AI is not an IT issue. It is a question of how leadership works

I use AI every day. I have a digital assistant, Ivar, who handles some of my monitoring and some of my correspondence. It has taught me more about what AI can and cannot do than every conference I have attended. And it has convinced me that leadership teams who do not try it themselves will make worse decisions about AI.

Leadership is the bottleneck

When Sweden's AI Commission presented its roadmap in 2025, it pointed to a lack of leadership, a conclusion the Swedish Association of Local Authorities and Regions endorsed in its consultation response. That matches what I see. The technology is there, and it keeps getting cheaper. What is missing is leadership teams who know what they want to use it for.

In many organisations, AI has ended up with the IT department or an innovation group. That is understandable, but it often leads to pilot projects that never become part of the business.

Ivar started as a sounding board. Today he helps me keep things in order, handles part of my communication and prepares material. He has taught me more about how AI can be used in organisations, for workflows and reporting, than anything else. But the most important lesson is that AI does not run itself. Ivar has to be followed up and given new instructions, just like a member of staff. We spend fifteen minutes a week getting better together: what worked, what went wrong and what should be done differently. It sounds simple, and it is. But that routine is exactly what makes it work.

Why leaders have to try it themselves

You cannot judge what you have not used. A leadership team that has never let an AI summarise a decision paper, draft a document or review a contract will struggle to tell what is realistic.

The risks only make sense once you see them. AI is sometimes wrong, and it is wrong with great confidence. It is easier to set sensible rules once you have seen that happen.

Your staff are already doing it. In most organisations, many people already use AI tools, with or without permission. The question is whether leadership wants to steer that or pretend it is not happening.

Three steps for a leadership team

  1. Have everyone on the leadership team use an AI tool for a month for their own work, and share at a meeting what worked and what did not.
  2. Decide what must not be done, for example which information must never leave the organisation. That matters more than a long policy.
  3. Pick one or two use cases where AI can make a real difference in the business, and give them an owner in the line, not in IT.

A question for your next leadership meeting

How many people on our leadership team have used an AI tool for their own work in the past month?

Sources: SKR consultation response on the AI Commission roadmap, SOU 2025:12 (Swedish) · Akavia, AI in municipal service, 2025 (Swedish)

What does it look like in your organisation?

Take the AI maturity assessment for leadership teams to see where you stand, or book 30 minutes with me about getting started safely.