Last updated on 23 June 2025

Government announces AI Initiative to speed up planning process and help build 1.5 million homes
Prime Minister Keir Starmer announced a significant new initiative to modernise the planning system using AI (artificial intelligence) on 9th June 2025. The aim is to help deliver 1.5 million new homes across the country by streamlining administrative processes, thus speeding up the planning process.
At the heart of this initiative is a new tool called Extract, developed in collaboration with Google DeepMind and powered by its advanced Gemini AI model. This technology is designed to rapidly digitise and interpret decades of complex planning policies and documentation – tasks that previously required many hours of manual effort by planning officers.
Initial trials with local authorities in Hillingdon, Nuneaton and Bedworth, and Exeter have shown encouraging results. Extract has demonstrated the potential to accelerate decision-making and improve access to information for planners, developers, and local communities. While full implementation details are still being developed, the initiative reflects a growing recognition that, when applied thoughtfully, technology can help address long-standing inefficiencies in the planning process. That said, it remains essential to ensure that increased speed does not come at the expense of quality or community engagement.
In this month’s newsletter from the Kingston Society, the Society says:
“While the introduction of AI into the planning system is broadly welcomed as a step toward greater efficiency, we have measured concerns about its limitations. We note that streamlining document processing, though helpful, does not address the deeper structural issues that cause planning delays, such as the complexity of legal frameworks, protracted appeals processes, and a lack of time-bound decision-making. AI should support, not supplant, the professional judgement of local planners, whose knowledge of place-based priorities remains essential. We have some scepticism around the framing of the tool as a breakthrough: much of its function, e.g. digitising old document, is a pragmatic application of existing technology rather than a leap into the AI future.”