AI in quantity surveying: where it's already used, and what the rules now say

Where AI is actually used in quantity surveying today — measurement, cost planning, drafting, transcription — and what the RICS AI standard requires firms to do about each use.

Ask a QS practice whether it uses AI and the honest answer is usually “more than we realised”. AI in quantity surveying rarely looks like a robot doing take-offs — it looks like features quietly switched on inside tools the firm already ran. Since 9 March 2026, the RICS AI standard attaches specific duties to that use wherever it materially affects a service.

Where AI actually shows up in QS work

• Measurement and take-off — OCR and vision models extracting quantities and elements from drawings and point clouds. The clearest material-impact candidate: output feeds the cost plan directly.
• Cost planning and estimating — assemblies suggested from historic data, benchmark rates surfaced automatically, anomaly flags on pricing.
• Document work — drafting assistants in Word and Outlook tightening reports, fee proposals and contract correspondence.
• Meetings — transcription and summary of site and client meetings.
• Programme and risk — schedule-risk flags and comparison of contract documents for deviations.
• Embedded features — AI inside estimating, BIM and document-management platforms, often arriving via a routine update nobody assessed.

The pattern in the rules

The standard doesn't rank tools — it ranks impact. For each use above, the same chain applies: decide and record whether the use is material; if it is, the tool belongs on the systems register, the supplier gets written due diligence, clients get advance written notice, outputs get named reliability reviews, and the risks live on a quarterly-reviewed risk register.

So a drafting assistant whose output is fully re-written by the surveyor may sit outside the heavy duties, while the measurement tool that feeds a cost plan sits squarely inside all of them. Same firm, two tools, two very different compliance postures — decided use by use, not brand by brand.

The opportunity being missed

The competitive story cuts both ways. Firms that govern AI openly get to use it aggressively — faster take-offs, quicker drafts, better benchmarks — because every use is registered, reviewed and disclosable. Firms that ban it or ignore it don't avoid AI; they avoid visibility of the AI their staff use anyway. In a fee-pressured market, the practices pairing modern tooling with a defensible record will win work from the ones pairing quiet AI use with crossed fingers. Start with the 12-point checklist.

Will AI replace quantity surveyors?

The standard's answer is structural: every material output requires professional judgement and a named, appropriately qualified surveyor accepting responsibility. AI changes what the surveyor spends time on — less extraction, more verification and advice — rather than removing the professional from the chain.

What should a QS firm do first?

Inventory the tools (including embedded features), make the materiality call per use, and stand up the systems register — an hour with a template. Then policy, risk register and client documents follow in order. The checklist article sequences all twelve steps.

This article is general information, not legal or professional advice. ComplyQS is not affiliated with or endorsed by RICS.