TeaThey issued a press release announcing a gleaming supercomputer on the outskirts of North London, depicting a glass and concrete building rising from a tree-lined street. With images of glowing blue robot faces, it looks like the epicenter of a technological revolution.
By the end of this year that artist’s concept will become reality.
But when the Guardian visited last month, there was no sign of it. Instead, the four-acre plot in Loughton was a depot filled with pylons and scrap metal under a corrugated roof, while flatbed lorries went in and out with the poles.
Nine months before the project was completed, it is still a functioning scaffolding yard.
The story of the Lawton supercomputer is an insight into the dizzying ambitions of artificial intelligence as an economic powerhouse in the UK and beyond – but also how those expectations can merge into a less exciting reality.
Around the world, billions of dollars and the fortunes of governments, banks and pension funds are at stake on the promises of a few big companies – that they can rapidly build AI infrastructure and fundamentally remake the global economy.
Nvidia’s chief executive, Jensen Huang, appeared to offer a vote of confidence in the UK for some of this AI investment, making two visits to London in June and September 2025. At London Tech Week, he appeared on stage with Keir Starmer and called the UK tech ecosystem “jealousy of the world“. Later that year, during the government’s AI dealmaking frenzy, Huang told people should invest in uk “If they want to get rich” and predicted that the UK would be an “AI superpower”.
But the major investments in Britain questioned by the Guardian are not as they have been presented. Big sums do not appear to be a solid commitment to the UK economy – and “new” datacentre projects have proven to be old buildings with new chips.
This is a global issue. A report this month found that half of the world’s datacenter projects coming online this year could be delayed. In the US, OpenAI’s $500bn Stargate project has been significantly pushed back due to stakeholder wrangling.
“There are a lot of narratives created by tech companies linking AI to economic growth,” said Cecilia Recap, an economics professor at University College London. “Companies claim they are making these big investments. But that’s a story that doesn’t hold water.”
A major investment is the Lawton supercomputer. The government announced the project in 2025, part of a plan to “turbocharge” the economy. reports say it three times stronger As the fastest supercomputer in the US, the “global top tier” for AI infrastructure.
The company that developed the supercomputer, Enscale, was a small London startup that never built a datacenter.
Enscale bills itself as an “AI hyperscaler” and advertises 11 datacenters on its website, including sites in Portugal and Lawton. All of those sites appear to be either under construction or contain premises acquired from other institutions.
The government announcement said Enscale “has signed a contract” to build supercomputers until 2026, and is investing $2.5 billion into the UK economy; Enscale said it had already purchased a site in Lawton, and promised that 750 jobs would be created as the supercomputer was built.
An announcement later in September said that Microsoft would partner with Enscale to build the site, part of the company’s $30 billion investment in the UK. Microsoft has suggested that $15 billion of this investment is for the Lawton supercomputer.
However, the Guardian understands that Enscale’s “contract”, and Enscale and Microsoft’s $17.5 billion “investment”, is not a commitment to the UK government, other UK companies or the UK economy.
According to an Enscale spokesperson, the contract the government was referring to appears to be a contract between Microsoft and Enscale. The government said it had no mechanism to audit the $2.5 billion investment, which “may also include equipment and capital financing” and that it was “not a formal contract, but the intention was to inject capital”.
Like US firm CoreWeave, another key player in the government’s AI ambitions, Nscale is set to spin off from Australian crypto firm Arcane Energy in 2024 as a Bitcoin mining company. The investment is expected to consist mostly of Nvidia chips.
As for Microsoft, it clarified to the Guardian that it was not building a supercomputer, but had agreed to become Enscale’s customer when the datacenter was built.
So the $17.5 billion investment represents a plan by a UK company to buy chips made in Taiwan by a US company, put them in a building in Loughton and rent them to another US tech company.
Asked how the site would create 750 jobs, Enscale could not say how the figure was calculated.
Whether the supercomputer will be built by the end of the year or not is a different matter. Land records appear to indicate that Enscale has not been registered as the owner of the site, more than a year after the claimed purchase. Enscale could not say whether the company owned the land or not, and could not give a date on which any purchases had taken place.
After the Guardian launched inquiries, Enscale filed planning permission to build the supercomputer in the last week of February. It is understood that Enscale intends to start construction “very soon” within the next few months.
The Guardian understands that the site is highly unlikely to be completed this year, as it typically takes 18 to 36 months to build a hyperscale AI site – like, potentially, one of the world’s most powerful supercomputers.
Enscale said: “As a UK-headquartered company, we remain committed to the UK investment we have announced – the Lawton project in support of Microsoft is progressing as we envisioned. We are investing not only in the site but also in offsite power infrastructure, local contractors and local suppliers.”
An inevitable feature of the global AI economy may be that whether or not datacenters are built, economies grow and jobs are created, companies like Enscale – and their shareholders – stand to make windfall profits.
CoreWeave is another central feature of the government’s AI plans, and is to invest £1.5 billion in an “AI hub” in Lanarkshire. Like Enscale, the US company was originally a Bitcoin miner before moving towards AI; It is expected to build AI datacenters across the US and Europe.
Last month, US shareholders filed a lawsuit against the company alleging it hid information about its ability to build these datacenters even after it went public and inflated its valuation. that suit The company says it hid facts about construction delays at its datacenters and misrepresented its ability to deploy AI “at scale.” When the information came to light, six months after CoreWeave came to market, its price dropped dramatically.
CoreWeave said it was aware of the lawsuit, and said the claims were baseless and that it would vigorously defend itself.
If the allegations are proven in court, the concealed information would have allowed its investors – companies like Nvidia, who poured money into it before it went public – to make higher profits when they offered their shares on the market.
“There’s a lot of pressure to have an IPO, to get everyone’s money, to get all the money to all the early investors,” said Alvin Nguyen, an analyst at Forrester.
Enscale, meanwhile, raised $1.1 billion in funding in September, shortly after the government announced the company would form a key part of the UK’s Stargate project. Around the same time, Nvidia acquired a significant stake in the company through a £500m investment.
None of this money has yet appeared on Enscale’s balance sheet, as the company has not yet filed accounts. However, in October, the company allotted more than 2.9 million shares at a price of 1 pence.
Enscale said on Monday it had raised $2 billion in funding, boosting its valuation to $14.6 billion. Those 1 pence shares now appear to be worth hundreds of thousands of times their allotted price. It’s not clear who owns Enscale’s shares, but its investors include Nvidia, as well as funds such as Aker SA and Blue Owl Management.
Enscale argues that there were different classes of shares that were allocated, and other factors complicated the analysis.
“That’s a 350,000% return on investment,” Nguyen said. “There are very few things that will give you that, right?”
