PlanAssist, an AI tool for council planning officers, is expanding from three pilot authorities to a 12-month trial involving 15 councils in England. The government announced the expansion on 30 October 2026. For homeowners seeking permission for an extension or loft conversion, it could change how their application is assessed.
PlanAssist supports the assessment; a planning officer remains responsible for the decision. It is not an automatic approval service. For applicants, the important distinction is between faster work on the paperwork and a shorter wait for permission to build.
What the tool does—and who decides
In its 31 October account, Google DeepMind says PlanAssist checks documents and drawings, planning policies, site constraints and earlier objections. It then drafts assessments with supporting references, which officers can edit.
The ministry’s programme explanation leaves the weighing of local context and competing considerations with the officer. That matters for applicants and neighbours: gathering evidence is a different job from deciding what weight to give it.
For a broader discussion of that distinction, see RohitAI’s explainer on human review of AI-assisted work. It is not an evaluation of PlanAssist.
Why less assessment time does not mean the same cut in waiting
Google reports a 58% average reduction in officer time on householder assessment tasks during the pilot. That measures staff effort, not the time from submission to decision.
The government’s October release also links the saving to cutting routine processing from eight weeks to four. Its June programme explanation described that reduction as an ambition. Neither announcement provides a case-level timing dataset establishing what wait an individual applicant should now expect.
Current official guidance still gives eight weeks for most non-major applications once validated, unless a longer period is agreed in writing; some categories have different limits. That is the decision framework, not an observed average or a guarantee that every case finishes on time.
There is also consultation with neighbours and other interested parties, normally lasting 21 days. Amendments can require further consultation. These stages help explain why quicker assessment does not translate directly into the same percentage reduction in calendar time.
A reasonable inference is that less paperwork could free staff capacity. But estimating the effect on a resident’s wait requires evidence about the whole process, including queues and requests for changes—not just the assessment task.
Google also reports that more than 70% of generated reports needed no edits. That is an editing measure, not an application-approval or accuracy rate. Its post gives neither sample sizes nor a detailed timing method.
Which councils are involved
The announced group of 15 councils consists of:
Original pilots: Barnet, Camden and Dorset.
Additional London councils: Lambeth, Redbridge and Westminster.
Other additional councils: Ashford, Bristol, Cumberland, Newcastle, North Lincolnshire, Nottingham, Rochdale, South Gloucestershire and Stoke-on-Trent.
Google describes a staged rollout using live applications. Inclusion in the council list therefore does not establish whether a particular case is using PlanAssist.
Wider availability across England is planned for the end of 2027, subject to further testing. Adoption will not be compulsory. The announcements do not set out exact council start dates or case-selection rules.
What to do before submitting an application
Check whether permission is needed. Some projects fall under permitted development rights. GOV.UK advises checking with your local planning authority if unsure.
Use your council’s usual application route. GOV.UK directs applicants to their local planning authority; the trial does not announce a separate public PlanAssist submission service.
Ask about your case, not just council participation. Ask your case officer whether PlanAssist is involved and what decision timetable to expect. Do not schedule work on an assumed four-week approval.
For councils, the Local Government Association’s test for rollout is broader than speed: accuracy, usability, value for money and human accountability. For households, the useful evidence will be how long comparable applications actually take to reach a decision—not just how quickly a draft assessment appears.