Data centre in Oslo, Norway amid rising AI electricity demand

Norway Moves to Rein In Data Centre Expansion as AI’s Demand for Electricity Surges

The artificial intelligence revolution is running into a problem that cannot be solved simply by developing a more powerful computer chip: electricity. As technology companies race to build larger AI data centres, the demand for reliable power is becoming a serious challenge for national energy systems. Norway has now taken a significant step to address that pressure, temporarily halting new grid-capacity reservations for data centres while preparing tighter rules for the industry.

The decision highlights a growing tension in the global technology sector. Governments want to attract investment, expand digital infrastructure and strengthen their artificial intelligence capabilities. However, they must also ensure that rapidly expanding data centres do not place excessive pressure on electricity networks or undermine the energy needs of other users.

Norway’s latest move is therefore about more than controlling construction projects. It raises a much larger question: how can countries support the growth of artificial intelligence without allowing its energy requirements to outpace the capacity of their power systems?

Why Norway Is Restricting New Data Centre Connections

On November 9, 2026, Norway’s electricity transmission system operator, Statnett, announced a temporary suspension of additional grid-capacity reservations for data centres.

The measure is intended to protect the safe and stable operation of the country’s electricity system as demand for new data-centre connections continues to grow.

According to Statnett, the suspension also affects the processing of certain new applications for approval of facility functionality under Norway’s electricity system regulations. These requirements are important because large data centres can have substantial and sometimes rapidly changing electricity demands.

Statnett said the temporary suspension is expected to remain in place until no later than August 1, 2027.

Importantly, the decision does not mean that Norway has banned artificial intelligence infrastructure or ordered existing data centres to shut down. Instead, it is a temporary measure designed to establish the technical requirements and regulatory framework needed to connect additional facilities safely.

The announcement comes as Norway faces an unusually large pipeline of proposed data-centre developments.

Norway’s Data Centre Demand Has Reached a Significant Scale

Aerial view of Hamar, Norway, near the growing data centre industry
Photo: Wikimedia Commons contributor. Licence: Creative Commons Attribution-ShareAlike 4.0 International (CC BY-SA 4.0).

The figures published by Statnett illustrate why the electricity operator is concerned about the pace of expansion.

According to its November 9 announcement:

  • 800 MW: Data-centre capacity already connected to the electricity grid.
  • 4,200 MW: Capacity that has received grid reservations.
  • 9,100 MW: Proposed data-centre capacity currently in the connection queue.
  • 18,000 MW: Additional connection requests submitted to the operator.

For comparison, Norway’s maximum electricity consumption is slightly above 25,000 MW.

These figures represent different stages of development and should not be interpreted as electricity already being consumed. Some projects are operational, others have reserved capacity, while many remain proposals or applications awaiting further assessment.

Nevertheless, the scale of the requested connections demonstrates how quickly the country’s data-centre ambitions have expanded.

The challenge is not simply whether Norway can generate enough electricity nationally. The location of available power, the capacity of transmission networks and the ability of the electricity system to handle large changes in demand are also important considerations.

Norway’s Government Plans a New Licensing System for Data Centres

Diesel-powered backup generator at a data centre illustrating rising electricity demand
Photo: Mikael Häggström / Wikimedia Commons. Licence: CC0 1.0 Public Domain.

Alongside Statnett’s announcement, the Norwegian government revealed plans to introduce a licensing framework for establishing and operating data centres.

In a statement published on November 9, the Ministry of Digitalisation and Public Governance and the Ministry of Energy said the government wants to ensure that data-centre growth delivers greater benefits to society.

The proposed framework is intended to establish requirements and conditions for both existing and future facilities, with appropriate transitional arrangements.

The government wants to prioritise data centres that contribute to important national interests, including:

  • Digital security and national preparedness.
  • Critical digital infrastructure.
  • Computing capacity needed for artificial intelligence.
  • Economic and technological development.
  • Greater digital independence for Norway and the wider Nordic and European regions.

The government also stated that data centres involved in cryptocurrency operations would not be considered socially beneficial under its proposed approach.

However, the licensing framework is still being developed. The government has not yet completed the full regulatory process, and the details will need further clarification.

Its broader objective is to manage the industry’s expansion rather than encourage unlimited demand for electricity and grid connections.

Why Artificial Intelligence Is Increasing Electricity Demand

Artificial intelligence systems require substantial computing resources, particularly when companies develop and operate advanced models.

Training large AI models involves processing enormous quantities of data through specialised computing hardware. Graphics processing units, specialised AI accelerators and high-performance servers work together to perform these calculations.

Once a model becomes available to users, the infrastructure must continue operating to process requests, generate responses and support applications.

This means AI companies need computing capacity not only during model development but also throughout the lifetime of their products.

Large data centres provide the physical infrastructure for these operations. They contain thousands of servers, networking equipment, storage systems and cooling technologies.

All of these systems require electricity.

