UK must develop sovereign AI to avoid foreign dependence, says Alan Turing Institute head

The UK must build its own artificial intelligence capabilities to avoid dependency on foreign systems that could be restricted or shut off without warning, according to George Williamson, the new head of the Alan Turing Institute. His call for "national resilience" comes as the US and China extend their lead in AI development, raising urgent questions about sovereignty and access for governments, healthcare systems, and critical infrastructure operators.

"As AI becomes more embedded in critical services, in domains like defence, national security, critical infrastructure and healthcare we shouldn't depend solely on systems we can't inspect or control, or that could be restricted or even switched off," Williamson said in an interview. He pointed to Washington's temporary block on the release of Anthropic's latest models outside the US in June, and reports that American AI labs were asked to withhold models from the UK's AI Security Institute until US counterparts had examined them first.

"We've seen now that both political and commercial pressures mean we may not have access in all the circumstances when we want it. That is not a place where a nation wants to be," he said.

A shift toward sovereign capability

Williamson, former head of His Majesty's Government Communications Centre - the secretive body that builds technology for UK spy agencies and once employed Alan Turing himself - said the country needs "more agency on access to technology." The ATI, which is the UK's leading AI research body, is refocusing under government pressure after the previous technology secretary demanded a pivot toward defence and security work.

"It feels pretty uncomfortable if you don't have some control or ability to use technologies when you want them, and also harder to imagine how you build an economy on the top of stuff that can disappear," Williamson said. His comments echo recent warnings from European Central Bank president Christine Lagarde and European Commission president Ursula von der Leyen, both of whom have urged Europe to pursue greater self-sufficiency in AI rather than relying on US and Chinese technology.

Williamson cited China's AI systems as another reason for domestic investment, noting that Chinese-made models might not "coincide with our values." He said: "Try asking some of these models about some events in the recent past in China and you're not going to get an answer that's particularly helpful if you're doing GCSE history."

Building resilience without matching frontier labs

The ATI does not aim to replicate the work of major AI labs with their vast financial resources. Instead, Williamson said the institute can produce "nationally owned IP" and smaller models that could be made available to startups. "We can be ambitious here without trying to copy what frontier labs are doing," he said.

The resilience agenda will span areas from weather forecasting and agriculture to countering disinformation. Williamson described state-led disinformation campaigns as "very dangerous for the state of democracy" and said monitoring them is work the institute can pursue. The focus on national resilience also allows ATI to continue environmental research that the current government had sought to sideline. "For me climate is as much a national security issue as defence," Williamson said.

On the debate around existential risks from AI, Williamson pushed back against the practice of assigning percentage probabilities to extinction scenarios. "I'm not sure it's helpful trying to assign particular percentages to it because I have not seen anyone show me a convincing evidence of how they've arrived at the number they then start throwing around," he said. Anthropic recently predicted a greater than 10% chance of AI wiping out humanity by 2030.

Why this matters for executives and government leaders

Williamson's warning signals a hardening view among UK and European officials that AI supply chains now carry geopolitical risk on par with energy or semiconductor dependence. For executives in defence, healthcare, and critical infrastructure, the prospect of foreign-controlled AI systems being restricted or revoked means procurement strategies must account for sovereign alternatives. Government leaders face parallel pressure: building domestic capability while funding institutes like ATI to deliver AI for Government Courses and tools that do not depend on permissions from Washington or Beijing. Strategy professionals should track whether ATI's push for nationally owned intellectual property translates into usable models for industry, and whether AI for Executives Courses begin incorporating resilience planning as a core competency rather than an afterthought.