The Planning Inspectorate (PINS) has recently strengthened and clarified its guidance on the use of Artificial Intelligence (AI) in appeals and examinations. The message is not that AI should be avoided but rather that it should be used responsibly, transparently and with appropriate human oversight. The guidance, and takeaways, are also applicable to submission of planning applications with some Council’s now asking for completion of declarations with regard to use of AI from applicants and/or agents.
The key requirement is disclosure. Where AI has been deliberately used to draft or substantially rewrite text, produce a summary or analysis, generate or alter images or videos, or create content beyond straightforward formatting prepared by a human, this should be declared. The declaration can be brief and can be included in a covering email, letter, statement or proof of evidence.
The declaration should explain that AI was used, what it was used for and how, together with details of any particular sections or paragraphs affected (or if it affects the whole document , this should be confirmed). Most importantly, the person submitting the evidence should also confirm what checks have been undertaken and accept responsibility for its accuracy. This applies to both applicants and agents but also interested parties responding to applications/appeals as appropriate. This reflects the Inspectorate’s underlying principle that the responsibility for the material submitted remains with the party making the submission.
Not every use of AI needs to be reported. Routine functions such as spellchecking, grammar suggestions, formatting and accessibility tools are expressly excluded and do not generally alter the basis of the case of content of submissions compared to AI generated content/submissions. The focus is on deliberate and substantive use of AI to create or materially alter casework evidence.
There are particular implications for local planning authorities and professional advisers. If an authority uses AI to summarise representations, for example, the original representations should still be provided to the Inspectorate. A summary alone may omit material considerations and may not satisfy the relevant procedural requirements. Similarly, where several experts contribute to a case, each should make an appropriate declaration rather than relying on a generic ‘catch-all’ statement.
For appellants and interested parties, the practical lesson is straightforward: AI can assist with research, drafting, translation and analysis, but it should not replace professional or personal judgment. Any AI-generated material needs to be checked carefully for accuracy, completeness and context. The Inspectorate specifically highlights the risk of fabricated text, evidence and images, as well as exaggerated or misleading illustrations of development proposals. There have been many examples reported of AI fabricating or vastly misrepresenting other cases referred to or case law – parties are responsible for fact checking their submissions thoroughly.
The guidance also creates a potentially important procedural risk. Improper or excessive use of AI could contribute to a finding of unreasonable behaviour and, potentially, a costs award. Inspectors may also ask parties to clarify how AI was used where they have concerns about the provenance or accuracy of material.
It is therefore strongly advisable to keep a simple record of substantive AI use, identify affected material, verify the output against authoritative sources and include a concise declaration with the submission. Personal and sensitive information should also be treated carefully, with the Inspectorate reminding users to consider their data-protection obligations before entering such information into AI systems.
Ultimately, the guidance is less about restricting AI than establishing an audit trail around its use and as outlined, this is an approach that should be extended to all inputs into applications and appeals by any party. The Inspectorate recognises that AI can bring efficiency benefits but expects parties to remain accountable for what they put before an Inspector.
It is for this reason many of our clients may have noted that, within our terms of business, we have procedures in place for clients seeking to use and/or rely upon AI reviews of any work we produce as a company. In practical terms, AI may help prepare an appeal (or application), but it cannot take responsibility for the evidence: that remains with the appellant, authority, professional adviser or other party submitting it.
AI is a tool that will undoubtedly assist as it evolves but, for now, the message is clear disclosure is key and if utilised, it should be fact checked fully with relevant parties taking clear responsibility for the content of their submissions.