UK government opens three-month consultation on AI for clean energy, backed by 34-recommendation network review

The UK Department for Energy Security and Net Zero published its Vision for an AI-enabled clean energy system on 8 September 2026, opening a call for evidence that runs until 6 November^[1]. The document is the starting point for what the government says will become the UK's first AI for Clean Energy Strategy, due next year^[1].

What the vision covers

The document maps out where AI might improve forecasting and asset management, while flagging technical and security risks regulators and operators need to manage. The vision sets out early thinking on how AI could improve efficiency, cut costs and strengthen energy security.

Minister for Local Energy and Jobs Martin McCluskey delivered the announcement at a Google for Startups event in London on 8 September. He argued that the growing complexity of the electricity grid means AI will be needed to manage generation and demand in real time. The grid context he cited is concrete: UK solar capacity surpassed 2 million installations for the first time in March 2026, and the country had nearly 3,000 fully commissioned offshore wind turbines across 46 wind farms producing up to 16.5 GW of grid-connected capacity in 2025^[1].

"The grid today is managing millions of decentralised energy generating assets from small- to large-scale, and the only way we can handle that generation and demand in real-time is through using AI," McCluskey said.

The Yu review

Alongside the vision, the government published an independent review of AI deployment in electricity networks led by Lucy Yu, the government's AI Champion for Clean Energy. The review makes recommendations in four broad areas:

• transitioning the electricity system to AI-enabled, risk-based operation and planning
• ensuring AI-driven flexibility delivers tangible benefits for consumers
• establishing clear governance for increasing autonomy
• accelerating the deployment of proven AI solutions

The review contains 34 recommendations in total. It recommends moving towards AI-enabled, risk-based network planning and operation, improving flexibility, establishing clearer governance as systems become more autonomous, and accelerating deployment of already-proven AI applications. Its scope covers grid connections, network planning, transmission and distribution, balancing, flexibility and optimisation.

One finding that has drawn less attention: recommendation 17 gives Ofgem and the Competition and Markets Authority 12 to 18 months to build capability against algorithmic collusion - not cartels run by people using AI, but independent learning systems arriving at anti-competitive behaviour with no communication and no intent.

Barriers the government has already identified

An Interim AI Adoption Plan published in June 2026 highlighted four key barriers to adoption: better data, clearer routes from innovation to deployment, trusted governance frameworks, and the skills and capabilities needed to scale solutions across the system. DESNZ said it will focus first on measures that could unblock easily achieved benefits and where the UK may gain a competitive advantage.

What to watch

The call for evidence closes at 11:45 pm on 6 November 2026. It asks stakeholders for evidence on opportunities, risks, barriers to adoption, and longer-term system impacts, to inform policy development and the UK AI for Clean Energy Strategy. The strategy itself has no published date beyond "next year." Whether the consultation produces binding regulatory changes - or remains a framing exercise - will depend on how Ofgem and the National Energy System Operator translate the Yu review's 34 recommendations into licence conditions and procurement rules.