The UK's ambitious plan to build national AI infrastructure independent from foreign control is colliding with local opposition, unfulfilled promises, and a fundamental credibility problem: most of the companies the government is backing to achieve sovereignty are still controlled by American firms. Scotland's government is considering a sweeping moratorium on new datacentres, directly threatening a cornerstone of the UK's AI strategy. This crisis reveals that building genuine technological sovereignty requires far more than funding announcements and infrastructure projects.

What Is Happening to the UK's Sovereign AI Strategy?

The Scottish National Party's national council passed a motion last week to freeze all new datacentres in Scotland, and that motion has been sent to the Scottish government for consideration. The move emerged after the Guardian revealed how the developer and UK government misrepresented the technical feasibility of a massive datacentre hub in Lanarkshire, despite community concerns about land use and unfulfilled promises of jobs and investments. This site was supposed to be an "AI growth zone," a key element of the government's strategy to build national AI infrastructure in rural areas of Britain.

The scale of the problem is striking. There are 24 hyperscale datacentre projects in various stages of planning in Scotland. Combined, they would use more than one-and-a-half times the power Scotland uses at peak demand. British officials had promoted Scotland as the prime location for datacentres because of its access to plentiful renewable energy, but the SNP's resolution suggests this capacity is being overwhelmed by overdevelopment.

"It is extreme overdevelopment. I'm very supportive of the local community and their endeavours to prevent this from happening," said Lesley Backhouse, a local councillor from one of the constituencies that put forward the motion.

Lesley Backhouse, Local Councillor

Why Is the UK's Sovereign AI Fund Failing to Deliver Real Independence?

The credibility of the UK's sovereign AI efforts has been undermined by the composition of companies receiving government support. The government released details of nine companies supported under the £500 million Sovereign AI Fund, which was launched in April to back homegrown AI founders. However, four of the nine companies awarded cash investment or offers to use the government's supercomputers in Bristol and Cambridge are ultimately controlled by American firms, according to a freedom of information response.

This contradiction exposes a deeper problem: the UK government has been "very opportunistic" in its approach to AI investment, according to Chi Onwurah, chair of the Commons science and technology select committee. She characterized the process as one where "X or Y or Z says they're going to invest [and so] we'll take that as part of an investment plan and then the investments aren't realised".

"They weren't matched, as I suppose Andy Burnham might put it, by a comprehensive place-based strategy to make sure that happened," Onwurah stated.

Chi Onwurah, Chair of the Commons Science and Technology Select Committee

Several major UK AI projects have been found to be "phantom investments" after the government failed to audit investment numbers or jobs claims. An "AI growth zone" in North Tyneside was more of a publicity stunt than a viable project, despite being supposedly backed by OpenAI, the maker of ChatGPT.

What Are the Geopolitical Risks of Depending on Foreign AI?

The warning from UK lawmakers is unambiguous: dependence on foreign AI providers creates vulnerability to geopolitical pressure. The White House's temporary export ban on access to advanced tools from Anthropic, a leading US AI company, served as a powerful reminder that even close allies cannot be counted on for access to vital technology.

"It is essential to ensure that the UK cannot be cut off from key technologies at the whim of a foreign government," the cross-party committee said.

Commons Science and Technology Select Committee

The committee called on the next government to set out how it intends to protect its sovereignty in AI. Onwurah added: "I hope the incoming administration will learn from the mistakes of its predecessors and move quickly to create a clear plan for how it will work internationally on science and technology".

Onwurah

How Should Governments Approach Real AI Sovereignty?

• Establish Clear Infrastructure Capacity Planning: Governments must conduct rigorous assessments of national energy capacity and resource constraints before approving datacentre projects. Scotland's 24 planned hyperscale datacentres would exceed peak power demand by 50 percent, demonstrating the need for coordinated planning rather than opportunistic project approval.
• Audit and Verify Investment Claims: Government support programs must include mandatory auditing of investment numbers and job creation claims. The UK's experience with phantom investments and misrepresented projects shows that without verification, sovereignty funds become vehicles for corporate marketing rather than genuine national capability building.
• Ensure Genuine Domestic Control: Companies receiving sovereign AI funding should be genuinely independent from foreign control. When four of nine companies in a £500 million sovereignty fund are ultimately controlled by American firms, the program fails its stated objective.
• Develop Comprehensive Place-Based Strategies: Infrastructure projects must be matched with broader economic strategies to ensure communities benefit from promised jobs and investments. Vague promises of development without concrete plans erode public trust and invite local opposition.

The broader challenge facing modern organizations mirrors the UK government's struggle. Companies are increasingly asking whether they should build their own AI systems to maintain sovereignty over critical data and decision-making, rather than relying entirely on massive external platforms controlled by foreign entities. This shift reflects recognition that political and economic developments in the countries where those platforms are based could directly determine a company's future.

For both governments and businesses, the lesson is clear: sovereignty requires genuine independence, not just funding announcements. It demands rigorous planning, verified outcomes, and a willingness to say no to projects that don't serve long-term strategic interests. Without these commitments, sovereignty remains a slogan rather than a reality.