What if planning officers could spend less time searching through documents and more time helping people get a clear, well-informed decision?
Early trials of a government-backed AI planning assistant suggest that some of that administrative work could become considerably quicker.
In an announcement published yesterday, 30 October, the UK government reported a 58% reduction in time spent on particular planning tasks during testing of PlanAssist in Camden, Barnet and Dorset councils.
That is an encouraging result. But it is a task-level time saving, not evidence that complete planning applications are now being processed 58% faster.
What Does PlanAssist Actually Do?
Planning officers need to bring together information from an application, relevant policies and the constraints affecting a site. Finding and cross-checking that material can involve a substantial amount of routine work.
PlanAssist is designed to gather those details in one place and produce an evidence-backed preliminary assessment for a trained officer to review.
The project involves the government’s Digital Planning Programme and Incubator for AI, working with Google DeepMind, Google Cloud, Faculty and local authorities.
The important point is who remains responsible. The AI supports the officer’s work; it does not grant planning permission. Officers still have to check the information, weigh the relevant considerations and exercise their professional judgement.
The department’s June explanation makes this explicit: each application must be considered on its own merits, with a human planning officer involved in the decision-making process.
What the 58% Figure Does and Does Not Mean
The government reports that early testing reduced the time spent on key planning tasks by 58%. That describes particular activities within the process, not the total time from submitting an application to receiving a decision.
A planning application can involve other stages and dependencies. Saving time on gathering policy information does not automatically remove every delay elsewhere.
The government’s ambition is to reduce routine householder application processing from around eight weeks to four. That remains an ambition, not a result established by the task-level trial figure.
The distinction matters because a faster administrative step is useful, but residents ultimately experience the whole process.
Fifteen Councils Will Test It for 12 Months
The next step is a wider, 12-month testing phase involving 15 councils in total. That includes the original three authorities and 12 additional councils, rather than 15 new councils on top of the initial group.
One of those joining is Ashford Borough Council in Kent, giving the development a particularly local connection for readers here.
Officers will use PlanAssist alongside their normal processes to test its performance at a larger scale. This is expanded testing, not a completed national rollout or proof that every participating council will achieve the same saving.
The June blog describes the earlier prototype and an earlier expansion proposal. The 30 October announcement supplies the updated figure of 15 councils.
Today’s Planning Rule Change Is Separate
There is another development in the same announcement that should not be confused with the AI trial.
From today, Saturday 31 October, new rules under England’s National Scheme of Delegation allow professional planning officers to decide more straightforward applications without waiting for a planning committee.
That is a change to how decisions are delegated, not something PlanAssist has achieved. Any reduction in committee-related waiting time cannot simply be credited to AI.
Both changes aim to improve the planning service, but they work in different ways and need to be assessed separately.
Why This Could Matter for People and Businesses
For householders considering an extension or loft conversion, a more efficient service could eventually mean less uncertainty and more useful contact with the officer handling their application.
For small businesses improving their premises, clearer information and fewer avoidable administrative delays could make it easier to plan work and coordinate suppliers.
For council staff, reducing repetitive document searches could create more time for site visits, discussions with residents and complex cases that need professional attention.
Those are possible future benefits, not outcomes demonstrated for every resident, business or planning team by this announcement.
A Practical Idea That Needs Wider Evidence
The broader trial needs to establish more than speed. Does the assistant identify the right policies and constraints? How much checking does it require? Does it support consistent, well-reasoned assessments across different councils?
Independent evaluation would help establish whether the early finding translates into reliable improvements in everyday use. Faster work is valuable only if accuracy, transparency and accountability are maintained.
There is a genuinely positive direction here: AI is being tested against a recognisable administrative problem, with skilled people retaining responsibility for decisions.
For now, the good news is a promising government-reported saving on selected tasks and a wider opportunity to test it properly. The next milestone is evidence that the complete service improves for the people who use it.
Latest announcement: UK government: Faster planning decisions as new rules and AI tackle delays ↗, published 30 October 2026. Source for the reported 58% task-level saving, the 15-council trial, Ashford’s participation and the separate delegation changes taking effect on 31 October.
Prototype background: MHCLG Digital: Using AI to support planning decisions, what it means for planners and residents ↗, published 19 June 2026. Explains the prototype, human oversight and the ambition to move householder processing from eight weeks to four; it is not an independent evaluation of the October results.
Less hype. More real-world AI progress.