GOV.UK founder warns Britain is ‘dangerously dependent’ on foreign AI
Britain risks becoming “dangerously dependent” on a handful of foreign tech companies as AI spreads through businesses and public services, the founder of GOV.UK has warned.
Mike Bracken, who founded the Government Digital Service and led the creation of GOV.UK, said organisations were steadily giving up control of critical systems as they turned to external tech and AI providers.
His warning comes as new figures from the Office for National Statistics (ONS) show just how quickly AI is working its way into British business.
Some 35 per cent of UK businesses with at least 10 employees were using at least one form of AI in June, up from around 12 per cent in 2023. Among companies with more than 250 staff, the figure has already reached 49 per cent.
Yet relatively little of that technology is being built in-house, or in the UK at all. Of businesses using AI, 45 per cent relied on free-to-use software and 43 per cent bought external or ready-made software, while just 12 per cent developed it themselves.
“Institutions rarely lose sovereignty in a crisis. They lose it one reasonable decision at a time” Bracken said. “The biggest risk facing many organisations today is not that somebody takes control away from them. It’s that they gradually give it away”.
The ONS itself highlighted the problem in its first steps towards measuring AI’s contribution to the economy.
It noted that the leading foundation-model providers, the likes of Anthropic, Google and OpenAI, are American, meaning much of the technology used by British companies is likely to be accessed through licensing agreements rather than owned as a UK asset.
The same problem exists further down the stack, where most of the specialist chips needed to run advanced AI are produced outside Britain and imported.
Britain’s sovereign AI push
The dependency question has shot up the political agenda as ministers try to work out which parts of the AI economy Britain actually needs to control.
The government has committed £500m to its Sovereign AI programme and £1.1bn to an AI Hardware Plan aimed at strengthening domestic capability in the chips and infrastructure underpinning the technology.
In its response to the Science and Technology Committee last week, the government acknowledged the risks posed by highly concentrated markets and “single points of failure”, but said sovereignty should mean diversification, building domestic capability in strategic areas and retaining access through trusted international partners.
Bracken said organisations often discover the problem only when they try to switch supplier or change direction.
“Every technology decision can make it harder or easier to change direction in the future”, he said. “Too often, institutions only discover they have lost that flexibility when circumstances demand it”.
MPs and peers have increasingly questioned how Britain should protect critical systems when cloud computing, data and AI are concentrated among a small group of global companies.
Recent parliamentary evidence has also highlighted the resilience risk created by technological “single points of failure”.
There is a domestic industry for ministers to work with, although the ONS figures suggest its strengths currently sit more in applying AI than building the underlying models.
Its analysis of roughly 5,860 UK AI companies found data analytics and decision support was the most common activity, appearing in 38.4 per cent of company descriptions. AI consulting and digital transformation appeared in 30.7 per cent.
At the other end of the scale, just 5.9 per cent related to AI platforms, model development and data infrastructure.