
On Monday, prime minister Keir Starmer said the government will “mainline AI into the veins” of the UK as he unveiled the flagship set of policies forming the UK’s AI Opportunities Action Plan.
The plan outlines uses of AI that range from the mundane – such as using AI to identify potholes – to the more ambitious vision of making the UK an AI “superpower”.
Despite the UK’s poor financial health, the government claims embracing AI will “lead to more money in the pockets of working people”, citing IMF estimates that its use could boost productivity by 1.5 percentage points every year. Ministers estimate this could provide an economic boost of £47bn to the UK each year over the next decade.
Peter Kyle, secretary of state for science, technology and innovation, also gave assurances the government’s AI proposals will “open up new opportunities” rather than threatening “traditional patterns of work”.
AI growth zones and energy council
Under the plans, the government will create “AI growth zones” (AIGZ) – geographic areas where restrictions will be relaxed on planning for AI infrastructure, such as data centres. The first AIGZ will be in Culham, Oxfordshire.
The government says these areas will also serve as a testing ground to explore how future sustainable energy systems, such nuclear fusion, might be able to power artificial intelligence.
Generative AI consumes large amounts of energy – creating an AI image uses the same amount of power as charging a phone. With relaxed planning laws, Labour has already opened the doors for more data centre construction.
To address the issues surrounding energy demand, the UK is setting up a new AI Energy Council, chaired by the Science and Energy Secretaries. It will work with energy companies to better understand the energy demands and challenges of AI and if miniature nuclear reactors can help solve the problem.
‘Sovereign’ compute power
The UK will commission a new supercomputer, to join two others at the University of Bristol, having scrapped plans for a supercomputer in Edinburgh set out by the previous government. This new machine will be at least as powerful as the UK’s current fastest supercomputer, Isambard-AI.
This commitment will be part of wider investment in government-owned AI computing power, as recommended by Matt Clifford, chair of UK research funding agency ARIA and an adviser to the prime minister on AI opportunities.
Clifford – who advised the previous Conservative government on tech policy and was pivotal to establishing the UK’s AI Safety Institute and AI Safety Summit – submitted 50 proposals to the government to help achieve its AI ambitions. Almost all of them have been accepted, including a recommendation to build the equivalent of 100,000 GPUs – graphics processing units, which are essential for AI – by 2030.
Data sovereignty is an increasingly important area of focus for countries as they seek to reduce their reliance on businesses headquartered elsewhere.
The government states its AI computer will be “owned and/or allocated by the public sector”. It hopes state ownership of this technology will allow the government to quickly and independently allocate computing power to national priorities – whether “mission-focused AI research” or ensuring access to AI in the event of “market disruption”.
This funding is separate to the private AI infrastructure investment being courted by the government. The Department for Science, Innovation and Technology (DSIT) will publish its long-term computing strategy in Spring this year.
A shift in AI regulation
The UK also has the “freedom” to diverge from the approaches to AI regulation set out by the US and the EU, according to Starmer. “We have the ability to choose the one that we think is in our best interest and we intend to do so,” he said in his speech.
Multinational law firm Pinsent Mason believes this could signify a shakeup of the UK’s approach to AI regulation. Clifford’s report suggests introducing AI sandboxes into areas with “regulatory challenges but high-growth potential”, such as where AI interacts with the physical world – such as autonomous vehicles, drones and robotics.
The recommendations also included overhauling the UK’s approach to copyright, so that it is more competitive with the US. The UK has already made some inroads here, publishing a consultation on AI and copyright reform in December last year.
Part of this will include a copyright-cleared dataset of British media that can be used to train large language models, the algorithms upon which generative AI platforms are built. That might mean partnering with British institutions such as the BBC and the National Archives.
The tranche of announcements demonstrate the government’s optimism about AI’s potential and contrasts with former prime minister Rishi Sunak’s outlook, which was much more focused on AI safety. Starmer will hope that his approach to AI brings with it the economic growth the UK needs.

On Monday, prime minister Keir Starmer said the government will “mainline AI into the veins” of the UK as he unveiled the flagship set of policies forming the UK’s AI Opportunities Action Plan.
The plan outlines uses of AI that range from the mundane – such as using AI to identify potholes – to the more ambitious vision of making the UK an AI “superpower”.
Despite the UK’s poor financial health, the government claims embracing AI will “lead to more money in the pockets of working people”, citing IMF estimates that its use could boost productivity by 1.5 percentage points every year. Ministers estimate this could provide an economic boost of £47bn to the UK each year over the next decade.