alice Roberts, CPRE’s Head of Campaigns.
Ian Pirie, co-ordinator of Havering Friends of the Earth, has questioned if the council “really know North Ockendon and its residents.”
Pirie expressed on behalf of the local environment group in a Cybernews report: “We get the distinct impression that the council gave the go-ahead for this proposal first and then tried to think of justifications for it afterwards.”
Investment surge
The financial stakes continue to escalate. Big tech has too much money invested and too much future growth to lose if they don’t find a way to appease opposing residents. Flashbacks of the 1970s anti-nuclear energy movement come to mind.
MIT Technology Review reports that Apple announced plans to spend $500 billion on manufacturing and data centres in the US over the next four years, while Google expects to spend $75 billion on AI infrastructure alone in 2025. This represents what the publication calls “a marked departure from Big Tech’s” previous spending patterns.
Deloitte research indicates that data centre capacity has grown 50% to 60% quarter over quarter globally since the first quarter of 2023, with similar growth expected through the first quarter of 2025.
No solutions? Good luck with community relations
The data centre industry is begrudgingly beginning to acknowledge the challenge. Data Center Dynamics analysis suggests that “opposition to data centre development often stems from a lack of understanding and engagement with local communities and stakeholders.” Industry experts are calling for better community engagement strategies to address these concerns.
However, Raconteur reported community opposition extends beyond communication issues, with protesters questioning promises of economic benefits as “greenwashing” and raising practical concerns about infrastructure capacity to support construction and ongoing operations.
Jobs created out of AI’s energy usage mess
Despite the challenges, significant research efforts are underway to address AI’s environmental impact. The National Renewable Energy Laboratory (NREL) announced in January 2025 that researchers are “investigating methods to make computing more energy efficient” through their collaboration with Intel on the Green Computing Catalyzer framework.
As far as two years ago, MIT researchers claimed to have developed promising energy reduction techniques, including early stopping methods that achieved “an 80% reduction in the energy used for model training” across computer vision, natural language processing, and material design applications.
Revolutionary memory chip technologies
One of the most promising developments involves processing data directly within memory chips, eliminating the energy-intensive process of moving data between storage and processing units. Researchers have unveiled CRAM (Compute-in-RAM) technology that “will dramatically reduce power demands for AI workloads” by performing computations directly inside memory cells.
The potential energy savings are extraordinary. Tom’s Hardware reports that CRAM technology gives RAM chips the power to process data, not just store it, with TrendForce analysis suggesting this approach could reduce AI energy consumption by 1,000 times or more.
Neuromorphic computing research also shows remarkable promise. Neuromorphic chips reportedly cut energy consumption by 1000 times compared to traditional processors, with the neuromorphic computing market projected to grow at 108% CAGR by 2025. Additional research demonstrates that three-dimensional memory architectures can achieve 6.9X less energy consumption than two-dimensional counterparts.
Singapore’s approach offers a model for policy intervention, with their Green Computing Funding Initiative providing SGD 30 million for research into energy-efficient software design and function.
AI accountability
The tension between AI advancement and community concerns reflects broader questions about sustainable technology development. Yale research has reported that legislators and regulators in both the US and EU are “starting to demand accountability” for AI’s resource consumption.
MIT experts suggest that while “AI workloads have sent data centre emissions skyrocketing,” there are viable solutions to reduce energy use and promote sustainable AI development.
As the AI revolution continues, the industry faces a critical challenge: balancing the transformative potential of artificial intelligence with the legitimate concerns of communities bearing the environmental and social costs of this technological advancement.
The resolution, or lack of one, could put a spanner in the works for AI data centres and determine the role of community consent of AI in the 21st century.