The growth of generative AI, cloud computing and AI-powered business applications is therefore increasing the importance of reliable electricity supplies.

However, the relationship between AI growth and electricity demand is not identical everywhere. Consumption depends on the size of the facility, its hardware, utilisation rates, cooling technology and energy efficiency.

Data Centre Cooling Creates Additional Infrastructure Challenges

Cooling tower installed on top of a data centre in Mesa, Arizona
Photo: Rsparks3 / Wikimedia Commons. Licence: CC0 1.0 Public Domain.

Electricity is not the only resource that matters when a data centre expands.

Powerful computing equipment generates heat, and facilities must maintain appropriate operating temperatures to prevent hardware damage and performance problems.

Different operators use different cooling technologies, including air cooling, liquid cooling and other advanced thermal-management systems.

Some cooling arrangements require significant quantities of water, although consumption varies considerably depending on the technology and local climate.

In areas experiencing water shortages, large data-centre developments can raise questions about how industrial and residential water requirements should be balanced.

Cooling systems can also influence electricity consumption. More efficient cooling technologies may reduce the amount of energy needed to operate a facility, but they do not eliminate the underlying demand from computing hardware.

For governments considering new data-centre projects, this makes resource planning increasingly important.

The Global AI Infrastructure Race Is Becoming an Energy Challenge

Norway’s decision reflects a wider issue facing countries seeking to attract major technology investments.

The expansion of artificial intelligence is encouraging companies to build larger computing facilities, secure long-term electricity supplies and invest in specialised infrastructure.

Technology companies are competing to develop more capable AI systems, while cloud providers are expanding their capacity to serve businesses, governments and individual users.

This competition has made access to electricity an increasingly important factor in determining where new data centres can be established.

Countries with reliable energy supplies, strong transmission networks and predictable regulations may have an advantage when attracting infrastructure investment.

However, having substantial electricity generation capacity does not automatically guarantee that every proposed data centre can connect to the grid.

Transmission infrastructure, regional demand, system stability and competing electricity requirements must also be considered.

Norway’s latest restrictions demonstrate why these issues are becoming central to national AI strategies.

Could Other Countries Introduce Similar Restrictions?

Electricity grid substation infrastructure supporting data centre expansion
Photo: Rsparks3 / Wikimedia Commons. Licence: CC0 1.0 Public Domain.

Norway’s decision does not automatically mean that other countries will follow the same approach.

Every national electricity system has different generation resources, transmission capacity, regulatory arrangements and patterns of demand.

Nevertheless, governments around the world are increasingly examining how large data-centre projects affect electricity networks and local communities.

Some jurisdictions may introduce stricter planning requirements, while others could encourage new power generation, energy storage or improved transmission infrastructure.

Another possibility is that data-centre operators will be required to demonstrate greater energy efficiency and provide more detailed information about their expected electricity consumption before receiving approval.

The most effective approach will depend on local conditions.

For countries hoping to become major artificial intelligence and cloud-computing hubs, the challenge will be to balance infrastructure investment with reliable and affordable electricity supplies.

What Norway’s Decision Means for the Future of AI

Norway’s temporary suspension is an important reminder that the development of artificial intelligence depends on more than software, algorithms and advanced processors.

Behind every major AI platform is a physical infrastructure network that requires electricity, cooling, land, equipment and substantial investment.

As demand for computing capacity increases, governments will face difficult decisions about how to allocate limited resources and which infrastructure projects should receive priority.

Norway is attempting to address these concerns through temporary grid restrictions and a proposed licensing system.

Whether its approach becomes a model for other countries remains to be seen. The outcome will depend on how effectively the government develops its new rules and how the data-centre industry responds.

One thing is becoming increasingly clear: the future of artificial intelligence will depend not only on who develops the most powerful models, but also on who can build and operate the infrastructure needed to support them reliably.

Why has Norway temporarily halted new data-centre grid reservations?

Norway’s grid operator, Statnett, introduced the temporary suspension to protect electricity-system stability and develop appropriate technical requirements for connecting additional data centres.

When will Norway’s data-centre restrictions end?

Statnett expects the temporary suspension to be lifted no later than August 1, 2027, subject to the necessary requirements and regulatory framework being established.

How much electricity capacity do data centres already use in Norway?

Statnett reported that 800 MW of data-centre capacity was connected to the grid as of November 9, 2026. A further 4,200 MW had received reservations.

Is Norway banning AI data centres?

No. Norway has not announced a general ban on AI data centres. Its measures are intended to manage new connections and introduce a licensing framework for the industry.

Why do AI data centres consume so much electricity?

They operate large numbers of specialised processors and servers to train and run AI models. Electricity is also required for networking, storage, cooling and other supporting infrastructure.

Could other countries restrict new AI data centres?

Governments may introduce stricter approval requirements or other measures depending on electricity availability, grid capacity and local infrastructure needs. Norway’s decision does not mean that other countries will automatically adopt the same policy.

Scroll to Top