'Control and sovereignty' to shape UK approach to AI

Principles come after robust discussions at top of government over rollout of fast-evolving technology

The UK government's approach to AI will be shaped by considerations about British interests, control and sovereignty, under principles agreed by ministers following robust cabinet discussions.

Andy Burnham and his senior team last week approved a framework that will govern the domestic adoption of the fast-evolving technology, and the build-out of infrastructure to support it, according to people familiar with the matter.

Seven core principles have been signed off, including that "AI and its adoption must work in the interests of the British people", "control and sovereignty matter" and "infrastructure must serve a national purpose and deliver for communities", the people said.

It comes after the government used the Budget last week to unveil funding for an AI lab to monitor biosecurity threats and a new data centre "covenant" aimed at heading off public opposition to the AI build-out.

The plan, first reported by the FT, will require developers to meet strict environmental and community requirements in exchange for priority access to the grid and fast-tracked planning permission.

The agreement on the UK statement of principles came after a cabinet meeting on October 13, where AI minister Kanishka Narayan updated colleagues on the government's direction.

Burnham said he wanted Britain to work towards a global set of principles and standards on AI, and chaired a discussion on the "risks and opportunities" around the technology, according to a Downing Street readout.

Government figures familiar with the meeting said a frank exchange of views ensued. "It is apparent there are very different views. Everyone is looking at AI through the lens of their own department," said one.

Another stressed that the talks had been "co-operative" but acknowledged the cabinet had "been on a journey", predicting that the prime minister would tread more cautiously on the issue than his predecessor, Sir Keir Starmer.

Culture secretary Lisa Nandy advocated a more circumspect approach during the meeting, the people said, after the creative industries expressed alarm about unauthorised data scraping and weakened copyright protections.

Conflicting opinions about the speed of the technology's rollout have come to the fore within Labour in recent months, alongside grumbling from some quarters of government about how AI policy is divided between departments.

While Nandy's Department for Digital, Culture, Media and Sport was inclined to proceed at a slower pace on AI, the Department for Business, Innovation, Science and Trade was pushing for a faster approach -- though not at the expense of safety -- based on the technology's potential to boost growth, several government figures said.

Official data suggests the AI boom is behind some recent better than expected UK economic figures.

Other officials stressed there was no disagreement on specific AI policies and that the cabinet was aligned in recognition of the opportunities and challenges posed by AI, as well as the importance of ensuring the public was onside.

The cabinet talks about the technology last month took place after a series of ministers requested a formal discussion. Those talks were followed by a strategy day that went into further detail to achieve consensus over the principles.

Questions remain outstanding on UK sovereign capability and capacity, however. Officials are still examining the digital infrastructure footprint required to ensure Britain achieves a baseline level to protect its own security.

Officials are also exploring the appetite and requirements for developing additional capacity and capability to be competitive in elements of frontier AI, such as quantum computing.

Last month Narayan told the FT: "The world of technology is moving so fast that the best defence we have to keep it safe and to do things that work for us is to build state capacity to exercise democratic control."

He promised a "fundamental state partnership with both academia and the private sector to shape much more deeply the trajectory of technology with our values", which he described as the "most distinctly Labour" part of the government's approach to AI in the UK.

Acknowledging voters' anxiety about the technology, Narayan said, "Frankly I share a lot of their concerns," but cautioned that abandoning attempts to build a domestic base in AI was not the answer.

Instead, he argued the government's approach must focus on AI adoption "in the interests of the British public".

Narayan is the first AI minister to attend cabinet. But the elevation of his role under Burnham was undercut by the abolition of the Department for Science, Innovation and Technology, which has folded into the business department. Nandy's department regained oversight of "digital" in the process.

The Institute for Government think-tank warned this year that instead of having a united department behind Narayan that could consider AI in the round, his split responsibilities between different departments would make cross-government co-ordination harder.

Last month Burnham said AI would be the "primary focus" of the UK's presidency of the G20 next year, with Britain hoping to retain a seat at the top table for its testing frontier models via the AI Security Institute.

A government spokesperson highlighted that the chancellor announced £700mn at the Budget last week to strengthen the AI Security Institute, establish a new AI biosecurity lab and build a national cyber shield.

The spokesperson added that Burnham was "determined that Britain leads the way on AI" and had created an AI task force "to realise this ambition . . . driving growth and prosperity, transforming public services and keeping the country safe".