Planning system ‘must keep pace’ with increased AI use in objections and representations

The Objector AI website offers members of the public an AI-driven automated way of creating planning objection process documentation, claiming to be able to instantly analyse an application and rank users’ objections to it based on impact.

If the AI system finds valid grounds for an objection, users can pay £45 to generate a ‘planning objection toolkit’, including letters, reports, video and speeches to use in the process. It also offers a bespoke service for larger developments.

The creators of Objector AI claim that using dual-AI systems simultaneously, with their results being cross-checked then combined, limits the risk of hallucinations or errors in the final outcome. Other, similar AI-based systems have already been launched into the market to generate objection letters and materials.

But critics have warned that without clear explanations that AI has been used in generating the objection material, it could lead to inaccurate or partial analysis which may then inadvertently be relied on by decision makers to rule on the application.

Iain Gilbey, a planning expert with Pinsent Masons, warned that without proper regulation of how AI is employed in the process, it could significantly harm the reputation of the planning system and lead to appeals and legal challenges.

“With this kind of technology only likely to proliferate, there needs to be a moderated and regulated approach as to how it can engage with the formal planning process,” he said.

“It would be a sensible safeguard if local authorities were to require that representations should only be accepted if they include a declaration as to whether AI has been used – in whole or in part – in compiling the objections. This would reflect wider guidance from the government on the use of AI in planning appeals.”

The move comes as the UK government looks to integrate AI further into the planning process, following the use of the technology in areas such as tidal predictions and peat mapping for developments, and the use of the Department of Science, Industry and Technology’s Consult tool to analyse casework for civil servants.

Sue Chadwick, a planning governance expert with Pinsent Masons, said introducing a dedicated AI segment to national planning practice guidelines and making amendments to The Town and Country Planning (Development Management Procedure) (England) Order 2015 and were needed to make sure the playing field remained level for both major and regular development applications.

“The planning system needs to embrace the role of AI in the process, but it also needs to regulate that involvement,” she warned.

“At the moment we risk the usage and pace of change of AI outstripping the government’s guidance and control at a policy and regulatory level.”

Robbie Owen, an infrastructure planning expert at Pinsent Masons, added that the same considerations applied to objections to, and other representations relating to, applications for consent for nationally significant infrastructure projects - but that a different response was required, given the separate legal regime governing infrastructure planning.

“There is a real concern that the concerted use of AI by those participating in the six-month DCO examination phase could exacerbate an already over-burdened process, by adding to the reams of submissions made by parties at each examination deadline,” he explained.

“This would work against government’s desire to improve the efficiency and effectiveness of examinations and consequently the overall Planning Act 2008 regime, including the speed of decision-making.”