Skip to main content
Where the Material Impact page records one firm-level judgement, the AI Use Case Register breaks it down: what your firm actually uses AI for, use case by use case. Open AI Use Case Register from the left-hand navigation. Four columns:
  • Use case, what the AI is used for
  • Client-facing, yes or no
  • Materiality, low, medium or high, or “Not yet assessed” until someone rates it
  • Actions, where you assess the use case
New entries start unassessed.

​
Where the entries come from

They are created automatically when you complete the onboarding questionnaire, from your answers about what your firm uses AI for.
You cannot add, rename or delete an entry on this page. Entries come from your onboarding answers. You can assess each one: select the pencil icon, choose a materiality level and give your reasoning. Any team member can do this.If your AI use changes, the firm owner can update the onboarding answer “What does your firm use AI for?” in Organisation, under your survey answers. You can add your own use cases there. The register is then rebuilt to match. Be aware that this clears every rating and reasoning already entered, so you will need to assess each use case again. The old ratings stay in the Audit Trail.
If a rating is wrong, assess the use case again. If a use case is missing, ask the firm owner to add it to your onboarding answers. The register is a section of the evidence pack, so whatever you record here is what a reader sees. In the evidence pack, some fields for each use case show “Not specified”. Oversight of individual outputs is recorded in the Oversight Log, not here.

​
What it is for

Clause 1.2 asks for a determination and reasoning at firm level. The use case register is the working underneath that judgement: it shows which specific uses drove the answer, and which of them touch client work. It is also the most useful prompt for a conversation your firm should be having anyway. A use case that reaches client work is one where clause 4.2 reliability decisions are most likely to be needed, and one your client disclosures should cover.
The client-facing column comes from one onboarding answer about whether your AI outputs reach clients, so it is the same for every use case. It does not tell you which uses reach clients. Use the reasoning on each assessment to say whether that use touches client work.