AI Governance - a good practice checklist

A RICS AI Standard readiness checklist for surveying firms

This checklist maps to the RICS Professional Standard ‘Responsible use of artificial intelligence in surveying practice’ (the “RICS AI Standard”), effective 9 March 2026. It sets out what good looks like across every clause in the standard when your firm uses Artificial Intelligence (AI) in its work.

Read our companion Explainer for a summary of the Standard and its implications.

Then work down each item below - the empty boxes become your action plan.

Governance and strategy

In place?	What good looks like	Clause
☐	AI Strategy - where AI fits, what you will adopt, your boundaries	s.3.2
☐	Responsible AI Use Policy - scope, roles, liabilities, human oversight, training	s.3.2
☐	AI Risk Appetite Statement - how much AI risk you accept and where	s.3.2 / 3.3
☐	Roles and Accountabilities (RACI) - who owns each AI decision and sign-off	s.3.2 / 4.2
☐	Governance cadence - an at-least-quarterly review rhythm	s.3.3
☐	Appropriateness assessment (or a standing written policy)	s.3.2

Registers and records (the living evidence)

In place?	What good looks like	Clause
☐	AI Systems Register - every material AI system, with materiality reasoning	s.3.2 / 1.2
☐	AI Risk Register - the four categories, with RAG, reviewed quarterly	s.3.3
☐	Reliability and Accountability Log - a signed decision per material output	s.4.2
☐	Supplier due-diligence records - the seven areas, gaps logged as risks	s.4.1
☐	Client disclosure records - AI disclosed in writing, before use	s.4.3
☐	Data and consent records - what data may go into which tools	s.3.1

Processes and templates

In place?	What good looks like	Clause
☐	Engagement-terms clauses - the six disclosure points	s.4.3
☐	Explainability response template - the five-point written explanation	s.4.4
☐	Data-handling and approved-tool policy	s.3.1
☐	Output-override and incident process	s.4.2
☐	Legislation-conflict and RICS-reporting process	s.1.2

People and competency

In place?	What good looks like	Clause
☐	Baseline AI competency framework - what every AI user must understand	s.2
☐	Baseline and annual privacy training, with records	s.2 / 3.1
☐	Board or leadership AI briefing	Governance

Evidence and readiness

In place?	What good looks like	Clause
☐	Compliance evidence pack - producible on demand	s.4.4 (covering 3&4)
☐	Audit trail - a versioned history of decisions and reviews	Derived
☐	Periodic readiness review - before PI renewal or audit	Derived

If your firm builds its own AI, section 5 Standard adds development records on data, permissions, sustainability and testing.

Above all, you need one living record

Most of the boxes above are records, and the Standard expects them to be live and producible on demand. The simplest way to hold them is one register that contains, in one place:

• AI systems and materiality calls
• AI risks
• reliability decisions and named sign-offs
• supplier due-diligence
• training and competency
• client disclosures

Keep that current and you can evidence your position to an insurer, a client or RICS at any time.

WHERE WE CAN HELP - AI GOVERNANCE REGISTER