You will find lists of twenty questions your board might ask about AI. Only three can hurt you, and the answer that keeps you safe in all three is the one no CEO wants to give: yes. Answered voluntarily, in the right order, those three questions stop being an ambush and become the strongest board meeting you have run in years. And the preparation starts before the meeting, at lunch.
Before the meeting: take the difficult director to lunch
If your board has a new member, often the younger one, brought in precisely to ask the tough AI questions, do the old-school thing. Invite them to lunch. Go to their office. Ask them directly: what are your expectations on AI? What should I be looking out for? How do you see it?
Get to know how they think and how they are wired. Their job is to help the company, and so is yours. If there are synergies in that conversation, they will tell you exactly what they will be looking for in the meeting, which is what you wanted to know anyway, and you gain a colleague on the board instead of an opponent. Human interaction is the one preparation tool that never gets replaced.
Open with the questions you fear most
Here is the sequencing secret. The questions you should prepare hardest for are the ones you absolutely do not want to answer, and you should open the meeting with them, unprompted. Eliminate the most embarrassing material first, before anyone can weaponise it, and the entire meeting changes character: the ambush never happens, the room relaxes, and you spend the remaining time on what you are proud of. Boards can smell the fear a CEO carries into an AI discussion. Removing it yourself, in the first ten minutes, is the whole trick.
Question one: “Have we failed with our AI strategy?”
Open with yes. Yes, we failed. Show exactly where, what you could have done better, and what you are changing. A board that hears its CEO name the failure before anyone else does stops worrying about what is being hidden, because evidently nothing is. Falling on your own sword this way beats every scapegoat structure ever designed, because it leaves you standing.
Question two: “Have we overspent on AI?”
Yes again, and show the exact numbers. Then show what exists now that did not exist when the money was spent: governance and policies that control the next round of spending. And add the line that turns the whole answer from defence into offence: the governance you have put in place is built to be compatible with the EU AI Act, which means you have not just fixed the spending, you have protected the company from fines reaching EUR 35 million or 7 per cent of global turnover. With the Act’s transparency obligations arriving on 2 August 2026, you may gently note that this is a question the board’s own legal voice might have raised with you six months ago. Asked or not, you prepared. That lands.
Question three: “Is our AI strategy world class?”
This is the trap question, and the one place where a blunt answer hurts you in either direction. Do not claim world class. Do not confess to a bad strategy. Say: here is our strategy. It is version 1.0, and we learned from it. Then move immediately to what you learned, how it has been improved, where the benefit will come, and, most importantly, in what timeframe and at what cost, in money and in the resources and people it will take. Boards do not need world class. They need a CEO who knows precisely where the strategy stands and what it costs to advance it.
Then the easy part, and how to end
With the heavy three cleared, the rest of the agenda is yours: the projects being implemented, the clients you are using AI with, the work you are proud of. End the meeting on the upside: where the company is going with AI, how it benefits, why clients are happier. The measure of success is simple. The board should leave the room less concerned about the company’s AI position than when the meeting began, and more convinced than ever that they chose the right CEO. That is the CEO still standing in 2027.
Frequently asked questions
What questions do boards ask CEOs about AI?
The three that matter: have we failed with our AI strategy, have we overspent on AI, and is our strategy as good as our competitors’. Everything else on a board agenda about AI is easier, and a CEO who answers those three honestly and first controls the entire meeting.
How should a CEO answer a board question about a failed AI strategy?
With yes, voluntarily, before being asked. Name where it failed, what could have been done better, and what is changing. Boards forgive experiments that failed. They do not forgive discovering a failure the CEO tried to hide.
How do I prepare for a board member who is an AI expert?
Meet them before the meeting. Invite them to lunch, ask their expectations of the company’s AI position and what they will be looking for, and learn how they think. Their mandate is to help the company, and a pre-meeting conversation converts the toughest questioner into an ally.
Should I tell my board our AI strategy is world class?
No, and do not call it bad either. Present it as version 1.0 that the company learned from, then show the improvements, the expected benefit, the timeframe and the full cost in money and people. Precision about where the strategy stands builds more board confidence than any superlative.
Thomas Anglero is a Strategic AI Advisor, keynote speaker and author of Intro to Artificial Intelligence. He has delivered over 450 keynotes across 30 countries for organisations including IBM, the WHO, the World Government Summit and the European Commission. He founded the IBM Watson AI Lab for Cancer at the Oslo Cancer Cluster and closed over $500 million in enterprise transformation deals as CTO and Chief Innovation Officer at Cognizant.
If you are leading your organisation through this, I work with a limited number of senior leaders each quarter. Get in touch at Anglero.com.