Guides

How to Measure Your Board's AI Governance Maturity

Tamara Gracon November 2026 7 min read

Most AI governance assessments produce a single score. A single score cannot tell a board whether it is doing well, because the same level of oversight can be more than enough for one company and dangerously thin for another.

Key takeaways

  • A useful AI governance assessment measures two things separately: AI exposure, meaning how material AI is to the business, and oversight maturity.
  • Exposure sets the priorities. Maturity shows the gap. The combination tells the board where to act first.
  • The Vela Board Navigator℠ rates maturity on four levels: absent, emerging, established and evidenced.
  • The step most boards miss is from established to evidenced: having a practice is not the same as being able to show it.

Why doesn't a single score work?

Consider two companies whose boards give themselves the same answers on AI oversight. One is a manufacturer using AI for internal document search. The other is a lender using AI to approve loans. Their oversight looks identical on paper, but the lender's board is carrying far more risk with it. A single maturity score treats them the same.

That is why the Vela Board Navigator℠ separates two questions. First, how material is AI to this company's products, operations, customers and strategic position? Second, how mature is the board's oversight in each area? The first question tells the board which areas matter most for it. The second tells it how far it has to go.

What does AI exposure measure?

Exposure is about the business, not the board. It asks where AI changes the business model or sources of value, how dependent the company is on AI vendors and models, and how many consequential decisions AI now makes or informs. In the Navigator, this sits mainly in the element called AI exposure and strategic dependencies.

Exposure changes the weight of the other 15 elements. For a company whose AI is mostly internal productivity, workforce transition and investment discipline may matter most. For a company whose AI makes decisions about customers, transparency and redress, testing and incident response, and third-party concentration move to the top.

What are the levels of governance maturity?

The Navigator rates each element on four levels:

  1. Absent. The board has not addressed this area.
  2. Emerging. The board has discussed it, but responsibilities, information and processes are informal or incomplete.
  3. Established. A defined practice exists: a committee owns it, reporting arrives on a schedule, and decisions follow a process.
  4. Evidenced. The board can show the practice working, from its own minutes, materials and decisions.

The step from established to evidenced is the one most boards miss. A board may have a committee charter that mentions AI and a regular agenda item, and still be unable to show what it asked, what it learned or what it changed as a result. In a dispute, an inquiry or a transaction diligence process, the difference matters.

How should a board read the results?

The useful output is not a score but a short list: the two or three elements where exposure is high and maturity is lowest. Those are the priorities for the next year. Elements where exposure is low can wait, even if maturity is also low.

The results should also prompt discussion, not settle it. In our sessions, the most valuable moment is often when directors on the same board rate the same element differently. That disagreement shows where the board does not share a picture of its own oversight, and it is usually where the next boardroom conversation should begin.

How can a board get started?

The Board Snapshot is a short version of this approach: five questions about your company and your board's oversight, producing a one-page brief with your most significant gap and questions for your next committee meeting. Boards that want the full picture can work through all 16 elements with us in a facilitated session, described on our services page.

To compare your board's approach with other directors', The Vela Barometer publishes how directors approach consequential decisions. Our 2026 research report, The Board's AI Moment, sets out what effective AI oversight now demands of directors.

Common questions

What is an AI governance maturity model?

A way of rating how developed an organization's AI oversight is, usually on a scale from no practice to a fully evidenced one. For boards, it should be read alongside how material AI is to the business.

What is the difference between AI exposure and AI governance maturity?

Exposure measures how much AI matters to the business. Maturity measures how well the board oversees it. High exposure with low maturity is where a board should act first.

How long does a board AI governance assessment take?

The Vela Board Snapshot takes a few minutes. A full assessment across all 16 Navigator elements is usually done in a facilitated board or committee session.

Ready to look at your board through the Navigator?

The Navigator is where every Vela engagement begins. Request access or a briefing to see what it surfaces for your board.