An AI planning assistant is expanding to 15 English councils

For a family waiting to add a bedroom or convert a loft, a planning application is not an abstract measure of government efficiency. It is the next step toward a home that better fits their life.

An AI tool designed to help council planning officers with that work is now moving into a larger test. PlanAssist, developed with Google DeepMind, Google Cloud, Faculty and the UK government, is expanding from three pilot councils to 15 authorities in England. The government describes a 12-month trial and intends wider availability by the end of 2027. Government announcement, October 30

Google’s October 31 account describes the expansion as a beta using live applications, with new councils joining in stages. It says broader rollout depends on successful testing. This is an expansion of a pilot, not a finished service already available to every homeowner. Google DeepMind’s update

Helping the officer, not replacing the decision

Google says PlanAssist checks application documents and drawings, brings together relevant policies and constraints, and drafts cited reports that officers can edit. Qualified planning officers retain responsibility for the final assessment and decision. The pilot began in Barnet, Camden and Dorset. How the tool works

That design puts AI into the supporting work around a consequential judgment. A tool can help an officer find material and prepare an assessment without becoming the authority that decides what someone may build.

For residents, the important outcome would not simply be a report produced quickly. It would be a timely decision that has properly considered the application, relevant rules and the interests of neighbors. Speed has value when it preserves the quality of that process.

What the early numbers actually measure

Google reports a 58% reduction in average officer time spent on householder assessment tasks in early testing. It also says more than 70% of generated reports were accepted without edits. These are findings reported by the developer, not an independently reproduced evaluation. Neither figure establishes that the full wait from submission to a final decision fell by 58%. Pilot results

The government separately presents faster processing alongside changes to decision-making rules and additional planning staff. Its announcement says officers will test PlanAssist alongside their normal process. Those overlapping changes mean that improvements in the overall service should not automatically be credited to AI alone. Government explanation

Nor does accepting a report without edits, by itself, establish that every conclusion in it was correct. The larger trial is an opportunity to examine accuracy, exceptions and how the tool performs outside the original councils.

Why the expansion is worth watching

The additional authorities include urban and rural areas. The government's list includes Bristol, Newcastle, Nottingham, Ashford, Cumberland and others, and it says the software will not be mandatory. Participating councils and conditions

Bright’s reading is that the next stage should be judged at two levels. First, does it make the officer's work easier without creating new checking burdens or missing important details? Second, does that improvement reach the applicant as a clearer, more reliable service?

Those are related outcomes, not interchangeable ones. Saving time on one part of an application could help reduce a backlog. It could also be absorbed by other parts of the process. Measuring the applicant’s experience alongside staff time would show where the benefit actually lands.

The human opportunity is practical: less effort assembling the same information, more attention available for the decisions that require expertise, and potentially less uncertainty for people trying to improve their homes. The expanded trial is a meaningful next step toward testing that promise, with people still accountable for the answer.

AI-assisted, source-reviewed Bright reporting. Cover: Bright editorial illustration, conceptual artwork rather than documentary evidence or measured outcomes. Bright has not firsthand-tested the systems described; interpretations are identified in the article.