Key Points
- Computer Weekly, in a report by Managing Editor Cliff Saran, says high UK electricity costs and weaknesses in domestic technology manufacturing are limiting the country’s ability to develop advanced technologies for national security and defence.
- The Joint Committee on the National Security Strategy examined the UK’s preparedness for artificial intelligence (AI) and quantum computing in a public evidence session on 7 September 2026.
- The 2025 National Security Strategy committed the government to strengthening UK capabilities in AI and quantum computing for security purposes. George Balston, strategy adviser at The Alan Turing Institute, raised concerns about Britain’s ability to keep pace with AI while maintaining technological sovereignty.
- Balston warned that frontier AI has applications across national security, including data collection, analysis, investigations and intelligence gathering, making dependence on foreign technology a sovereignty risk.
- The UK has AI supercomputing infrastructure, including Isambard-AI in Bristol, but Balston said much of the country’s existing capacity is focused on research rather than the inference workloads required to operate powerful AI models.
- UK electricity costs were described as more than three times those in the US, creating a significant economic barrier to large-scale AI compute infrastructure.
- Kenneth Payne, professor of strategy at King’s College London, said high energy costs restrict the amount of computing available for AI training and inference.
- Payne also highlighted Britain’s dependence on overseas suppliers for critical hardware, including high-end computer chips and semiconductor manufacturing equipment.
- The UK does not control major parts of the AI hardware supply chain, while companies such as ASML in the Netherlands supply critical lithography equipment used to manufacture advanced chips.
- Payne suggested that cooperation among “middle power” nations could help countries such as the UK reduce dependence on the US and China, although future changes in computing could alter today’s technological balance.
- Michael Cuthbert, director of the National Quantum Computing Centre, said the UK remains a strong third player in quantum computing behind the US and China, following a mature national programme developed over roughly 12 years.
- Cuthbert nevertheless identified device fabrication as a weakness, noting that many UK clean rooms remain focused on academic research.
- Britain’s photonics industry, particularly in Scotland’s central belt, was identified as an important potential strength for future quantum technologies.
- Cuthbert also highlighted UK expertise in quantum software, error correction, applications and algorithms, but said these capabilities will require continued nurturing.
Why is the UK struggling to build sovereign advanced technology for defence?
The UK is facing a combination of high energy costs, limited domestic hardware manufacturing and insufficient computing capacity as it seeks to strengthen the use of artificial intelligence and quantum computing in national security.
The issue emerged during a 7 September evidence session of the Joint Committee on the National Security Strategy, which examined “AI and quantum computing in the national security context”. The UK Parliament records the witnesses as George Balston, strategy adviser at The Alan Turing Institute; Kenneth Payne, professor of strategy at King’s College London; and Dr Michael Cuthbert, director of the National Quantum Computing Centre.
The session is significant because the government’s 2025 National Security Strategy had already committed to increasing the UK’s capabilities in AI and quantum computing for security.
As reported by Cliff Saran of Computer Weekly, Balston expressed concern about Britain’s ability to keep pace with AI and linked that challenge directly to the question of sovereignty. His warning centred on whether the UK could maintain sufficient control over critical AI capabilities rather than becoming dependent on systems, infrastructure and suppliers controlled elsewhere.
For national security, that dependence can extend beyond ordinary commercial technology. Frontier AI can be used for activities such as intelligence gathering, investigations, data analysis and other defence-related functions. This makes access to computing infrastructure and advanced models a strategic issue as well as an economic one.
How are high electricity prices limiting UK AI capability?
One of the clearest obstacles identified during the session was the cost of electricity required to operate large-scale computing infrastructure.
According to Cliff Saran’s Computer Weekly report, Balston said UK electricity costs are more than three times higher than those in the United States. He illustrated the difference by comparing the cost of operating a one-gigawatt data centre: approximately $800m in the US compared with $2.5bn in the UK.
The comparison highlights why the economics of AI infrastructure are becoming increasingly important to national security.
