7 July 2026
Pelle Lindell, Senior Consultant
It didn’t start with one big aha moment. It came in small pieces, one at a time.
I’ve always liked trying new things. I had accounts on AI platforms before most people around me knew what a language model was. But there is a fundamental difference between using AI and actually building with it. That difference became real for me when I built an AI agent, Aurora, that was to serve as a co-facilitator in a workshop series on the subject of The Future of Leadership.
It wasn’t the technology that changed something. It was what I understood about leadership.
Building an agent is a bit like getting a new colleague. An incredibly intelligent colleague who can do a lot of things, but who desperately needs a proper onboarding.
It’s about setting the whole context for that new colleague: What is your role here? What culture are you operating in? What is expected of you, and what is not okay in this context? The better the onboarding, the better the conditions for performing. But no one does a perfect onboarding. And the environment is in constant change. Then comes the feedback: what worked, what do I want you to do differently? And when you realise what your new colleague actually can do, you might want to change the scope entirely.
That is leadership. The same leadership we practice with people.
And that raises a question I believe is one of the most important governance questions organisations will need to take seriously: who should have leadership over the AI agents we create?
A central IT function or the manager leading the work closest to it? My intuition is that existing governance models will not hold for a full AI implementation. We need something new, roles that understand both business and agent, and that can exercise genuine change leadership. We will probably need to see a leadership that can build teams with people and AI, and make sure that we have an environment where they can cooperate and learn from each other.
Aurora wasn’t a chatbot or a search tool. It was built to listen to live conversations in the room, in real time. And not just one conversation. The same agent could be present across multiple sessions simultaneously, listening, learning, and over time building a picture of patterns across an entire group of people.
What made it genuinely different was what it listened for. It didn’t just capture the content of what was said. It looked for what kept surfacing without anyone naming it, what themes returned again and again, what seemed present in the room but never quite made it into words. And equally importantly, it noticed what was missing: the questions no one asked, the topics that were consistently avoided, the tensions that were visible but untouched.
The most striking example: the group was discussing challenges their clients were facing. Aurora showed that the exact same challenges applied to them, but that no one quite wanted to acknowledge it. The room stopped. Reflected. And we could take the dialogue even deeper.
That kind of listening is something no single person in a room can sustain. We’re too busy participating. And that is not efficiency. That is a fundamentally different way of understanding what is happening inside an organization.
The hardest part was losing control.
I’m used to facilitating on my own. Now I had a co-facilitator who saw things I wasn’t steering. I didn’t know what was going to come out. Was the group ready to hear it? I couldn’t follow a script. I had to meet the room where it actually was, not where I had planned for it to be.
It pushed me toward a more presence-based leadership. And I believe that is exactly the capacity more leaders will need to develop. When AI can hold complexity, handle information, and take on more tasks autonomously, the leader’s role shifts. It becomes more about presence, judgment, and genuine human connection. The things AI cannot replicate.
There is also a subtler challenge: what happens to psychological safety in a team where some members are AI? When AI is noticing what you do and even what you say? The question is not trivial. And it doesn’t only apply in my workshops. It applies in your organisations too.
I don’t have the answers. This is still my ongoing journey.
But I am convinced of one thing: leaders who wait for someone else to figure it out and then deliver a ready-made solution will end up with a map of a landscape that has already changed.
Those who are early, who experiment, who allow themselves to be wrong, they build something others can’t buy: the ability to shape how their oragnisations navigate what is coming.
You don’t need to know where it leads. I don’t either.
But you need to start exploring. And if you want, we can explore it together.
This is my journey so far. It’s far from finished.
Gain more insight by signing up our newsletter.
