Skip to content
Neural Daily

BusinessNews

PlanAssist: DeepMind AI cuts UK planning assessment time

🇭🇺 Olvasd el magyarul →

Google DeepMind's Gemini-based PlanAssist cut the average time English planning officers spend assessing householder applications by 58% in pilots. It is now rolling out to 12 more councils.

By: Mikola Máté Patrik · · 2 min read

Abstract 3D illustration of miniature houses on layered, wavy terrain with architectural plans in the foreground.
Image: Google DeepMind · Press kit · Official hero image from the Google DeepMind blog post · Source

What's new

Google DeepMind announced on October 31 (Google DeepMind) that its PlanAssist tool will be made available to 12 more local planning authorities in England. Built with Gemini, it handles routine householder applications such as loft conversions, conservatories and extensions. In pilots across Barnet, Camden and Dorset, the average time a planning officer spent on householder assessment tasks dropped by 58%.

How it works

Google DeepMind developed the tool with the UK government, Google Cloud and Faculty. It first validates application documents and architectural drawings to check that the required information has been submitted. It then checks the application against heritage constraints such as tree preservation orders, as well as local and national planning policies. It also looks at whether objections were raised before, for example by local residents. Finally, it drafts fully cited reports that officers can edit before making a decision.

According to DeepMind, householder applications make up as much as 70% of planning officers' workloads. In the pilot, officers approved more than 70% of the generated reports with no edits.

What's next

PlanAssist now enters a 12-month beta on live applications. The new councils join in three cohorts, bringing the total to 15 across metropolitan and rural areas. Subject to successful testing, the government plans to make the tool available to every council in England in 2027. Every assessment will still be reviewed and approved by a qualified planning officer before any decision is made.

Why it matters

This is one of the more concrete examples of a frontier model moving into day-to-day public administration with a measurable time saving. The UK government wants 1.5 million new homes by 2029, and slow planning approvals are a real bottleneck. The model does not decide anything here; it takes over the paperwork, which is also the safer route from a regulatory point of view. The beta will show whether the pilot numbers hold up on live cases across very different councils.

    Produced with the help of AI, reviewed by an editor. How we make our articles

    Reviewed and edited by the Neural Daily desk. About

    Related articles