Large AI models require substantial computing resources for both training and inference. Training involves developing or refining models, while inference refers to using trained models to generate results. As advanced AI becomes embedded in defence, intelligence and security systems, the ability to provide inference at scale becomes strategically important.
The UK already has major supercomputing capacity. Isambard-AI in Bristol, for example, represents a significant national investment in AI computing. However, Balston said this infrastructure is heavily focused on research, while the UK lacks sufficient inference capacity.
That distinction matters because possessing a powerful research supercomputer does not automatically provide the infrastructure required to operate AI systems at the scale potentially needed for national security applications.
Why does the UK lack control over the AI hardware supply chain?
Energy costs are only one part of the problem. The UK also lacks control over several critical stages of the hardware supply chain required to expand AI computing.
Kenneth Payne of King’s College London, according to Computer Weekly, told the committee that Britain’s high energy costs compared with countries such as France restrict the amount of computing available for AI model training and inference.
Payne also pointed to the country’s lack of control over key hardware supply chains.
The UK does not manufacture the high-end computer chips required by advanced AI models. That means expanding domestic computing capacity can depend on international suppliers and the availability of advanced semiconductor technologies.
Payne specifically referred to ASML, the Dutch company whose lithography equipment is used in semiconductor manufacturing. His comments illustrated the broader challenge facing Britain: even when the country wants to expand domestic AI infrastructure, it does not necessarily control the critical technologies required to construct that infrastructure.
This creates what Payne described as a problem around the strategic “choke points” of the technology stack.
For a country seeking greater technological sovereignty, the issue is therefore not simply whether it owns data centres. It is whether it can secure the chips, manufacturing equipment, energy and other infrastructure needed to keep those data centres operational and competitive.
Could partnerships with other countries reduce Britain’s technology dependence?
International partnerships were also discussed as a potential response to the concentration of advanced technology capabilities in a relatively small number of countries.
Computer Weekly reported that former technology secretary Liz Kendall had argued in April that Britain should work more closely with international partners, particularly “middle power nations”, to avoid becoming overly dependent on technology from a single country.
Payne was asked about whether cooperation among middle powers could help develop frontier AI capabilities capable of competing with the US and China.
His response suggested that the current technological landscape may not remain unchanged. Graphics processor-based AI accelerators, including those produced by companies such as Nvidia, currently play a central role in advanced AI computing. However, Payne said future changes in computing could alter the importance of today’s dominant technologies.
That creates both an opportunity and a risk for countries attempting to establish technological independence.
Payne also pointed to Mistral, the French AI company, as an example of the difficulty involved in maintaining pace with frontier AI development through a home-grown model. The implication is that developing an alternative to the largest US and Chinese AI systems is not simply a matter of securing investment; maintaining technological leadership requires sustained access to computing, talent, infrastructure and hardware.
What role can quantum computing play in UK national security?
Quantum computing represents a different stage of technological development, and witnesses indicated that Britain has some important strengths in this field.
Dr Michael Cuthbert, director of the National Quantum Computing Centre, told the committee that the UK remains a strong third player behind the United States and China.
The UK Parliament confirms that Cuthbert appeared alongside Balston and Payne at the 7 September session focused specifically on AI and quantum computing in national security. Cuthbert said Britain’s position was supported by a mature national quantum programme developed over approximately 12 years.
However, he also identified a major weakness: device fabrication.
According to Computer Weekly, Cuthbert said many UK clean rooms are primarily focused on academic research. That creates a gap between research capability and the industrial manufacturing capability required to develop quantum technologies at scale.
The problem resembles some of the UK’s challenges in AI hardware. Strong research does not necessarily translate into sovereign manufacturing capacity.
For national security applications, the ability to develop, manufacture and maintain critical technology domestically can be as important as the ability to conduct world-leading research.
Which UK quantum capabilities could provide a strategic advantage?
Despite those weaknesses, Cuthbert highlighted areas in which Britain has significant strengths.
