Free · self-reported · no sign-up
Free AI Governance QuickScan
Answer 16 short questions about how your organisation uses and governs AI. In a few minutes you get preliminary observations by area, exactly three practical next actions and an evidence checklist — produced by fixed rules you can read, not by a model.
How it works
16 questions in 8 areas
Each question has an Unknown and a Not applicable choice. Unknown is never counted as a pass; Not applicable is left out of the area.
Fixed, explainable rules
Each area's level is its weakest applicable answer. Nothing is averaged, weighted or scored as a percentage, and no AI model is involved.
Critical concerns on their own
A few answers raise a critical concern. They are listed separately and are never offset by stronger answers elsewhere.
Three actions and an evidence list
You get observations per area, exactly three rule-chosen next actions and the evidence an assessor would ask to see.
The eight areas
The areas follow the governance questions that come up first in AI assurance work and map to the Govern, Map, Measure and Manage functions of the NIST AI Risk Management Framework. Every question is about what you could show, not what you intend.
- 01AI system inventory
- 02Ownership and accountability
- 03Authorization of AI tasks and agents
- 04Data handling
- 05Vendors and access
- 06Human oversight
- 07Incidents
- 08Evidence availability
What happens to your answers
- Your answers and results are produced and kept in your browser only (local storage on this device). They are not sent to AuditPartners and are not stored on our servers.
- You can change any answer, restart (which deletes the saved answers from this browser) or download your results as a text file or print them to PDF — all without contacting us.
- Nothing leaves your browser unless you choose to use the optional contact form on the results page. That form sends your name, e-mail, company and message to us; your QuickScan summary (area levels and flagged concerns, not your individual answers) is attached only if you tick that box.
- If analytics cookies are accepted, we record only that the QuickScan was started, completed or downloaded, with a fixed tool identifier — never your answers, levels, flagged concerns or anything else from your results.
- Details are in the Privacy Policy.
What this is — and is not
The QuickScan is preliminary, self-reported guidance for AI owners and teams starting a governance conversation. It reflects what you tell it and nothing is verified.
It is not an audit, a security scan, a certification, a compliance rating or a legal conclusion, and it does not test any control. It produces no percentage or overall score. A scoped engagement such as AI & Agentic AI Assurance is where evidence gets examined.
Want to see the output first? Read the fictional worked example for Northwind Claims Services (fictional).
Ready? It takes a few minutes.
Answer Unknown wherever you are not sure — finding out is often the first useful action.