Pillar 02 · Scope
S1 AI Use-Case and Agent Register
You can't govern AI you haven't found.
Form Workbook
What it's for
This register is the inventory the Governing for Answers framework asks for in Chapter 2 (Scope) and the owner record Chapter 4 (Accountability) depends on. It lists every AI use case, every agent, and every AI feature switched on inside tools you already use, with a business owner, the sources each one reads, and a risk tier that sets how much governance it needs before go-live.
Gate one of the framework's roadmap is simple: every live AI use case and agent is in this register with a named owner.
What you get
- Register The inventory. One row per use case, agent or SaaS AI feature.
- Summary Measures for your governance report, worked out from the register.
- Tier gates The three risk tiers, the four rules and the Low-tier self-certification checklist.
Worked out for you Risk tier, Gate before go-live, Next review due, Review status
How it connects
Each use case names the S2 sources it reads; S3's scans find the AI nobody registered, and S4 checks the model providers. Its tiers drive the rest: M4 builds a context pack per use case, A1 and A3 set a check's severity by the highest tier reading the data, and C1 to C4 give each use case its owners, RACI, charter and flagged answers.
A flavour of the worked example
| Use case or agent | Invoice coding agent |
|---|---|
| Type | Agent |
| Q1 · What is the worst a wrong answer could do? | Mislead a business decision |
| Q2 · Does it act without a person checking each action? | Yes |
| Risk tier worked out | High |
| Gate before go-live worked out | Full review, DPIA and AI impact assessment, senior sign-off |
A few fields from one record of the toolkit's worked example, with long text shortened.
Next step
Build it with the guided assistant
The guided builds are open to invited people: you sign in first.
Request access
Not invited yet? Ask for access to the guided builds.