> ## Documentation Index
> Fetch the complete documentation index at: https://help.guardbook.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# guardbook: RICS AI compliance for surveying firms

> What guardbook does, which RICS obligations it carries, and what your firm still has to do outside the product.

guardbook is a record-keeping system for the RICS standard *Responsible use of artificial intelligence in surveying practice*, which takes effect on **09/03/2026**.

It holds the records that standard requires, generates the documents it expects, and logs what your firm did and when.

## What it covers

| Area | What guardbook holds | Clause |
| - | - | - |
| [**Material Impact**](https://help.guardbook.ai/materiality/overview.md) | Your firm's determination that AI use is material, and the reasoning | 1.2 |
| [**Tool Register**](https://help.guardbook.ai/tools-register/overview.md) | Every AI tool, with an appropriateness assessment and supplier due diligence each | 3.2, 4.1 |
| [**Risk Register**](https://help.guardbook.ai/risk-register/overview.md) | Scored risks with controls, appetite and RAG banding | 3.3 |
| [**AI Policy**](https://help.guardbook.ai/ai-policy/overview.md) | A generated, versioned, signed policy | 3.2 |
| [**Oversight Log**](https://help.guardbook.ai/oversight-log/overview.md) | Written reliability decisions on AI outputs | 4.2 |
| [**Client Disclosures**](https://help.guardbook.ai/client-transparency/overview.md) | A versioned client-facing AI statement | 4.3 |
| [**Evidence pack**](https://help.guardbook.ai/evidence-audit/evidence-pack.md) | All of it, as one PDF | 4.4 |

Everything is written to an [audit trail](https://help.guardbook.ai/evidence-audit/audit-trail.md) that cannot be edited or deleted through the product.

## What guardbook does not do

Worth knowing before you rely on it.

> [!WARNING]
>
> * It does not **train your staff**. Three separate training duties sit in the standard and none of them are satisfied by using software.
> * It does not **record your quarterly risk review**. Clause 3.3 requires one and guardbook has no feature for it yet.
> * It does not **write to your clients**. Neither the clause 4.3 disclosures nor the clause 4.2 letter about an unusable output are sent for you.
> * It does not **schedule dip samples** for automated or high-volume AI use.
> * It does not **decide anything**. Every judgement in it is yours, and it will happily record a poor one.

## Where to start

**1. Create your account**

See [create an account](https://help.guardbook.ai/getting-started/create-account.md).

**2. Work through the questionnaire**

Eighteen questions across four sections. It seeds your Tool Register and everything guardbook later generates. See [the questionnaire](https://help.guardbook.ai/getting-started/onboarding-questionnaire.md).

**3. Complete your registers**

Tools first, then risks. The policy is generated from them, so the order matters.

**4. Generate and adopt your policy**

See [AI Policy](https://help.guardbook.ai/ai-policy/overview.md).

**5. Start logging decisions**

The [Oversight Log](https://help.guardbook.ai/oversight-log/overview.md) is the record that has to be kept as you work, rather than assembled later.

New to the standard itself? [The RICS AI standard](https://help.guardbook.ai/rics-standard/overview.md) covers what it requires, independently of the product.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.