How can power system balancing be improved? The UK looks to AI

On 8 September 2026, the UK government launched a consultation on a document intended to help shape a national AI strategy for the energy sector. It explores the use of artificial intelligence in forecasting, grid management and the coordination of millions of distributed devices. The scale of the challenge is illustrated by figures from June last year, when the system operator had to issue more than 200,000 instructions to balance the system—ten times as many as in December 2023.

Why has the number of instructions issued by the operator increased tenfold?

The UK’s electricity system is becoming increasingly decentralised. The operator must coordinate renewable energy sources, energy storage systems, electric vehicles, heat pumps and other flexible devices. Figures from the National Energy System Operator (NESO) illustrate the scale of this shift. In December 2023, around 20,000 instructions were issued over the month. By June 2025, that figure had exceeded 200,000. These were dispatch instructions issued by NESO’s control room to market participants to balance the system.

The launch of the Open Balancing Platform in December 2023 also contributed to the increase. The platform enables instructions to be issued more efficiently to energy storage systems and smaller units, increasing their participation in system balancing.

The UK government notes that a growing number of devices and decisions increases the risk of errors and inefficiencies. The costs of managing an increasingly complex system are also rising.

What role could AI play in the UK energy sector?

The UK sees forecasting and system optimisation as the main applications for AI. Algorithms can also support renewable energy integration and real-time grid management.

AI can also manage demand-side flexibility. One example is scheduling the charging of electric vehicle fleets at times that better suit the system. Such decisions affect system balancing, the local grid and wholesale prices simultaneously. However, the UK authorities stress that artificial intelligence is not simply intended to replace system operators. Over the coming years, its main role will be to support human decision-making.

The UK is also seeking views on cybersecurity and AI oversight

Automation also creates new risks. The UK government highlights cyberattacks, data manipulation and errors in AI models, among other concerns. Another potential challenge is explaining decisions made automatically.

The document therefore does not yet commit to introducing specific regulations. Rather than a finished strategy for deploying artificial intelligence, it is a call for evidence and views.

The consultation opened on 8 September and will run until 6 November 2026. The findings will help the government develop future policy and an AI strategy for clean energy. With more than 200,000 decisions required in a single month last year, digitalising system management is becoming an integral part of the energy transition.