This article is also available in Danish.
Bestyrelse & AI

The board and AI: From passive recipient to active strategic player

By Daniel Wegener 12 April 2026 6 min read

A board turns passive when the meeting is built around what management wants to present rather than what the board needs in order to decide. The fix is not more slides. It is one shared picture that all four groups start from, analysed before the meeting instead of presented during it.

The board sits in the room with a drive full of slides. One presentation every ninety minutes. Finance, operations, people, all of it recited. A few questions are asked. A few answers are given. Next quarter the pattern repeats.

It is not bad work. It is just not strategic work.

The board as recipient

Most boards are arranged as recipients of information. The meeting is structured around what management wants to tell, not around what the board needs to know in order to make a better decision.

That produces three reliable problems:

The result is a board that sees problems once they are already urgent. It approves budgets and risk policies and says yes or no to proposals, but rarely shapes direction before the problems appear.

Four groups talking past each other

It helps to be precise about who is actually at the table. In the Leadership Room there are four groups, and each runs on a different time horizon:

  1. Owner or investor. Holds the vision and the values. Wants growth and stability.
  2. Board. Should look past next quarter and ask the uncomfortable questions.
  3. Chief executive. Translates strategy into action and balances ambition against reality.
  4. Management team. Executes this week and needs clear priorities.

The problem is not disagreement. The problem is that the groups use the same words for different things.

The board says "we should invest in innovation." The chief executive hears "a demand is arriving without a budget." The management team hears "another project on top of what we did not finish last year." None of them is wrong. They are simply starting from different pictures. How to get the four groups talking to each other is covered in the leadership room.

What changes when the analysis comes before the meeting

The real difference is not that a model writes faster. It is that the analysis happens before the meeting, so the meeting can be spent deciding rather than briefing.

The landscape has been worked through in advance. Instead of arriving unprepared, the strategic landscape is examined systematically, lens by lens: market, capability, digitalisation, ESG, risk. Each lens reduces hundreds of data points to a single question, namely what this means for us, here, now. How the lenses fit together is covered in from data to decision.

The work is visible. The meeting does not open with "we analysed it, here is the answer." It opens with "here is the picture, what do you see?" The board can see which assumptions were made, which data weighed most, where uncertainty is greatest, and what the analysis did not look at. The judgement stays with the board, but it is made informed.

There is one story instead of twelve. Board, chief executive and management team start from the same picture. The discussion stops being "I don't even agree with how the problem is framed" and becomes "given that this is the picture, what is the right next step?"

An example with arithmetic in it

A furniture factory with 45 employees has run the same product line for twenty years. Revenue is flat. Some people in the building think they should digitalise. Others think they should double down on craft. The board thinks the price needs to come down.

In the classic meeting that becomes three sets of slides and three conclusions, and the item is deferred.

With the analysis done beforehand the picture looks like this:

The story resolves: do not compete on price, live on craft sold digitally with customisation. That calls for investment in sales and design, not in production equipment.

Now the board can say yes to something concrete rather than to a word.

What you can do before the next meeting

None of this requires a purchase. It requires one person to do the preparation.

Try it in 90 minutes. Take the next board meeting and change three things:

  1. Ask for the material five days ahead, not the night before.
  2. Ask for one page with it: what we know, what is uncertain, and which two decisions this meeting has to make.
  3. Spend the first fifteen minutes asking whether anyone disagrees with how the problem is framed, before anyone presents a solution.

If disagreement surfaces in that quarter of an hour, you have found it before the decision instead of after it.

Three questions to ask yourself as chair:

Why we built the analysis as a chain

The reason twelve presentations never become one story is not laziness. It is that they were produced independently of each other.

In 360° Sprint the connections between nodes on the board are what determine the order the analysis runs in. A node receives its context from the nodes upstream of it, through the knowledge graph. The capability analysis therefore knows the market analysis's conclusion because it came earlier in the chain, not because somebody remembered to mention it.

That is the mechanical reason a coherent story comes out instead of twelve loose sections. The chain is described in the strategy chain.

What a board should demand before letting AI into that work is covered in three things your board should demand.

None of this makes the board operational. It gives the board clarity about its own role: ask the right questions, challenge the assumptions, hold the direction, and get owner, chief executive and management team speaking the same language.

If you are still weighing whether to have a board at all, and what it costs, start with boards in smaller companies.