From my perspective, the RICS Rules of Conduct are genuinely well-written for the Ai era, even though they don’t mention Ai explicitly. Rule 3 — the duty to undertake proper enquiries and provide a clear basis for professional opinions — maps directly onto how Ai should be used in surveying. Specially because the Ai isn’t replacing the surveyor’s professional judgement; it’s organising the evidence that supports it.
Rule 3 in the Ai context
Rule 3 requires members to act with diligence, make proper enquiries, and provide a clear basis for professional opinions. In practice this means every paragraph in a survey report should be traceable to a source: the field notes, the photos, the comparables, the planning data. Historically, the trace is in the surveyor’s head and loose notes. With Ai, it can be explicit — every paragraph carries a citation back to the specific source.
I believe this is a net gain for Rule 3 compliance, not a risk to it. Specially for PI insurance purposes, a source-cited report is a better defence than a professionally-drafted narrative without traceability.
How the Ai-assisted survey workflow works
Typical instruction. Surveyor voice-notes through the property, takes photos of every defect. Field data uploads to the instruction vault. Ai transcribes the voice notes, categorises by RICS section (movement, damp, roofing, services, external elements). Drafts each section with inline citations back to the specific photos and voice notes.
Surveyor reviews draft. Edits for professional opinion. Signs. Report goes to client with the source chain attached — every paragraph has a link to its supporting evidence. If a client or PI insurer ever queries a specific finding, the evidence is already attached.
Apart from this, the per-instruction vault means the survey for No. 12 Station Road cannot leak into the survey for No. 14. Each instruction is isolated at the retrieval layer — specially important in residential work where similar addresses might otherwise confuse.
Valuation work: where Ai helps, where it doesn’t
I believe Ai should not generate valuations. Valuation is a registered professional opinion, and the Red Book requires it to come from the RICS Registered Valuer. What Ai can do is organise the comparables, flag adjustments, draft the narrative that explains the valuation — with the surveyor’s number and reasoning.
Specially since the 2022 Red Book changes, comparable evidence needs to be clearly documented. A well-structured Ai summary of 12 comparables with adjustments for condition, floor area, and location gives the surveyor a stronger starting point than reading 12 pages of Land Registry data from scratch. The number remains the surveyor’s; the organisation is Ai’s.
PI insurance and the audit chain
Apart from this, PI insurers are starting to ask questions about Ai use in surveying. The firms I’ve seen pass this conversation cleanly have three things. A written Ai policy (we’ve published a free template for UK SMEs that adapts for surveying). A tool that produces source-cited outputs. A tamper-evident audit chain that the insurer can examine.
Specially for professional firms, the audit chain actually strengthens PI posture — if a claim ever arose, the chain lets you reconstruct exactly what evidence supported which paragraph of which report, and who signed it off.
Other Me for property and surveyors
Other Me is built for RICS-regulated practices, with per-instruction vaults and source-chain citations on every Ai-drafted output. Full workflow on the Property and Surveyors solution page. You can start a free 7-day trial, no credit card, run one real instruction through it, and see whether the audit chain satisfies your PI insurer’s standard.