OpenAI is sending the UK two messages at once. The first: your electricity is too expensive and your regulations are too unclear to justify building a data centre. The second: your talent is so good we are doubling our London headcount and making it our biggest research hub outside San Francisco.

Both messages arrived within days of each other. Together, they reveal something important about where AI infrastructure is actually heading: the people who build AI and the machines that run AI are splitting apart geographically, and the UK just became the sharpest example of that divide.

Why Stargate UK Died

The Stargate UK data centre project, a joint venture with Nvidia and Nscale, has been paused indefinitely. OpenAI told CNBC the project would restart "when the right conditions, such as regulation and the cost of energy, enable long-term infrastructure investment." No timeline was given.

The numbers explain the decision. UK industrial electricity prices are four times higher than in the United States and nearly double those in France, according to the Institute of Economic Affairs. For a data centre running thousands of GPUs around the clock, those costs make the math impossible regardless of political ambition or public backing.

Grid access is the second problem. Building a data centre takes 18 to 24 months. Connecting one to the UK power grid can take three to eight years. The government's AI Growth Zones policy, launched in November 2025, replaced a first-come, first-served system with a first-ready, first-connected model. It was not fast enough.

Ben Peters, CEO of industrial efficiency company Cogna, was blunt: "If a project of this scale cannot stack up in the UK, that tells you the underlying costs and constraints are still too high. Nothing material has changed, and that is exactly the problem."

Why London Won Anyway

Days after the Stargate pause, OpenAI signed a lease for 88,500 square feet of office space in King's Cross, London. The location is in the heart of the city's tech hub, alongside Google DeepMind, Meta, the Alan Turing Institute, and the Francis Crick Institute. The office will house over 500 employees, more than doubling OpenAI's current UK staff of roughly 200.

London is ranked second in the world for tech talent, behind only San Francisco. OpenAI is making it the company's largest research center outside headquarters. The move follows the exact playbook Google DeepMind executed a decade ago: establish a London research base, recruit from the UK's deep bench of AI academics, and ship the compute-heavy work to cheaper jurisdictions.

OpenAI is not alone in this calculation. Anthropic announced plans to expand to London as well, securing space for 800 employees. The two biggest US AI labs are racing to hire the same London talent pool while refusing to build the infrastructure to run their models there.

The Sovereign AI Problem

The UK government has been positioning the country as a global AI hub. Prime Minister Starmer's administration treated the Stargate project as a flagship initiative. The pause is a direct blow to that ambition.

The problem is structural. Sovereign AI requires two things: the people who understand the technology and the infrastructure to run it. The UK is about to have an abundance of the first and a shortage of the second. That makes the country a research colony for American AI labs: producing the ideas, training the researchers, and then watching the compute (and the economic value it generates) ship overseas.

Rich Pleeth, CEO of AI logistics platform Finmile, put it simply: "AI is becoming infrastructure. Once you hit that point, decisions are driven as much by energy grids, planning laws, and geopolitics as they are by engineering."

Where the Compute Goes Instead

The answer is wherever power is cheap and politics are friendly. Stargate's US sites are moving ahead, including a large campus in Texas. Companies like Crusoe are building facilities that use energy from natural gas flares instead of the main grid. The Middle East is investing heavily. The Nordics offer cheap hydro power.

The pattern is now clear across the industry: put talent in cities where you can hire fast, and place infrastructure where energy is cheap and abundant. London gets the offices. Texas, Abu Dhabi, and the Nordics get the data centres. The UK gets the brains but not the machines. Until that changes, sovereign AI remains a slogan, not a strategy.

Sources: Reuters, CNBC, TechFundingNews, Institute of Economic Affairs, Chatham House.