Energy minister Martin McCluskey has set out the government’s early thinking on AI in the energy system. An independent review recommends moving towards a probabilistic grid, ahead of a full strategy next year.
The UK government wants AI to squeeze more out of a grid that grows more complex by the month. Martin McCluskey MP, Minister for Local Energy and Jobs, spoke at Google’s London offices on 8 September. There he launched an AI for clean energy vision. He also published an independent review of AI in electricity networks. Both feed into an AI in Energy Strategy due next year.
The minister framed the problem in numbers. The UK passed 2 million solar installations in March 2026. In 2025 it had nearly 3,000 offshore wind turbines across 46 wind farms. Together they give up to 16.5 GW of grid-connected capacity. A system once run from a handful of coal plants now balances millions of small assets, second by second. McCluskey argued that only AI can manage that.
Three announcements
The vision document is a consultation. It sets out how AI could cut waste, cut costs and make energy supply more secure. It then asks the public and industry three things: what role government should play, how to manage the risks, and what is blocking adoption. The answers will shape the strategy the department intends to publish next year.
The independent review recommends moving towards a probabilistic grid. Lucy Yu, the government’s AI Energy Champion, led it. It also calls for clear governance and faster roll-out of AI tools that give consumers measurable gains. McCluskey described the work as pushing at the frontier of global thinking on AI in energy.
Third, the Interim AI Adoption Plan published in June identified four barriers. They are better data, clearer routes from pilot to roll-out, trusted governance, and the skills to scale solutions across the system. For start-ups, the minister said, each of those barriers is also a market opportunity.
Security and public consent
McCluskey set two conditions. The first is security. Energy networks are a known target for hostile states across Europe, he said. The UK therefore can’t add weak points through technology it hasn’t fully explored. The second is public consent. Every industrial revolution has raised questions about how people live. So governments have a duty to take communities with them.
He also warned that voters won’t forgive governments that fail to prepare for the future. AI, he argued, has a part to play in updating the energy system, in keeping British industry competitive and in protecting national security.
What it means for a constrained grid
The UK’s transition problem is increasingly one of wires. Connection queues are long and new transmission lines take years to build. Better forecasting and a grid run on probabilities can squeeze more capacity from the wires already in place. That is therefore one of the few levers that works on a short timescale.
Whether the strategy delivers will depend on the plain work the review flags: data, governance and skills. The vision is, by the minister’s own admission, a starting point. Still, it tells the sector that Whitehall now treats AI as part of the grid. The full speech is on GOV.UK.
