Nine questions, five stages. How a board oversees AI.
Most boards get an aggregate answer when they ask where the company is on AI. This paper sets out the nine questions a board should ask every quarter, a five-stage scale that separates saying from doing, and what governed looks like for each question. Written for chairs, NEDs, PE operating partners and the CEOs who report to them.
AI is not adopted by a company. It is adopted inside individual workflows, and each one carries its own risk and its own case for AI. An aggregate answer hides the detail a chair needs, and it tilts towards cost savings because savings are easy to count.
This paper gives a board two things: a short, fixed set of questions to ask every quarter in the same order, and an honest scale for how far each answer has got. It pairs with the interactive AI Board Diagnostic, which scores a board on the same questions in about ten minutes, and the Claude prompt, which scores them from real evidence.
What is inside
- Why "where are we on AI?" gets the wrong answer, with what 120 companies disclose and what NHS board packs show
- The five stages: not discussed, discussed, owned, measured and governed, and the rule that a policy on its own is never proof
- The nine questions, each with why it matters and what owned, measured and governed look like, including where the EU AI Act, ISO/IEC 42001 and the NIST AI RMF fit
- A worked example: one fictional board scored on all nine, the question to fix first and a 90-day plan
- Running it every quarter: the persistence rule, two people scoring separately, and what goes in the board pack
Who it is for
Chairs, non-executive directors, PE operating partners and CEOs of B2B software companies. It is general guidance for boards, not legal, regulatory or financial advice.
Written from an operating perspective
By Andrew Wyatt, Principal of Ortent Advisory. Four exits across enterprise software and telecoms, over 30 years operating, and board roles across PE and VC-backed businesses since 2006. The paper is the companion to the field guide AI and the Board.