Alan Turing Institute chief urges UK to build its own AI

The UK should stop relying on foreign artificial intelligence that could be withdrawn at short notice and develop its own systems, George Williamson, head of the Alan Turing Institute (ATI), said in remarks reported on October 10, 2026.

Williamson framed the issue as one of control over tools that are moving into the core of the state. Speaking to The Guardian, he argued that Britain needs a firmer domestic footing in the field as the US and China pull ahead, and that “national resilience” will now anchor the institute’s work.

Why access to foreign models looks fragile

“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,”

George Williamson, head of the Alan Turing Institute

His concern is not abstract. Recent decisions in Washington have shown how access can be shaped by politics.

Williamson also named China as a reason to strengthen home-grown capability, voicing concern that Chinese-built systems may not “coincide with our values”. The worry is shared across the Channel: Christine Lagarde, president of the European Central Bank, and Ursula von der Leyen, president of the European Commission, have both recently pressed Europe to become more self-sufficient in AI, with Lagarde warning of the danger of being “cut off” from American and Chinese technology.

An institute under pressure to change course

ATI is the country’s leading AI research body, and the government is its main funder. That relationship drove the current overhaul. In 2025 Peter Kyle, then technology secretary, wrote to the institute’s chair insisting that defence and security become its main focus. Jean Innes, Williamson’s predecessor, stepped down soon after.

Williamson arrives from His Majesty’s Government Communications Centre, a discreet body that builds equipment for British intelligence agencies and once employed Alan Turing himself. He suggested the resilience framing gives ATI room to keep environmental research that Keir Starmer’s government had wanted to push aside. “For me climate is as much a national security issue as defence,” he said.

What national resilience would cover

According to Williamson, the new focus has three strands:

• protecting critical systems from disruption;
• countering threats that come from AI itself;
• helping define the strategic AI capability the UK should have.

He added that the institute could help develop AI applications “for the public good”, from weather forecasting to agriculture, and support efforts to detect disinformation. On state-led campaigns, he said: “It’s very dangerous for the state of democracy and it’s an area that we can monitor.” The context is pointed: last month prime minister Andy Burnham announced a new national centre to fight state-backed disinformation.

Competing without frontier-lab budgets

Williamson accepts that neither ATI nor the UK is likely to match the spending of the biggest AI companies. Instead, he sees room for “nationally owned IP” or smaller models that startups could use. “We can be ambitious here without trying to copy what frontier labs are doing,” he said.

He also wants the institute, a registered charity, to act as an “objective and helpful” arbiter of fears about AI, existential risk included. That stance puts him at odds with how some in the industry talk about danger. Last month, Anthropic put the odds of AI wiping out humanity by 2030 at more than 10%; Williamson said he doubts such percentage estimates rest on convincing evidence.

Featured image. Source: Wikimedia Commons. Credit: Derrick Coetzee from Berkeley, CA, USA. License: CC0.