AI minister defends billions amid ‘phantom funding’ scrutiny
Britain’s multi billion-pound push to build sovereign AI capacity will inevitably include projects that are still taking shape rather than fully deployed capital, the UK’s AI minister told Morning Wire, as scrutiny intensifies over the government’s headline investment figures.
Kanishka Narayan, who was appointed as AI and online safety minister last September, said the scale of infrastructure promised across Britain’s AI sector should be understood as a pipeline of projects moving through development, not money already flowing through operational data centres.
“It definitely takes time to get all the bits of building done right”, Narayan said. “You have to have spades in the ground, you have to hire people, you have to line the projects up, you have to build the whole thing, you have to operate it.”
The comments come as questions mount over a series of multibillion-pound AI infrastructure announcements made by government and industry over the past year.
Recent reports by the Guardian suggested that “the money isn’t real, the data centres may not be new” and “the jobs are unaccounted for”.
But, “no one is saying that we’re operating sixty-eight billion or whatever of data centres today as a result of those announcements”, Narayan said.
Instead, he argued the numbers reflect an investment pipeline beginning to materialise across several parts of the UK.
“What we are saying is that we’re making concerted progress,” he said. “We have live data centres in Lanarkshire already. We have spades in the ground in parts of the Northeast”.
“Projects will be at different stages of evolution, but let us be very clear. There are very practical signs of this happening in the real world”.
Sovereign AI drive
This comes as ministers attempt to turn a flurry of announcements, startup fundraises and infrastructure deals into what they describe as a coherent sovereign AI strategy.
Narayan argued the point is not full technological independence but making sure Britain has meaningful leverage over the AI stack.
“I think the way I think about sovereign AI and sovereignty generally is: What are all the things we can do to have strategic leverage as a country when it comes to AI?” he said.
“AI is increasingly the primary aspect of economic power and of hard material power. And so Britain needs leverage in this context.”
That leverage, he argued, will come from selectively backing British strengths while working with allies in other areas.
“I don’t think this is about Britain duplicating every part of the stack”, he added. “It’s about us being selective: where do we have amazing strengths, heritage and the ability to lean into the future in Britain?”
The government is already championing several companies as part of that very effort.
London-based Nscale this week announced a $2bn funding round valuing the firm at $14.6bn, as it builds large-scale compute clusters used to train artificial intelligence models.
Narayan said firms like Nscale are hugely important to this push, because Britain was left without meaningful capability in critical parts of the infrastructure stack during the last wave of tech development.
“For us to have a British-headquartered company as a neo-cloud providing compute clusters for AI is a really important thing,” he said.
Yet, he added that infrastructure alone would not define the UK’s AI future and that chip design, scientific models and applied AI startups would also form part of what he described as a broader ‘jigsaw’ of British strengths.
The prize, in the government’s view, is economic as much as technological. “I think the future of AI is the central economic question for the UK,” Narayan said. “Britain’s central crisis is a crisis of productivity. AI is the best shot that we have to get out of it and launch into a more exciting future”.