Jacob Langvad Nilsson — Programme Leadership, Digital & AI, Copenhagen
All writing

Insights

What a Fractional CAIO Looks Like in a Danish Company

The fractional Chief AI Officer is an American idea. In a Danish company it works differently, because decisions are made differently. Here is what changes, and what the seat has to respect to work.

  • AI
  • AI leadership
  • fractional CAIO
  • Denmark
  • governance

· 5 min read

Most of what is written about fractional Chief AI Officers comes from the US. The model there is simple. A senior person takes the AI seat for a few days a week, gets a mandate from the CEO and starts deciding. The org chart does the rest.

In a Danish company the org chart does less of the work. Decisions are made through agreement more than through instruction, and a new senior voice without the trust of the organisation gets polite nods and very little else. The fractional model still fits most Danish mid-sized companies well. But it runs differently, and the differences decide whether it works.

Decisions are made in the room, not from the top

Danish management is flat by international standards. A direktør can decide alone, but the culture expects them to bring people along first. Middle managers push back openly. Specialists expect to be heard before their tools change.

For a fractional CAIO that means the mandate from the top is necessary and nowhere near sufficient. If the seat decides which AI tools the company uses and announces it on a Monday, the decision will be followed on paper and worked around in practice. People will keep their personal accounts and their own workflows, and the official tool will sit unused.

What works is slower at the start and much faster after. Decisions go through the ledelsesgruppe with the reasoning on the table. The people who will live with them get a real say in the use cases. Then the decision holds, because it was made with them and not for them.

The board asks different questions

Danish boards have started asking about AI, and the questions are mostly about risk and responsibility. Who owns it. What data goes where. What happens when it gets something wrong.

A fractional CAIO in a Danish company should expect to spend real time with the bestyrelse in the first months. That is not a distraction from the work. A board that understands what the company does with AI and why gives the direktion room to move. A board that feels kept outside will slow everything down with good intentions.

I have written separately about what to tell the board about AI. The short version is that boards want to see ownership and a way of stopping things, more than they want to see ambition.

Employees are part of the decision

Danish workplaces have a strong tradition of involving employees in changes that affect how they work. In companies with a cooperation committee, new technology that changes jobs is the kind of thing that committee expects to discuss before it happens, not after.

This is good news for AI adoption, even if it feels like friction. The committee is where worries about surveillance, workload and job security get said out loud. If those worries are handled there, they do not turn into quiet resistance later. A fractional CAIO who treats the committee as a stakeholder and brings concrete use cases to it gets a lot of goodwill. One who treats it as a box to tick gets the opposite.

Regulation lands on people who already have other jobs

The EU AI Act applies the same way in Denmark as everywhere else in the union. The regulation itself is long, and the national authority work sits with the Danish Agency for Digital Government. What is particular to Danish companies is the size of the team that has to deal with it.

In most mid-sized Danish organisations there is no AI compliance function. The work falls on a legal counsel who also does contracts, an IT manager who also does infrastructure, or a CFO who also does everything else. A fractional CAIO can take the parts of that load that are really about decisions. Which use cases are in scope. What the inventory says. Where the risk sits. Then the specialists can handle their part without carrying the whole thing.

Small market, long memory

Denmark is a small market. People know each other across companies, and a reputation travels fast in both directions.

That works in favour of a fractional seat that behaves well. Someone who has done the job in another Danish company comes with references people can actually check. It also raises the bar. A fractional CAIO who pushes a vendor they have a stake in, or who leaves a mess at the end of an engagement, will not get many more engagements here. Independence is not a nice extra in this market. It is the condition for doing the work at all.

What the seat looks like in practice

In a Danish company of a few hundred people, the shape that tends to work looks like this.

  • One or two fixed days a week, the same days every week, so people can plan decisions around them.

  • A seat in the ledelsesgruppe, with the decision rights for AI agreed in writing at the start.

  • A regular slot with the board in the first six months, then quarterly.

  • The cooperation committee brought in early, with real use cases rather than a policy document.

  • A clear end. Either a permanent leader is hired into a role that is now understood, or the decisions move into an existing role.

None of this is exotic. It is how good change already happens in Danish organisations. The fractional CAIO who works with that culture instead of around it gets further in a year than one who imports a model built for a different way of deciding.

If you are considering the model, I have written about when a fractional seat fits and when to hire permanently. How I run the engagement is on the fractional CAIO page.

Initial consultation. Thirty minutes.