16th June 2026
For years, the technology industry measured progress by faster computer chips.
Today, there is a new bottleneck - Electricity.
As companies race to build ever more powerful artificial intelligence (AI) systems, they also need huge new data centres to house the computers. The problem is that these facilities consume extraordinary amounts of electricity—often as much as a medium-sized town.
Building a data centre can take two or three years.
Building the power stations and electricity networks to supply it can take ten years or more.
That mismatch is becoming one of the biggest challenges facing the AI revolution.
Why does AI use so much electricity?
Traditional data centres mainly stored information and hosted websites.
AI is very different.
Training and running large AI models requires tens of thousands of specialised computer processors operating around the clock.
A single hyperscale AI data centre may require hundreds of megawatts of electricity. The largest proposed campuses are approaching one gigawatt—roughly enough to power hundreds of thousands of homes.
The United States is taking a different approach
America's largest technology companies—including Microsoft, Amazon, Google and Meta—are increasingly concluding that waiting for electricity companies is no longer an option.
Instead they are:
investing directly in new nuclear power
signing long-term renewable energy contracts
building solar and wind farms
installing battery storage
in some cases planning gas-fired power stations next to their data centres.
The aim is simple.
If the national grid cannot supply enough electricity quickly enough, build your own.
What is happening in the UK?
Britain faces a more difficult situation.
The UK has excellent internet connections and attracts global technology investment.
But it also has:
some of Europe's highest industrial electricity prices
long waiting times for National Grid connections
planning delays
limited spare capacity in parts of the electricity network.
As a result, electricity availability is becoming one of the deciding factors in whether new AI data centres are built.
The UK Government has responded by creating AI Growth Zones, where planning and infrastructure are intended to be accelerated to attract investment.
Why have some projects already stalled?
The problem is no longer attracting investors.
It is supplying enough affordable electricity.
Earlier this year, OpenAI paused its proposed Stargate UK project, saying high energy costs and regulatory uncertainty meant conditions were not yet right for long-term investment.
That decision highlighted a growing reality:
Money alone cannot build AI infrastructure if the power is unavailable.
What about Scotland?
Scotland could become one of Britain's biggest opportunities.
It has:
abundant renewable electricity
significant hydro-electric generation
growing offshore wind capacity
cooler temperatures that reduce cooling costs
land available outside crowded urban areas.
Several new data centres have been proposed or are under consideration across Scotland, although relatively few have yet reached construction.
Many industry experts believe Scotland could eventually become one of Europe's preferred AI locations—provided electricity transmission and grid capacity are expanded.
For areas such as the Highlands, this could present both opportunities and challenges.
So why did Norway say "enough"?
Norway has long been one of Europe's favourite locations for data centres.
It offers:
plentiful hydro-electric power
relatively low electricity prices
a cool climate
political stability.
But success has brought new problems.
The Norwegian Government has become increasingly concerned that very large AI data centres consume electricity that could otherwise be used by households and manufacturing industries.
Officials have also questioned whether some facilities create enough local jobs to justify their enormous energy demand.
As a result, Norway has tightened planning rules and indicated that future projects must demonstrate clear economic benefits before approval.
In simple terms, Norway is saying:
"If our electricity is limited, should it power AI computers—or industries that employ more people?"
That is a question many other countries may soon face.
A global competition for power
Electricity is rapidly becoming one of the world's most valuable economic resources.
Countries able to offer:
abundant electricity
reliable grids
affordable energy
faster planning approvals
are likely to attract the next generation of AI investment.
Those that cannot may watch billions of pounds of investment move elsewhere.
The Bigger Picture
The AI revolution is often described as a battle over computer chips.
Increasingly, it is becoming a battle over electricity.
Whoever can generate the cleanest, cheapest and most reliable power may become tomorrow's AI superpower.
That is why governments are no longer talking only about digital infrastructure.
They are also talking about power stations, transmission lines, nuclear reactors, offshore wind farms and battery storage.
In the age of artificial intelligence, electricity is becoming the new strategic resource—just as oil was in the twentieth century.