By James Titcomb
Down a single-lane road, a couple of hundred metres from the local secondary school, and through a creaky metal gate, a field of oilseed rape is starting to break out in yellow flower.
This farmland, dotted with old oak trees, separates the Hertfordshire town of Potters Bar from the M25 and a busy service station, acting as a buffer zone between residents and London’s orbital motorway.
Locals guard their patch of green belt fiercely. So in September 2024, when the newly elected Labour Government endorsed plans for a £4bn project marketed as Europe’s largest artificial intelligence data centre, opposition quickly mounted.
Despite protests from close to a thousand residents, the project, which will generate millions in local tax revenues, was approved by the council. Equinix, the US data centre giant that owns the site, is planning to start construction next year.
Campaigners are determined to make life difficult. The residents’ association opposing the plans vows to fight Equinix at every corner.
“We’re just going to keep throwing spanners in the works,” says Ros Naylor, one of the lead protesters. Fleur Albrecht, another local opponent, suggests if that doesn’t work, “we’ll strip naked and tie ourselves to the trees”.
Politicians and major businesses are racing to infuse AI into as much of the economy and public sector as possible, claiming it could lead to scientific breakthroughs, boost profits and make public services more effective.
But among the wider public, enthusiasm has been much more muted. As residents see more data centres pop up and AI increasingly starts to affect their employment prospects, both Westminster and Washington are facing a growing backlash.
The Equinix data centre is one of more than 100 new sites planned across the country amid voracious demand for computing power from AI companies. The Government has welcomed such investments with open arms.
In Potters Bar, residents say that the 250-megawatt data centre, which will rely on 240 diesel generators for backup power will blight the area, creating noise and environmental pollution. Posts on the campaign group’s Facebook page suggest that after recent attacks on data centres in the Middle East, the site could even be a target – a particular concern given its proximity to the local school.
Equinix says the site would be powered by renewable energy and would attract AI investment into the country, as well as creating 2,500 construction jobs and 200 permanent roles. The company claims the data centre will also be subject to strict noise and air quality standards.
“We understand and appreciate how deeply people feel about their communities,” a spokesman says. “Data centres support regional and national growth, and as we operate within communities, we are committed to doing so responsibly.”
Local opposition to major infrastructure projects is nothing new. Similar reactions occur up and down Britain when a new solar farm, housing estate or transport link threatens to disrupt rural life. But even if their campaign is ultimately fruitless, the opponents of the large data centre believe they have tapped into a wider backlash.
None of the Potters Bar campaigners who say the data centre will scar their community is particularly enthusiastic about AI, pointing out that it has made it more difficult to speak to a human when calling customer service.
Naylor likens it to sugary sweets placed near a supermarket till: you didn’t ask for it, but when it’s there, you end up taking it. “It’s being thrown in our faces every day,” she says.
A growing proportion of the population agrees. According to polling by YouGov last August, just 25pc of Britons had a positive perception of the technology compared with 38pc who had a negative perception.
The numbers are similar in the US, with English-speaking countries generally more negative about AI than in other parts of the world.
The prospect of unemployment is by far the biggest reason for its unpopularity. Two thirds of Britons said they believed more jobs would be lost to automation than would be created.
The increasing visibility of AI in everyday life has also coincided with people being more negative about it. The US polling group Pew Research asks members of the public every year for their views on AI. In 2021, before the release of ChatGPT, 37pc said they were more concerned than excited about the technology. By last year, this had risen to 50pc.
The Silicon Valley companies building AI are becoming increasingly conscious of this.
“You can’t just go around saying, ‘we’re going to create all this abundance, a lot of it is going to go to us, and we’re going to be trillionaires’,” Dario Amodei, the boss of tech giant Anthropic, recently told Axios. “You’re going to get a mob coming for you if you don’t do this in the right way.”
This opposition is likely to become more acute as a growing number of data centres pop up, and if AI starts to have a more visible influence on the job market. Block, a payments company, laid off 40pc of its 10,000 staff in February – proportionately one of the biggest ever rounds of redundancies by a major US company – with Jack Dorsey, the chief executive, saying AI could handle the work instead.
Scepticism, however, comes from many places. While most people say they support the idea of AI being used to make medical breakthroughs or saving time at work, they are concerned it will pollute online information with “fake news” and images, weaken democracy and make romantic relationships harder.
AI has raised the price of computers and games consoles, as the supply of memory chips is directed to the boom. It has also led to an avalanche of “slop” – low-quality LinkedIn posts and pictures on Facebook that, despite being widely disliked, hit the algorithm’s right buttons.
But it is because of data centres that the opposition is most fierce, and where the political backlash to AI is likely to hit first.
The release of ChatGPT in late 2022 prompted soaring investment in AI, much of which has gone into the humming, warehouse-like structures stuffed with computer servers where new systems are trained and costly microchips transform queries into answers.
Politicians have jumped on the opportunity for new investment and jobs. Since Labour came to power two years ago, data centres have been categorised as critical national infrastructure and in some cases have led to prioritised planning and power decisions.
