Resources

The RICS standard

What the RICS standard on artificial intelligence asks of your firm, in plain English.

What it is

Responsible use of artificial intelligence in surveying practice is a professional standard published by RICS, the Royal Institution of Chartered Surveyors. It was published in September 2025 and became mandatory on 9 March 2026 for RICS members and RICS‑regulated firms.

It’s a conduct standard. It sets out how a surveyor should behave when AI is involved in the work, and it applies to firms of every size, sole practitioners included. Most of its heavier duties depend on one judgement: whether your use of AI has a material impact on the service you deliver (1.2).

What it asks of your firm

The standard has five sections. We find it easiest to read through the five GUARD™ pillars, which sort its requirements by the job they do. The clause numbers are the standard’s own.

Governance

Decide whether your AI use has a material impact and, if it does, record that decision and your reasoning (1.2). Adopt an AI policy, informed by your risk register, that sets out who is responsible for what and how human judgement works alongside AI (3.2). Keep private and confidential data out of AI systems unless those affected have given express written consent in advance and you’re satisfied the risk is acceptable (3.1).

Use

Know which AI systems your firm uses, including AI built into everyday software, and keep a written register of them (3.2). For each one, assess whether it’s appropriate for what you actually use it for (3.2).

Accountability

Where an output has a material impact, a suitably qualified, named surveyor decides in writing whether it can reasonably be relied on. If it can’t, the client is told in writing, with the reasoning. For automated or high-volume tools, check random samples at regular intervals instead (4.2). On request, explain to a client in writing how your AI systems work and how you manage them (4.4).

Risk

Keep a risk register covering bias, erroneous outputs, limited information about each system and what happens to the data you put in, and review it at least quarterly (3.3). Before buying an AI system that will have a material impact, put set questions to the supplier in writing and record what they tell you, and what they don’t (4.1).

Documentation

Everyone using AI needs a baseline understanding of how it works and where it goes wrong (section 2), with training at least once a year (3.1, 3.2). Where AI use is material, tell clients in writing and in advance when and why it will be used on their work (4.3).

What it doesn’t do

  • It doesn’t approve or ban particular tools, or tell you how to use AI for a particular kind of survey.
  • It doesn’t require AI cover on your professional indemnity insurance. Where AI use is material, it asks you to tell clients what cover you have, if any (4.3).
  • It doesn’t ask you to record a decision that your AI use isn’t material, though writing one down is still sensible (1.2).
  • It doesn’t override the law. Where the two conflict, legislation takes precedence, and the firm records the conflict and reports it to RICS (section 1).

This page is a summary. The standard itself is the authority, so read it for the full wording.

View the standard on the RICS website (opens in a new tab)

We’re independent: guardbook isn’t part of RICS and isn’t endorsed by it.

Start building the record the standard asks for, a few minutes at a time.

Start free trial