Keir Starmer’s resignation on 22 June leaves Andy Burnham likely on course to enter Downing Street within weeks. While Burnham may be less vocal on questions of tech than his predecessor, he takes office at a truly critical moment in shaping the UK’s economic resilience, its geopolitical independence, and its status as a defender of democratic values in the machine age.
Sovereignty reigns
For two decades, the reflex of British tech policy has been to prioritize its special relationship with Washington. The UK’s dependencies on US technology run extremely deep. With them come expectations that this close relationship buys guarantees to market and frontier technology access, intelligence-sharing, investment and a seat at the top table for the UK’s outstanding AI governance and research institutions.
US technology companies have invested heavily in the UK for decades; the Tech Prosperity Deal signed during President Trump’s state visit, which trumpeted £150 billion of future investment into the UK, looks now to be a high-water mark.
That deal is now, reportedly, on ice. Europeans have been further alarmed by the US imposing temporary limits on access to the latest AI models – notably Anthropic’s Mythos 5 and OpenAI’s GPT 5.6 – apparently at the sole discretion of the Oval Office.
The UK-US relationship has always had its difficulties: the two countries are commercially entwined but normatively estranged. The UK public are sceptical of most US technology companies, particularly on questions of tax and safety.
And despite US pressure, the UK has kept its digital services tax and hardened, not softened, online regulation – its June 2026 ban on under-16s using social media goes further than Australia’s. The ban drew a warning from the US Embassy that the rules would burden American firms, echoing the president’s own executive order targeting ‘overseas extortion’ of US tech companies.
UK polling suggests tech sovereignty – as wonkish a subject as there is – is beginning to filter through to the public, with growing calls to curb foreign ownership of infrastructure and maintain the UK’s commitment to online safety.
This friction has accelerated a broader trend across middle powers, who are now realizing that access to US technology – which they previously took for granted – may be less stable or come with more strings than they had hoped. The sovereignty question is hardly Britain’s alone; it animates Paris and Berlin too. Should the White House continue to entangle itself with its frontier technology companies, or should the AI boom undergo a painful course correction, these positions will continue to harden.
Reducing reliance on US technology is reportedly an emerging strand of Burnham’s thinking. So, what should the incoming government do? Three priorities stand out.
Build leverage
The first task is to make Britain AI ready, not just AI-adjacent: the best of the rest, behind the US and China. The returns from cheap intelligence (in the form of AI) won’t be limited by which models the UK can access, but rather by the physical and industrial capacity to turn this cheap intelligence into output: energy, grid connections, factory floors, lab capacity.
Britain should audit its own economy for exactly these blockages and clear them, prioritizing the 2025 Industrial Strategy’s own growth sectors in life sciences, advanced manufacturing and clean energy.
It should also prioritize routes to safely using NHS and other sources of public data in AI development: they remain a potentially decisive asset that years of fragmented, siloed systems and public trust failures have kept locked up, and could be the keystone of a credible UK AI bio and medical research programme.
On compute, the UK cannot, and should not, try to out-build the US or China: money is limited, energy is expensive and data centres are unpopular. The right target is enough sovereign onshore capacity for critical inference – the compute used to deploy and use an AI model – with any surplus committed to UK start-ups, universities and towards a shared middle power technology stack capable of picking some battles with the US and China.
Lead at home
Second, the incoming government should do distinctive things well at home, including making good use of digital identities and AI in public services.
Digital identities have long been treated as a potential liability for civil liberties. Burnham seems to have embraced this position too following the government’s disastrous attempts to relaunch the idea last year. This is the wrong approach. Digital identity should be reframed for what it can and must be: a piece of digital public infrastructure that makes everyday life easier and lays a foundation for public services and new digital businesses.
The same logic applies to AI in public services, where the government should build on the work of Greater Manchester’s AI and Data Innovation Office and work on health data by Health Innovation Manchester. Both prove that modernising public services through the use of digital technology is achievable.
Lastly, the UK should also build on its leadership position in AI governance, where the UK’s AI Security Institute gives it international standing that capital cannot buy.
Pool what you can
The final priority is maximising sovereignty by coalition with other countries rather than hopeless attempts at autarky. Dozens of countries share both the UK’s needs in public service technology and share the anxiety about entrenched dependencies on the current providers. To squander this opportunity by building rickety national silos of technology would be a disaster when the UK has historically shown its capacity to build technology that enables collaboration rather than shutting it off. The success of gov.uk – forked and used by many of Britain’s friends and allies – is just one example.