The Government has also attached itself to a series of corporate press releases trumpeting new data centres, including the Hertfordshire campus and announcements from Google and Microsoft.
In the US, the Trump administration has called for a building boom of data centres. On his first full day in office, the Donald Trump promised that “we’re going to make it as easy as it can be” to build.
But on the ground, in local politics, the US president’s stance is at odds with a rapidly growing network of grassroots opposition against the AI rollout – and particularly the development of the data centres behind it.
It extends deep into the Maga heartlands. And it is fast becoming arguably the most unifying force in American politics.
‘Private companies taking advantage of rural communities’
Three years ago, Kyle Schmidt and his wife moved to the countryside outside a 20,000-person town called Sand Springs in Oklahoma, a state that voted for Trump in 2024 with a 34 percentage point margin.
“We wanted to have a little farm, raise our kids and see the stars at night,” says the 43-year-old.
When he found out in November that an 827-acre stretch of land about a mile from his home was getting rezoned to build a hyper-scale data centre for Google, he devoted his life to opposing it.
“We’ve invested our life’s work and savings to build this quiet country life and this would absolutely destroy that.”
Schmidt founded the Protect Sand Springs Alliance, which has filed two lawsuits against the city of Sand Springs over the plans.
“I still work my regular job. So I work from 6.30am to 11pm at night. And I’ve been doing that seven days a week for the last four months straight.”
Schmidt, who is a Trump voter, agrees with the president’s framing of AI as an arms race, but says this does not mean that he wants a data centre nearby.
“I can agree with an objective and also disagree with the means by which that objective is accomplished,” he says.
“I think this is more of a problem with private companies taking advantage of small rural communities for their own personal profit.”
Despite tech executives’ insistence that AI will bring a productive utopia, a slew of survey data show Americans simply do not believe them.
Polling by Public Opinion Strategies (POS) for NBC News, released this month, showed that Americans believe the risks of AI outweigh the benefits by a margin of 23pc. The fear was strongest among Democrats at 33pc, but even among Republicans the margin was 8pc.
Americans fear the threat of AI
“Everywhere I go and speak I just see more concern and more energy to do something about it,” says Brendan Steinhauser, a Republican strategist who runs the Alliance for Secure AI, which launched last June.
Much of this fear hinges on employment. A CNBC poll found 72pc of Americans think that AI will eliminate more jobs than it will create. But there are also fears about how AI could be used for surveillance, the impact of data centres on electricity bills and the environment, whether children are safe online, tech executives having too much power and how deep fakes could influence politics.
‘Losing our jobs to machines’
Republicans may be less likely to share Democrats’ concerns over the environmental impact of data centres. But Right-wing circles have their own point of contention with AI: a theological dimension.
“We’re working with a lot of faith leaders around the country and they’re very concerned about what AI could do to the soul and to the worship of God,” says Steinhauser.
“If we lose our jobs to machines, what does that do for our purpose and our dignity? They’re very concerned about people having relationships with AI companions instead of with each other. They’re worried about what that can mean for the nuclear family and for procreation,” he says.
“Priests and pastors and rabbis want to make sure that we don’t try to replace God with a machine. That would be sacrilegious and it would be detrimental to the souls of all believers.”
Campaigning is now moving from disorganised grumbling to co-ordinated resistance.
Growing environmental concerns
Last spring, there were 188 community groups opposing data centre construction across the US, according to Data Center Watch, a national tracker. By the end of 2025, the number had doubled to more than 375, says Miquel Vila, a lead analyst.
Since 2023, activists have held up US data centre projects worth more than $150bn, the group says.
Campaigning has also thwarted plans in Britain. In January, the Government was forced to admit it had wrongly given planning approval for a data centre in Iver, Buckinghamshire, 30 miles anti-clockwise along the M25 from Equinix’s proposed site, after campaign groups Foxglove and Global Action said the site’s electricity demands had not been considered.
Edinburgh city council blocked plans for a site at the former headquarters of Royal Bank of Scotland last month, despite the proposal being supported by planning officers.
Politicians are starting to notice. So far this year, 382 US state bills specifically targeted at data centres have been either proposed or carried over from 2025, including around whether or not to allow tax incentives for developers.
“It’s hard to overstate how big a conversation this is on the state and local levels right now,” says Morgan Scarboro, the vice president at MultiState, which tracks state legislation. “It has just been exponential growth.”
When it comes to AI in general, there have been an additional 1,576 bills so far this year. Anticipating a spate of local crackdowns, the AI industry has lobbied hard for federal laws blocking states from getting in the way.
The appetite for regulation is growing in Britain, too. Research from the Ada Lovelace Institute and Alan Turing Institute last year found that 72pc of Britons wanted to see AI regulation, up from 62pc when asked two years earlier.
“There’s a potential mismatch between where the Government is and where the public are,” says Roshni Modhvadia, a researcher at the Ada Lovelace Institute.
The AI backlash could grow further as consumers start bearing more of the costs. Today, the vast majority of ChatGPT users interact with the service through a free subscription, subsidised by hundreds of billions of dollars of venture capital cash that has been thrown at the company.
