AI only becomes profitable when it saves 30–40% or more in time — and for most surveyors, we’re simply not there yet.
The Reality of Surveying Workflows
Even with AI in the mix, a surveyor still needs to:
- Physically visit the property
- Photograph the defect
- Log notes using a voice recorder or digital tablet
So what’s the real gain from asking AI to turn those inputs into a report clause?
Standardised Phrases Already Work
Most modern surveyors already rely on standardised phrases within reporting software. These are:
- Fast to apply
- Easy to review
- Familiar to the surveyor and client alike
And most defects are repetitive:
- Slipped slates → Nail sickness
- Cracked tiles → Impact damage
- Leaking gutter → Failed joint
In fact, back in the early ’90s — long before AI was a buzzword — I introduced a system where we stored pre-written report paragraphs on a floppy disk. Surveyors would cut and paste the relevant ones, tweak them based on their notes, and produce the final report within hours of inspection. That was not AI, it was standardisation with a hint of automation.
The Limitations of AI and ChatGPT
Asking ChatGPT to analyse a photo and suggest a diagnosis might feel innovative — but it’s not fit for professional use. Why?
- You have no control over how it was trained
- You don’t know where it got its data
- There’s no guarantee of accuracy, consistency, or liability coverage
In regulated work like surveying, confidence and traceability matter more than novelty.
The Real Disruption May Come from Clients
The real shift might not be surveyors using AI — but buyers using AI tools themselves to take photos of defects, identify potential issues, and only bring in a surveyor if something concerning is flagged.
That’s where AI could truly change the game — and it’s a shift we should be watching closely.