You run a leadership programme. Last spring you added a module called "AI for leaders", a good speaker, a nice deck. The feedback forms were warm. Six months later, ask the participants what they do differently on a Monday morning, and you get a polite silence.
The short answer: AI should not be a module in your leadership programme. It should be how the programme works. Four changes do most of the job: start each leader from a real problem, put peers at the same level in one room, make them practise on their own files between sessions, and measure usage a month later instead of satisfaction on the day.
Why the "AI module" does not stick
A module teaches about AI. A leader needs to use it on something that is keeping them awake. Those are two different things, and only the second one changes a calendar.
I have seen this from the other side. I once ran a session for a large production company where I was pushed to arrive with spectacular use cases, to impress the top team. We lost everyone, and people left thinking their job was about to disappear. A demo makes a decision possible. It does not make one.
There is a quieter flaw too. In most programmes, the leaders who know the least say nothing, because the room contains their boss, their peers from other divisions, and the facilitator. Nobody admits in public that they have never opened the tool.
Four changes to make to the programme
1. Start from the problem, not from the tool. Ask each participant for the one file they dread: the board paper, or that supplier negotiation. We call these the pebbles in the shoe. You enter AI through the problem, never to look at the AI. A questionnaire about needs does not replace this, because someone who has never used the tool cannot tell you what it could do for them.
2. Put peers in the same room. A cohort of executives at the same level changes what people dare to say. In our 12-week bootcamp, there is a live session every week, and the best moments are often a participant showing another how they handled the same file. This is my observation, not a measured result.
3. Make the practice happen between sessions. The live hour matters less than the days around it. Give each leader a first use case to test within the week, then ask for it back. The format is close to what we run in one-to-one coaching: a short exchange, a real file, a next step with a date.
4. Measure what people use, not what they liked. Satisfaction scores tell you the speaker was good. Ask instead, a month later: which use did you keep, and what did you stop doing by hand? Three or four answers per leader are enough. If the answer is "nothing", the programme did not work, whatever the form said.
What stays the same
Leadership development was never only about tools. Delegation and difficult conversations still need the classic work. What changes is that AI now touches each of them. A leader who delegates a first draft to an assistant has to decide what must stay human. A leader whose team uses AI without rules has a governance problem, not a technology problem. I would fold these questions into the existing modules rather than isolate them.
The paper trail
If you operate in the EU, Article 4 of the AI Act asks companies to take measures to promote AI literacy among their staff and the people using AI tools on their behalf (Regulation (EU) 2026/1744 of 8 July 2026). A leadership programme with a clear syllabus, attendance records and a follow-up of usage gives you something to show. It is one measure among others, not a certificate of compliance. We cover the rule in AI literacy training under the EU AI Act.
This week's move
Take your current programme and ask three or four participants from the last cohort one question: what did you use from the AI module last month? Write down the answers. If you hear a file and a date, keep the format. If you hear a polite silence, start with change number one.
If you are choosing between formats for your senior team, executive AI programs compared sets four questions next to the brochures.
Our live cohorts run in French today. If you lead in English and want the same format for your leaders, join the waitlist: you will hear first when an English cohort opens. In the meantime, see AI training for executives.
Join the waitlist →Frequently asked questions
How do you use AI in leadership development?
Treat AI as the way the programme works, not as one more module. Start each leader from a real problem, practise on their own files between sessions, and measure what they actually use a month later. A half-day lecture on AI trends changes nothing.
Do leaders need a separate AI course, or should AI sit inside the leadership programme?
Both can work. A separate cohort of peers at the same level lets leaders admit what they do not know without an audience. Then the leadership programme can take the decisions further: usage policy, delegation, team rituals. What fails is a single module wedged between two classic ones.
