> ## 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.

# Client disclosures: what clause 4.3 requires

> The six things your governing documents must tell clients about AI use, and how guardbook produces the wording for your terms of engagement.

Clause 4.3 requires firms using AI with a material impact on service delivery to make clear to clients, in writing and in advance of use, when and for what purpose AI is to be used.

In advance is the part to hold on to. A disclosure that first appears in the report is not advance notice.

## The six required disclosures

Your terms of engagement, contracts, service agreements and other governing documents must detail, in writing:

1. when AI will be involved in delivering a surveying service
2. which parts of the delivery process it will be involved in
3. if available, the extent of professional indemnity cover for the firm's AI use
4. the internal process to contest the use of an AI system
5. the process to seek redress if a client feels negatively affected by AI use
6. how a client can opt out of AI use, if at all

Item three is the only place in the whole standard that touches PI insurance, and it does not require your firm to hold AI cover. It requires you to disclose the extent of whatever cover exists, if that is available to you. "We are confirming the position with our insurer" is an honest answer, and guardbook offers it as one of the options.

## How guardbook handles it

Open **Client Disclosures** from the left-hand navigation. It works in two stages.

**1. Answer five questions**

Three are required, because guardbook will not generate anything until it knows your firm's position. See [your disclosure answers](https://help.guardbook.ai/client-transparency/terms-of-engagement.md).

**2. Generate and adopt the statement**

guardbook builds a versioned statement, **Use of artificial intelligence in the delivery of our services**, from your answers and your Tool Register. It is written to form part of your terms of engagement. See [generate and adopt](https://help.guardbook.ai/client-transparency/disclosure-statements.md).

> [!NOTE]
>
> Only the firm owner can generate and adopt the statement. Members can read it.

The result is one firm-level statement. guardbook does not produce a separate one per client or per job. Where a particular instruction involves AI in a way your standard wording does not cover, that needs saying to that client separately.

## Two more client-facing duties

**Explaining an AI system on request (clause 4.4).** Section 8 of your statement promises clients written information about each AI system you use, if they ask. guardbook compiles that for you. See [client explanations](https://help.guardbook.ai/client-transparency/client-explanations.md).

**Telling a client an output could not be used (clause 4.2).** Where you conclude an AI output cannot reasonably be used for its intended purpose, you must tell the client in writing with your reasoning. Your terms of engagement do not cover this. See [outputs you rejected](https://help.guardbook.ai/client-transparency/unusable-outputs.md).


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