Skip to content
PTRequest the Diagnostic

Free self-assessment

AI governance maturity self-assessment

A quick read of how your company governs AI use: 15 questions in 5 dimensions, a 0 to 100 score per dimension and three priorities. It takes about 6 minutes, the result shows on screen and is computed in your browser.

  • 15 questions
  • 5 dimensions
  • About 6 minutes
  • No sign-up to see the result
Start the self-assessmentSee the methodology

Questionnaire

Answer for what happens today, not for what is planned

Pick the level that describes current practice. If torn between two, take the lower one: the score is for deciding, not for pleasing.

Step 1 of 50 of 15 answered

Inventory and use

Knowing where AI is used, with which tools and at what cost.

01Do you know where AI is used in the company today?
02Do you know how much the company spends on AI?
03Have the AI tools in use gone through any approval?

Answers stay in this browser. Nothing is sent to Reveno.

Methodology

How the score is calculated

The formula is open and the same for everyone. You can redo the math by hand with your answers.

The formula

Dimension score
100 × Σ(weight × level) ÷ Σ(weight × 4), rounded. The question that is the foundation of the dimension has weight 2; the others, weight 1.
Overall score
Simple average of the scores of the dimensions that apply, rounded.
Not applicable
Only on the agent execution questions. The question leaves the math; if all three leave, the whole dimension leaves the average.
Lock
If the question on sensitive data or on approval before action is at 0, the overall band cannot go above First rules, whatever the score.

The five levels of each answer

  1. 0Does not exist
  2. 1Informal
  3. 2Partial
  4. 3Defined
  5. 4Applied and reviewed

Overall score bands

  1. 0 to 24Improvised. AI is already in the company, with no list, rule or owner. The risk exists and nobody can size it.
  2. 25 to 49First rules. There are initiatives and some guidance, but each area still does it its own way. The rule depends on who is watching.
  3. 50 to 74Organized. A list, a policy and owners exist. The rule still has to reach execution and be measured on a routine.
  4. 75 to 100Governed. The rule holds at execution, is recorded and the board decides with numbers. The work now is to keep it.

How the three priorities are chosen

  1. Questions below level 4 are candidates.
  2. Critical questions (sensitive data and approval before action) at level 0 or 1 come first.
  3. Then, highest impact: (4 − level) × weight.
  4. Ties: dimension order, from Inventory to Value, then question order.
  5. At most one per dimension; if short, fill with the next ones.
  6. Level 0 or 1 gets the foundation action; level 2 or 3 gets the next step.

No AI picks the priorities: it is the rule above, applied in your browser.

The five dimensions

Inventory and use

Knowing where AI is used, with which tools and at what cost.

  • Do you know where AI is used in the company today?weight 2
  • Do you know how much the company spends on AI?weight 1
  • Have the AI tools in use gone through any approval?weight 1

Policy and owners

A written rule, an owner for each use and a clear path to approve what is new.

  • Is there a written rule for AI use?weight 2
  • Does each AI use have an owner?weight 2
  • Who decides on new AI uses?weight 1

Data and privacy

What may go into the tools, what the vendor does with it and data protection law.

  • What kind of data goes into AI tools?weight 2critical
  • Do you know what each AI vendor does with the data?weight 1
  • Were AI uses involving personal data assessed under data protection law?weight 1

Agent execution control

Who checks before AI acts, what gets recorded and which access it has.

  • When an AI carries out an action (send, change, approve), who checks first?weight 2critical
  • Is every AI execution recorded?weight 1
  • What access do AI agents and automations have?weight 1

Value measurement

A metric per case, priority by criteria and board-level follow-up.

  • Do you measure the results of AI uses?weight 2
  • How are AI use cases prioritized?weight 1
  • Does the board follow AI on a routine?weight 1

References

The dimensions map to the NIST AI RMF functions (govern, map, measure and manage) and to ISO/IEC 42001, the backbone of Reveno's method. The self-assessment is not an audit, a certification or a legal opinion.

Method version 1, published on 2026-11-02.

What the self-assessment does not measure

The score is what was declared. The Diagnostic checks it.

  1. 01

    Actual use

    You answered from what you know. The Diagnostic crosses interviews, contracts, access and invoices, and usually finds use nobody declared.

  2. 02

    The real cost

    There are no amounts here. The Diagnostic reaches spend by area and tool, including what sits on cards and personal accounts.

  3. 03

    The risk of each case

    The self-assessment looks at the whole company. Risk lives in the use case: which data goes in, which action comes out, who approves.

  4. 04

    Differences between areas

    One person answered for the company. Different areas often score the same question differently.

  5. 05

    The plan with owners and dates

    Three priorities are not a plan. The Diagnostic delivers the 30/60/90 roadmap with an owner for each step.

  6. 10 days. Inventory, risk and invisible cost map, maturity score per dimension and 30/60/90 roadmap.

    See the Executive AI Diagnostic

Frequently asked questions

How long does the self-assessment take?

About 6 minutes. There are 15 multiple choice questions, in 5 steps of 3 questions.

Do I need to sign up to see the result?

No. The result appears on screen as soon as you answer the last question. Sign-up exists only if you want the short roadmap, through the Diagnostic form.

Are my answers sent to Reveno?

No. The score is computed in your browser. Answers live in the result link, after the # sign, a part the browser does not send to the server. Only what you send through the form reaches Reveno.

How is the score calculated?

Each answer is worth 0 to 4. The dimension score is 100 × Σ(weight × level) ÷ Σ(weight × 4) and the overall score is the average of the dimensions. The full formula, weights and bands are in the methodology on this page.

Who should answer?

Someone who sees the whole company: CEO, COO, CIO, Compliance or HR. It is worth answering with more than one area and comparing links: the gap between scores is already a finding.

Does the self-assessment replace an audit or a data protection assessment?

No. It is a declared reading, with no evidence check. For data protection law and contracts, validate with legal or the data protection officer.

How is it different from the Executive AI Diagnostic?

The self-assessment shows what was declared in 6 minutes. The Diagnostic checks it in 10 days with interviews, contracts, access and invoices, and delivers an inventory, a risk and cost map, a score per dimension and a 30/60/90 roadmap with owners.

Can I share the result?

Yes. The result link carries the answers and opens the same reading in any browser. It is not indexed by search engines.