This week, British Prime Minister Andy Burnham assumed office promising major reform against a backdrop of strained public services, an unmanageable cost of living, and low trust in politicians. In his first speech as prime minister, he promised his government would be a “circuit breaker for Britain, bringing forward the biggest changes in the last forty years.”

Burnham has no shortage of work if he wants to deliver on his philosophy of “Manchesterism,” which focuses on regional investment and decentralizing governance beyond Westminster. This is true in any number of policy areas, including AI.

By the government’s own admission, some data center projects are facing waits of fifteen years amid severe backlogs and oversubscribed queues for grid connections. The UK also has some of the most expensive industrial electricity in Europe, with prices three times American levels. While these delays persist, 80 percent of UK operational data center capacity remains concentrated in London, and the promise of thousands of new jobs and billions in new investment to the UK’s northeast region remains unfulfilled. 

The challenge for Burnham on AI will come in balancing these domestic priorities with a sharpening geopolitical environment. Advanced AI development is concentrated in just a few companies in the United States and China, and it is not clear whether and on what terms the UK will have access to capabilities that will fundamentally transform economic potential and national power. As newly promoted Minister for AI Kanishka Narayan warned on Monday, “Nations have a narrow window to decide whether they shape AI or get shaped by it. Britain is in that window right now.”

Whether Britain is successful will depend largely on the choices the Burnham government makes in the first few months of its tenure.

Fortunately, the prime minister inherits many of the tools to get this right. Across multiple Conservative and Labour governments, the UK has spent the past three years developing capacity to nurture domestic AI assets and punch above its weight in international conversations. Specialized strengths, such as those in developing AI models specifically designed for scientific research, have drawn U.S. and Chinese companies such as Alphabet’s Isomorphic Labs to do drug-discovery research in the UK. And a growing cohort of innovative AI infrastructure companies like Fractile and OLIX is blossoming at home—some of which have been backed and scaled by public money through the UK’s Sovereign AI Fund or its billion-pound “moonshot” agency ARIA. 

Backing this type of company can itself become a driver of localization, bolstering specialized manufacturing capabilities in chips and models to drive UK economic growth. But just as important is recognizing these companies’ potential for providing geopolitical leverage to the UK—something to offer the United States and China in return for their capabilities. Thinking geopolitically, and helping the UK navigate a fraught international environment, will leave the country less susceptible to global shocks that would impede growth across the country.

Yet the UK’s industry is not alone in providing this sort of value. Britain has been a leader in international conversations on AI governance, shaping discussions about how to navigate the technology’s future. Its anchor here has been the AI Security Institute (AISI), which tests the most powerful AI models before they are released, in the same way one might crash-test a car or trial a new drug. 

So trusted is AISI that it was the only non-U.S. body granted early access to Anthropic’s Mythos models earlier this year, in order to discover vulnerabilities before those systems are deployed into the critical infrastructure the country relies on. It has convened over 100 of the world’s leading AI experts, backed by an expert advisory panel nominated by more than thirty governments, to write the International AI Safety Report. In so doing, the UK has shaped the state of the science of advanced AI and positioned Britain as one of the world’s deepest reserves of AI expertise.

 This is not just a question of prestige and clout: Privileged model access is an enabler for national security and growth, and a national AI assurance industry employing more than 12,500 people has spun up around AISI. As the newly expanded Department for Business, Innovation, Science, and Trade navigates Britain through the advanced AI era, it should recognize that strong industrial assets and geopolitical leverage are mutually reinforcing. 

Burnham has promised the “biggest changes in the last forty years.” An unprecedented technology will demand nothing less. Supporting and expanding unique areas of British excellence and ensuring the UK can navigate a tumultuous geopolitical environment is not antithetical to delivering for local communities. In focusing on one, the government should not forget the benefits of the other.