New research from CUDO Compute shows how rising energy costs and limited infrastructure are pushing British companies to run AI systems outside the country.
20% of UK firms have already moved AI workloads overseas because of high power costs. This comes as many companies say they would prefer to keep operations at home within the UK.
In the study, 46% of UK organisations say geopolitical instability is pushing them to keep AI within domestic borders. Cost and performance take priority for 43% of UK organisations when choosing where AI should run. This leaves a disconnect between stated preferences and what companies do when budgets come into play.
Matt Hawkins, CEO of CUDO Compute, spoke on the situation. He said, 鈥淎I sovereignty is being hotly discussed as a priority for UK organisations, but it only works if the infrastructure exists to support it.鈥�
How Much Are Energy Costs Affecting Decisions?
Energy pricing has become a constraint for companies building AI systems across the UK economy. According to the report, 33% of UK organisations say energy costs are limiting their ability to scale AI operations properly.
The data gives more insight into that trend across different regions. 37% of UK respondents say volatility in energy prices is increasing instability for business planning. 31% say long term pricing makes investment difficult to justify across AI projects. 22% say they have already scaled back activity due to energy costs rising.
These numbers come from the CUDO Compute research across the UK, EU and US markets. There is also a budget element influencing how companies think about infrastructure spending decisions. 26% of UK organisations say energy already accounts for more than 25% of their AI infrastructure budget.
That level of spending changes how companies choose locations for running workloads. Running workloads in regions with lower and more stable energy costs becomes an easier option.
Hawkins explained the physical side behind these numbers in more detail. He said, 鈥淎I is not abstract software. It is physical infrastructure that depends on power, land, cooling and grid access.鈥�
More from News
- What Do The April 2026 ONS Market Figures Mean For UK Businesses?
- FinanceWire And Symex Global Partner To Boost PR And IR Reach For Euronext Paris Companies
- Could You Be Answering A Normal Call When It鈥檚 Actually A Deepfake?
- Do People Trust AI More Than They Trust Humans?
- What Will Happen If EU Regulators Win At Getting Google To Share Its Data?
- Uber Eats Makes Influencers Central To Its UK Growth Strategy
- It Sounds Ridiculous, So Why Is Allbirds鈥� AI Pivot Actually Working?
- Tanzania Is Dealing With Digital Fraud Through Legislation – What Are The Changes?
Which Countries Are Attracting UK Workloads?
The research also tracks where UK firms are sending their AI workloads as costs rise. The United States ranks highest, with 72% of respondents viewing it as an attractive location for new AI capacity.
India follows at 62% and Eastern Europe reaches 58% among surveyed respondents. Western Europe scores 45% and the Nordics reach 44% in the same data. China stands at 55%, placing it above regions such as Latin America at 40%, the Middle East at 39%, Africa at 38% and APAC at 29%.
These preferences show how companies are choosing locations where energy, land and infrastructure are easier to access at a lower cost. That often outweighs the desire to keep operations within the UK domestic market.
32% of UK organisations say they are actively considering relocating workloads due to geopolitical pressures. This adds another element to how companies choose where to run AI systems at scale.
Can The UK Keep AI Work At Home?
The data suggests that many companies would prefer to build and run AI systems in the UK if conditions allowed it. 31% say they are prioritising sovereign or regionally controlled compute, even when it comes at a higher cost.
45% say data sovereignty, regulatory compliance or national security concerns are influencing their choices. This creates a complicated situation for businesses trying to manage cost and control at the same time.
Access to land, power and grid capacity continues to hold back expansion across the UK. Hawkins addressed this disconnect in his comments on the report findings.
He said, 鈥淲hat we are seeing is a growing tension between where businesses want to run AI and where they actually can.鈥�
He added another observation that puts all the findings together. He said, 鈥淩ight now, every UK boardroom is talking about AI, but almost nobody is talking about the infrastructure needed to power it.鈥�
The study done with Censuswide, is based on responses from more than 700 senior AI decision makers across the UK, US and Europe.
