FCA authorised · FRN 724952 0117 325 0027Quote & buy →
Apex Insurance Brokers
Speak to a brokerGet a quote →

Aipedia: Surveyors and valuers

RICS AI standard: a compliance checklist for surveying firms

To comply with the RICS AI standard, a firm needs written records: material-impact decisions, an AI register, a quarterly risk register, supplier due diligence and signed reliability decisions. The standard has been mandatory since 9 March 2026, so missing paperwork is now a conduct issue as well as a claims weakness.

In short

The RICS standard on responsible use of AI in surveying practice is mandatory from 9 March 2026 for all members and regulated firms. Compliance is mostly about written records. Firms should record which AI outputs have a material impact and why (1.2), keep an AI register and policies (3.2), rate risks and review them at least quarterly (3.3), send written questions to suppliers and log gaps (4.1), document a named surveyor's reliability decision on outputs and dip sample high-volume work (4.2), tell clients in advance (4.3) and explain systems on request (4.4).

What the standard is and when it applies

RICS published the professional standard on responsible use of artificial intelligence in surveying practice on 10 September 2025. Paragraph 1.3 makes it mandatory from 9 March 2026, and RICS confirmed on 18 March 2026 that it is now in effect. It applies to all RICS members and RICS-regulated firms worldwide. Because it is a professional standard rather than guidance, departing from it is a regulatory matter and is likely to be relevant if a negligence claim is brought.

DocumentStatusDateMain ask
RICS AI standard (1st edition)Mandatory professional standardPublished 10 Sept 2025; in force 9 March 2026Records, governance, reliability decisions, client notice
RICS Rules of ConductMandatoryEffective 2 Feb 2022Integrity, competence, diligent service, accountability
Red Book (Valuation: Global Standards)Mandatory for Red Book valuationsEffective 31 Jan 2025Transparency and judgement in technology use

Step 1: decide and record material impact (1.2)

The standard bites on AI outputs that have a "material impact" on delivering the surveying service. List every AI tool in use, including features built into software you already own, and decide for each use whether the output could materially affect advice, a figure or a report. Record the determination in writing with reasons. A transcription tool that only drafts internal notes may be judged differently from one whose text goes straight into a client report. If the law conflicts with the standard, the law prevails, but you must record the conflict and report it to RICS.

Step 2: knowledge, training and data (sections 2 and 3.1)

Step 3: the AI register and policies (3.2)

Before using AI for material-impact work, write down whether AI is the most appropriate tool for the task. Then keep a written register showing, for each system, its purpose, the date it was first used and the next review date. Adopt a responsible-use policy and a procurement policy that cover human oversight and at least annual training. Our page on what to document about AI use has general record-keeping points that sit alongside this.

Step 4: the risk register, reviewed quarterly (3.3)

Keep a risk register covering, as a minimum, bias, erroneous outputs, information limitations and data retention. Rate each risk using RAG or a similar method. The standard requires review at least quarterly, so put four dates in the diary and minute each review, including any change of rating and the action taken.

Step 5: supplier due diligence (4.1)

Send written questions to each AI supplier and record the answers. The minimum topics are compliance with data law, permissions to use the data, dataset quality and bias, environmental impact and the provider's liability. Where a supplier will not or cannot answer, log the gap as a risk in your 3.3 register rather than leaving it unrecorded.

Step 6: reliability decisions and dip sampling (4.2)

This is the step most likely to matter in a claim. Apply professional judgement and scepticism to each output and document in writing the decision on whether it is reliable, prepared by or under a named, appropriately qualified surveyor. If an output cannot reasonably be used, tell the client. For automated or high-volume work, such as AVM-supported valuations, use randomised dip sampling and keep the results.

Step 7: client notice and explainability (4.3 and 4.4)

Tell clients in writing, in advance, when and why AI will be used. Terms must cover contesting AI use, redress, opting out where applicable and, if available, the extent of PI cover for AI use; see AI clauses in terms of engagement. On request, provide written information on the system type, how it works, its limits, your due diligence, your risk management and your reliability decisions. Prepare a short template now so requests can be answered quickly.

Firms that build their own AI tools have extra duties under section 5, including recording the application, risks, benefits and alternatives before deployment, a sustainability impact assessment, stakeholder involvement and data law compliance.

What this means for PI

SectionRecordFrequency
1.2Material-impact determinations with reasonsPer tool or use, on change
3.2AI register and policiesReview date per system
3.3Rated risk registerAt least quarterly
4.1Supplier questions and answersOn procurement and renewal
4.2Signed reliability decisions; dip samplesPer output or batch

CMS has noted that the standard is likely to inform how negligence claims against surveyors are judged, so these records may become central evidence in a defence. Insurers may also ask about them at renewal. For the insurance detail, read what the RICS AI standard means for PI and how the RICS minimum wording treats AI claims. Apex arranges PI for surveyors and valuers, and a broker can talk through how your compliance records support your renewal presentation. See the AI in surveying and valuation hub for the wider picture.

Sources

Frequently asked

Does the standard apply to every AI tool we use?

It applies to AI outputs with a material impact on delivering the surveying service. You must decide which uses meet that test and record the decision with reasons. Tools that only support internal admin may fall outside, but the decision itself still needs to be written down.

How often must the AI risk register be reviewed?

At least quarterly under section 3.3. The register should cover bias, erroneous outputs, information limitations and data retention, rated using RAG or a similar method. Keeping a short note of each review helps show compliance if RICS or a claimant asks.

Who signs off AI outputs?

Section 4.2 requires the written reliability decision to be prepared by, or under the supervision of, a named surveyor who is appropriately qualified for the work. It should record the judgement and scepticism applied, and whether the output can reasonably be used.

Is a sole practitioner exempt?

No. The standard applies to all RICS members and RICS-regulated firms. A sole practitioner can keep simpler records, but still needs the material-impact decisions, register, quarterly risk review, supplier checks, reliability decisions and client notices that the standard requires.

Check your PI covers how you use AI

Tell us about your firm and the work you do. A broker will review how your current PI treats AI-assisted work, or arrange new cover. No fee for a quote. Or call 0117 325 0027.

Start a PI quote Request a callback

Apex Insurance Brokers Limited is authorised and regulated by the Financial Conduct Authority. Registered in England and Wales, company number 07014570. This page is general information about how AI is used in the profession and how insurance may respond. It is not legal, regulatory or insurance advice on your circumstances, and it does not guarantee that cover will be available or on what terms.