That is unlikely to last: OpenAI, ChatGPT’s owner, is experimenting with adverts to help fund its spending, a development that has raised privacy concerns even as the company insists it will not sell data about what users type into the bot.
Sam Altman, OpenAI’s chief executive, recently suggested that AI would become a pay-as-you-go service, charged like an energy bill.
“We see a future where intelligence is a utility, like electricity or water, and people buy it from us on a meter,” he told a conference last week. Clips of the event posted on social media prompted outrage.
However, the direction of travel appears clear. Whether with dollars or eyeballs, users will soon be paying.
National politicians are starting to wade in on the issue. Bernie Sanders, one of the Democratic Party’s leading figures, has called for a moratorium on data centre construction.
Trump is becoming increasingly cognisant of the mismatch between his big tech backers and his Maga base as he approaches the November Midterms.
The president recently told the leaders of Google, Microsoft, xAI and Meta at a White House meeting that their companies needed “PR help”.
In his state of the union address at the end of February, Trump announced a “ratepayer protection pledge” that he said will oblige tech companies to build their own power plants for their data centres.
For now, Republican leaders are divided on AI. On one side, JD Vance, the vice president, Marco Rubio, the secretary of state, and Ted Cruz, the Texas senator, have become fierce proponents.
‘Broligarchs’ vs the Maga masses
On the other, Ron DeSantis, the Florida governor, has become an unlikely champion of AI regulation, while Steve Bannon, Trump’s former chief strategist, has warned that AI is “maybe the most dangerous thing we’re dealing with right now”.
Bannon depicts the fight for the future of the Right as a battle between the tech “broligarchs” and the Maga masses.
James Fishback, a far-Right candidate running to replace DeSantis as the governor of Florida, has made opposition to data centres a key pillar of his campaign.
He told Politico last week: “They’re not building an AI data centre next door to Mar-a-Lago. They’re building AI data centres in our agricultural communities like Loxahatchee and Fort Meade, and I’m running for governor to stop that.”
British politics has been slower to wake up to the issue, but perhaps not for long. Zack Polanski, who has pushed the Green Party into second place in the polls, recently warned against embracing the technology.
“AI – in many ways – has potential to be a force for good, but is already causing people to lose their jobs, consuming huge amounts of energy and water,” he said.
“Planned data centres will produce little employment and blight communities – plus further jeopardise our climate targets.”
Polly Billington, the Labour MP for East Thanet, called for a “national conversation” about data centres, telling Politics Home last week: “We can hurtle towards a data centre future, but is that overall in the best interests of the country?”
Naylor, in Potters Bar, says she had been offered the chance to talk to Nigel Farage about her campaign, but has turned it down, saying: “I don’t want it to become a political football.”
Campaigners (L-R) Fleur Albrecht, Eamonn Lynch, Odette Garvey, Cherry Chasteauneuf and Ros Naylor oppose plans to build data centres on green belt land near Potters Bar
Campaigners (L-R) Fleur Albrecht, Eamonn Lynch, Odette Garvey, Cherry Chasteauneuf and Ros Naylor oppose plans to build data centres on green belt land near Potters Bar Credit: Ben Montgomery for The Telegraph
Oliver Dowden, the local Conservative MP and a former deputy prime minister, has voiced his opposition to the data centre, despite having been a vocal AI advocate when in government.’
Last month, MPs on the environmental audit committee launched an inquiry into whether Labour’s data centre ambitions were at odds with its net zero targets, after Energy Secretary Ed Miliband said this was “inherently uncertain”.
Popular protest is becoming more visible. In February, activists holding signs saying “Pull The Plug” and “No to AI” held a demonstration outside big tech offices in London.
More of this can be expected should AI affect employment prospects to the extent that many predict, as happened in 2020, when A-level students chanted “F— the algorithm” as they protested against an automated system used to give them grades in lieu of exam results.
There are signs that the AI fever that has swept Westminster is starting to break. On Wednesday, the Government dropped plans to let AI companies train their systems on copyrighted work without permission. It followed a well-publicised campaign by musicians and artists, including Sir Paul McCartney and Sir Elton John.
‘I have no doubt that Trump will respond’
Trump has also sounded more sceptical. This week he said that AI “can be very dangerous” after viewing fake social media images appearing to show an attack on a US aircraft carrier.
Steinhauser argues that the issue will become more pressing for the US president in the coming weeks and months as AI continues to change the economy. A fleet of major companies has signalled that they were replacing human employees with AI, including HP, Amazon, IBM and Salesforce.
“I have no doubt in my mind that, as the job losses continue to tick up, Trump will respond,” says Steinhauser.
“This is an iceberg up ahead. We can turn the ship but you can’t wait until the last minute, otherwise you’re going to face a huge political challenge.”
The backlash is concerning for anyone who sees the technology as this century’s Manhattan Project; the race that will determine where power lies in the coming decades. Polling from Ipsos in 2024 found that while just 39pc of Americans believe AI offers more benefits than drawbacks, 83pc of Chinese people do.
“AI is not very popular in the US right now,” Altman said last week. “If we don’t move as quickly as other countries … then I think we will lose the advantage that we have.”