AI makes objecting cheaper, not planning committees slower

A public consultation closes on Friday. The planning officer expects twenty responses. She gets two hundred. Half are templated, most repeat the same four concerns, and three appear to come from addresses outside the catchment. This is not a new problem. AI-generated planning objections just make it cheaper to produce at scale.

The Guardian reported this week that AI could grind the UK planning system to a halt. The concern is that objectors can now draft submissions in seconds, flooding consultations and overwhelming planning officers. The framing assumes the system works smoothly now. It does not. The bottleneck is not the volume of objections, it is the number of officers employed to read them before a committee meeting that was booked six weeks ago.

Do AI-generated planning objections slow down the planning system?

They add volume, not complexity. A planning officer still reads for substance, not word count. Templated objections are easy to spot and easier to summarise. The system slows because there are fewer officers than there were five years ago, not because the objections are written by a machine.

The real shift is cost. Drafting a coherent objection used to take an hour. Now it takes three minutes and a prompt. That lowers the barrier to entry. More people object, not because they care more, but because it costs them less time. The planning committee still decides on the same grounds: material considerations, planning policy, and whether the concerns hold weight. AI does not change what counts as a valid objection.

What happens when a consultation gets flooded?

The officer writes a summary. If fifty objections raise the same concern, the report notes it once. If the concern is material, it goes to committee. If it is not, it does not. Volume might delay the officer’s report by a week, but the decision still rests on the merits. A planning committee does not count objections like votes. It weighs them.

The risk is not that AI-generated planning objections break the system. The risk is that planning departments are already stretched, and anything that adds to the reading list makes a slow process slower. That is a resourcing problem, not a technology problem.

Where does AI automation fit in planning work?

On the other side of the table. Developers and consultants already use automation to draft submissions, track consultation responses, and flag material objections worth responding to. A working system can pull objections from a public portal, tag them by theme, and draft a response for the planning officer to review. That does not replace judgement. It just means the judgement happens faster.

I have built systems that do this for clients with schemes in consultation. The automation reads, it does not decide. A planner still reviews the summary before it goes into the committee report. The system saves an afternoon, not a decision.

How planning officers handle templated planning objections

Officers are already trained to identify and group templated responses. The Royal Town Planning Institute guidance on handling public consultations emphasises material considerations over volume. When half the submissions use identical phrasing, the officer notes the concern once and attributes it to multiple respondents. The committee paper reflects the substance, not the count.

Does this mean planning gets easier?

No. It means objecting gets cheaper, and responding gets faster. The committee still meets on the same cycle. The decision still rests on policy and material considerations. AI does not make a weak objection stronger, and it does not make a strong application weaker. It just compresses the time between submission and decision, assuming the planning department has the tools to match the volume.

Most do not. That is the gap.

If you are running a scheme through consultation and the responses are stacking up, send a paragraph to jamie@goldspink.co. I will write back.