We argued about who should own AI. We were asking the wrong question.

Hi Reader,

Last week I joined Todd Michaud on his podcast, the Work Intelligence Loop, alongside my partner from AI Leader's Compass, Giorgio Zampirolo. Partway through, we hit a question that turned into a bit of a debate.

Who should own AI in an organization?

Three of us had three different answers.

Todd went first, and he was direct about it. The CEO is the Chief AI Officer — who else could it possibly be? If you lead the organization, AI adoption is your job. It's too big and too strategic to belong anywhere else.

Giorgio saw it differently. To him, the title matters less than the knowledge behind it. Anyone can lead the AI effort, he said, as long as they have the skills and understand what makes the company valuable in the first place — its market, its expertise, the things it's good at. Lead from there, and let the technology scale it.

And then me. I've watched too many CEOs say "let's appoint an AI officer, and they'll fix it for us." They won't. Naming one person to own AI doesn't make the problem disappear — it just moves it somewhere else. My answer was that the CEO's real job is to build the culture and hold the vision, to make sure the right conversations are happening. But not necessarily to lead the process. That belongs to the team.

Three answers to the same question. But listening back, I don't think we were really disagreeing about AI.

We were disagreeing about what it means to be "in charge."

It's natural to expect the CEO to be responsible for everything. Culture. Transformation. Strategy. And that's fair — those things do sit with them.

But the way we interpret that responsibility is where it goes wrong.

When we say someone is "in charge," we usually mean they should make all the important decisions. That's the assumption baked into the word. And it's exactly the assumption that breaks under pressure — because leadership doesn't scale that way.

A leader who is the single point every important decision passes through doesn't get more powerful as the organization grows. They become the bottleneck. The thing everyone waits for.

The leaders I see doing this well are not trying to weigh in on everything. They're designing the conditions in which good judgment happens without them.

Their real work isn't deciding. It's designing the workflows and the communication flows so that decisions get made consistently — in line with their intent, their principles, their strategic direction.

That's a different definition of being in charge. Less about making the decisions yourself, more about building the system that makes good decisions possible.

I wrote about this in my latest Forbes Coaches Council piece, Leader Burnout: A Team Design Problem, Not A Personal One. It's about how leaders become bottlenecks — often out of genuine care — and why the fix isn't more personal discipline but a change in the underlying design.

Which brings me back to the AI officer question.

The reason it's so hard to answer is that we keep hunting for the right person to put in charge. But AI transformation isn't one decision someone makes well. It's a stream of them — which tools, which risks, which processes to redesign, when to step in — arriving faster than any single person can keep up with.

So maybe the question isn't who should own AI.

It's whether your organization is designed so the people closest to the work can make those calls well — consistently, and in the direction you're trying to go.

And that's not really an AI question. It's a design one.

See you next Thursday.

Daria

P.S. If you want the full back-and-forth — Todd, Giorgio, and me going a few rounds on AI, governance, and what actually breaks inside organizations — the whole episode is worth a listen.