Does it apply to you?

Who it covers

The standard applies to RICS members and RICS‑regulated firms that use AI in delivering surveying services (section 1). Size doesn’t change that. A sole practitioner who asks an AI assistant to summarise a document, then relies on that summary in a report, owes the same duties as a large firm. What changes with size is how much work it takes: a one-person risk register is still a risk register.

What counts as AI

More than most firms expect. If an output can influence the service you deliver, it counts, whatever the product is called. That includes:

• general-purpose assistants used for drafting, summarising or research
• transcription and meeting-notes tools
• AI features built into software you already pay for, such as your word processor, email or valuation platform, where the output feeds your work
• systems bought as part of a wider package, where the AI is built in rather than sold as AI

The material impact decision

Clause 1.2 asks whether your AI use has a material impact on the surveying services you deliver. The test is whether an output is capable of influencing the service and, if so, how. It’s judged on the facts of each use, so the same tool can be material in one job and not in another.

If you decide your use is material, record that decision and your reasoning (1.2). Several of the heavier duties then apply: a risk register (3.3), supplier checks before you buy (4.1), written decisions on outputs (4.2) and telling clients in your terms of engagement (4.3).

The standard doesn’t require a record when you decide your use isn’t material, but it’s worth writing one. A firm that asked the question and noted a reasoned answer is in a far stronger position than one that never asked. Some duties are worded without reference to material impact, so treat them as applying whenever you use AI: baseline knowledge (section 2), protecting confidential data (3.1), your AI policy and training (3.2), and explaining your AI use on request (4.4).

Worked examples

These are starting points. Your own facts decide the answer.

• An automated valuation model producing figures that go to clients. Material. The AI is doing core professional work.
• An assistant summarising a lease that you then rely on when writing your report. Material. This is one of the standard’s own examples.
• A tool suggesting which part of a building to investigate for a fault. Material. Another of the standard’s own examples.
• AI meeting notes that feed into drafting a report. Likely material. The AI shapes content that reaches the client.
• Tidying the wording of internal emails. Probably not material. Nothing reaches a client deliverable.
• Drafting a post for your firm’s website. Probably not material. It doesn’t touch the surveying service.

Your firm makes the call

The decision is your firm’s, and so is the record. We can’t make it for you, and neither can any software. In guardbook you write down your reasoning, sign it and keep it with the rest of your records. Revisit it whenever your use of AI changes, and at least each time you review your risk register.