Britain risks becoming a major buyer of artificial intelligence rather than a country capable of building and shaping the technology unless it urgently strengthens its domestic AI capabilities, government advisers have warned.
The UK’s Council for Science and Technology said the country remains heavily dependent on a small number of mostly US-based companies for advanced AI models, computing infrastructure and other critical technology.
The council has urged ministers to develop a national AI sovereignty plan covering computing power, data, skills and access to leading models.
It also recommended creating a publicly governed British AI capability that could support researchers, startups and public services.
FCRF’s Chief AI Officer Certification Returns on November 14 to Prepare Future AI Leaders
UK Warned Against Depending on Foreign AI
The Council for Science and Technology advises the government on major scientific and technological issues.
In recommendations published on October 19, it defined AI sovereignty as having enough control across the AI supply chain to make strategic choices and protect national interests.
The council warned that Britain could otherwise become a “high-value AI customer”.
That would mean UK businesses and government departments increasingly using advanced AI while depending on foreign companies to provide the underlying models and infrastructure.
Such dependence could create economic as well as security risks.
If access to advanced systems became restricted because of commercial disputes, export controls or geopolitical tensions, the UK could have limited alternatives.
The advisers therefore want Britain to maintain access to the best global AI systems while also developing capabilities it controls domestically.
Computing Power and Data Become Strategic Assets
One of the council’s central recommendations is greater investment in computing infrastructure.
Training and operating advanced AI models requires large numbers of specialised chips, high-capacity data centres and enormous amounts of electricity.
The companies with access to this infrastructure gain a significant advantage in developing new models.
Most of the world’s most powerful AI systems are currently produced by US companies including OpenAI, Anthropic, Google and Meta.
The UK has strong universities, AI researchers and startups, but it does not currently host a frontier model developer operating at the same scale.
The council says Britain needs stronger foundations if it wants to retain influence as the technology develops.
That includes not only hardware but also high-quality datasets and people capable of building and evaluating advanced systems.
Data can be particularly valuable in fields such as healthcare, science and public administration.
British public institutions hold large amounts of information that could support AI research, but using it safely requires clear governance and privacy protections.
Advisers Propose Publicly Governed British AI
The council also suggested developing a publicly governed UK AI capability.
Such a system could potentially be used by universities, researchers, startups and government departments that cannot afford access to the most expensive commercial models.
The recommendation does not necessarily mean Britain should try to build a direct competitor to every major US AI company.
Instead, advisers argue that the UK should selectively develop capabilities in areas that are strategically important.
Those could include healthcare, national security, scientific research and public services.
The council also recommended stronger testing of AI systems used in sensitive sectors.
Government agencies need to understand how models behave before relying on them for decisions involving national security, healthcare or education.
Testing could include checking whether systems produce unreliable results, expose confidential data or behave unexpectedly when given complex tasks.
One Leader Could Coordinate UK AI Policy
The advisers also raised concerns about fragmented decision-making.
Different departments currently oversee AI infrastructure, regulation, science funding, public-sector adoption and cybersecurity.
The council recommended clearer governance and a single senior figure responsible for coordinating AI sovereignty policy.
That could make it easier to decide where Britain should build domestic technology and where it should rely on international partners.
The government has already invested in AI computing capacity and safety research.
The UK AI Security Institute has become a prominent centre for evaluating advanced models, while Nvidia recently pledged $250 million in computing resources to support British AI safety research.
But the council argues that safety expertise alone will not give the country technological independence.
Britain also needs the capacity to build, operate and adapt important AI systems itself.
The warning reflects a wider geopolitical shift.
AI models, chips and data centres are increasingly being treated as strategic infrastructure alongside energy, telecommunications and defence technologies.
Countries are therefore asking not only whether they can use AI, but whether they retain meaningful control over the systems on which their economies may depend.
What this means for you:
The UK debate shows that AI policy is increasingly about economic independence as much as innovation. Countries that rely entirely on foreign models and computing infrastructure may have less control over costs, access and how the technology is used.
Follow for daily updates on cybercrime, corporate fraud, DFIR, hacking, investigations, and digital forensics
