Use of artificial intelligence in casework evidence
- TP Editorial Team
- 1 day ago
- 2 min read
New guidance was announced by PINS on Friday 20th February 2026
What the guidance says
You can use AI to help prepare appeal/application/examination material, as long as it’s used responsibly and transparently.
You must tell PINS if AI was used to draft or substantially rewrite text, create summaries/analysis, generate or alter images/video, or do anything beyond straightforward formatting.
You do not need to declare routine tools like spellcheck, grammar suggestions, formatting tools, or accessibility features.
What your declaration needs to include
You should briefly state:
AI was used
Which tool (e.g., Copilot / ChatGPT / Midjourney)
What it was used for (which parts, and whether images/video were altered)
What checks you made, and that you take responsibility for the accuracy.
You can put this in the cover email/letter, or inside the statement / proof of evidence.
Extra expectations (professionals vs interested parties)
Professional parties: expected to take responsibility for accuracy and lawfulness; also use AI consistently with relevant professional codes; PINS points to the general obligations around expert evidence in the Procedural Guide.
Interested parties: can use AI to help draft/translate, but should still add a short disclosure statement and take responsibility.
Why PINS is doing this
It helps Inspectors understand source/handling of what they’re assessing, supporting fair and transparent decisions.
PINS flags risks: AI can be wrong, and can also generate fake text/images/evidence. Improper use may be treated as unreasonable behaviour, potentially exposing parties to costs.
Data protection point
Do not put personal or sensitive data into public AI tools (PINS signposts ICO guidance).
Practical significance for you (planning work)
If you’ve used AI to help draft a representation, SoCG text, PoE, or any planning statement, this is now a simple “AI used” audit trail you should build into your templates.
It’s also a gentle warning that AI-generated “evidence” without verification (especially anything that looks like an expert opinion, technical conclusion, or image-based analysis) could backfire at inquiry/appeal and become a costs risk.
Comments