One is the UK’s photonics industry, particularly in Scotland’s central belt. Cuthbert described photonics as an important enabler for quantum technologies in the longer term.
The second is quantum software.
Computer Weekly reported that Cuthbert highlighted UK strengths in quantum software, including error correction, applications and algorithms, alongside a strong research base.
Quantum error correction is particularly important because practical quantum computing requires methods for dealing with errors that arise during quantum operations.
Cuthbert nevertheless stressed that these strengths cannot simply be assumed to remain competitive. He told MPs that the UK’s capabilities will need to be nurtured.
That means maintaining investment in research, developing industrial capabilities, building specialist skills and ensuring that promising technologies can move from academic environments into practical applications.
What does the evidence mean for UK defence and national security?
The evidence presented to the committee suggests that Britain’s advanced technology challenge is broader than simply increasing spending on AI or quantum computing.
The country needs access to affordable electricity, sufficient computing infrastructure, advanced chips, semiconductor equipment, manufacturing capacity and specialist talent.
The AI challenge is particularly acute because frontier models depend on increasingly large amounts of computing infrastructure. If the cost of operating that infrastructure is substantially higher in Britain than in competing countries, UK organisations may face an economic disadvantage when attempting to scale domestic capability.
At the same time, dependence on foreign suppliers can create strategic vulnerabilities.
The issue is therefore closely connected to technological sovereignty. A country can have world-class universities, researchers and technology companies while still depending on foreign infrastructure for critical components.
The committee session demonstrated that the UK has genuine strengths, particularly in research, quantum software and photonics. But the witnesses also identified gaps between research capability and sovereign industrial capacity.
The government’s challenge will be to determine how those strengths can be developed while reducing exposure to energy costs and foreign supply-chain dependencies.
What could the UK do to strengthen its advanced technology strategy?
The evidence does not point to a single solution. Instead, it highlights several interconnected areas that policymakers will need to consider.
First, energy costs remain fundamental. AI infrastructure requires large amounts of electricity, making the cost and availability of power directly relevant to the country’s ability to develop computing capacity.
Second, Britain faces a hardware challenge. The country does not manufacture the high-end chips needed for leading AI systems and lacks control over some of the most strategically important parts of the semiconductor supply chain.
Third, the UK needs to consider how research infrastructure can be complemented by commercial-scale inference capacity. The existence of systems such as Isambard-AI demonstrates national computing capability, but Balston’s evidence indicates that research capacity alone does not solve the inference requirement.
Fourth, international partnerships could provide an additional route towards resilience. Cooperation with other middle powers could potentially diversify Britain’s technological relationships, although Payne’s evidence also indicates that today’s AI architecture may change as new computing approaches emerge.
Finally, the UK will need to maintain its strengths in quantum technology. Cuthbert’s evidence suggests that Britain has a substantial foundation, but weaknesses in device fabrication mean that research leadership needs to be matched by stronger industrial capabilities.
Why does advanced technology matter to Britain’s future security?
AI and quantum computing are increasingly connected to national security because they can influence intelligence, cyber operations, data analysis, decision-making and other defence capabilities.
The 7 September parliamentary evidence session therefore examined a strategic question: whether Britain can build enough domestic capability to use these technologies securely while retaining meaningful control over the infrastructure on which they depend.
The evidence presented by George Balston, Kenneth Payne and Michael Cuthbert does not suggest that the UK lacks technological strengths. Instead, it shows that those strengths exist alongside structural weaknesses in energy costs, hardware supply chains, inference capacity and device fabrication.
For Britain’s national security strategy, the challenge is consequently one of turning research and expertise into resilient, scalable and sufficiently sovereign technology capabilities.
As the country develops its approach to AI, quantum computing and national security, expertise in Artificial Intelligence (AI), IT and Computer Science, Power Systems and Maintenance, and Safety and Security can be relevant to understanding the infrastructure, technology, energy and risk-management issues highlighted by the parliamentary evidence session.