Sunday night, 80 pages of reporting to read before Monday's leadership meeting. Every department sent its numbers in its own format, and nobody had time to write the one note that says what actually needs deciding. When you suggest using AI for this, you get the eye roll. "We already did the digital transformation, thanks."
The short answer: AI prepares the meeting, it does not make the decisions. Before the meeting, it turns the board pack into a one-page note per topic and lists the questions each item will raise. Afterwards, it turns your notes into an action log. Use it on your own first, in a tool your company has approved, and show the team the result instead of pitching them AI.
I have spent the last two years working with executive committees on exactly this. The gap I see everywhere is the same: everyone agrees it is urgent, and nobody knows where to begin.
What AI actually prepares for a leadership meeting
Start with the decision note. Hand over the full pack and ask for one page per topic: the decision needed, the options, the numbers that matter and what is still missing to decide. Then read it. It sometimes gets an amount wrong, so every figure in the note gets checked against the source document before it reaches the table.
Then the objections. Ask it what your CFO is going to push back on for each agenda item. That is often where prep saves the most, because you walk in with the answer instead of promising it for next time.
KPI consolidation. When every department sends its own spreadsheet, AI puts them in one format and flags the changes since last month. It does not replace your finance team. It spares them the copy-paste.
And the action log, after the meeting. Your notes, or the transcript if everyone agreed to be recorded, become a list of who does what by when. Sent the same day, not the following week.
One rule I give every leader: AI does not get a vote. No recommendation it drafted goes into the room until you have read it. And shit in, shit out: a badly built pack gives you a badly built note, AI will not rescue sloppy reporting.
What never goes into the tool
A leadership pack holds the most sensitive material a company has: unpublished results, HR files, sometimes an acquisition. France's national cybersecurity agency, ANSSI, is blunt about it. Its recommendation R34 says not to use generative AI tools on the Internet for professional use involving sensitive data (Security recommendations for a generative AI system, ANSSI, 29 April 2024).
In practice: a company account, in a tool your IT team has approved, never a committee member's personal account. If the tool is not approved yet, work on an anonymised version of the pack, or wait. In our sessions, confidentiality comes up often when participants compare tools. That is good news: the question comes up before the incident, not after.
Training matters too. Since the EU's digital omnibus, Article 4 of the AI Act asks companies to take measures that support AI literacy among their staff and anyone using these tools on their behalf, without requiring a guaranteed level for each person (Regulation (EU) 2026/1744 of 8 July 2026, Article 1, point 5). A leadership team that preps its meetings with AI is part of that group.
Prep one meeting yourself before you talk about it
The first mistake is walking in with a slide deck on trends and billions invested. Your CFO does not care. Neither does your head of HR. They want to know what changes for them, this week.
So do it the other way round. Take the prep for the next meeting and handle it with AI, on your own, in an hour. At the next meeting, you do not talk about AI, you put the result on the table: "I prepped the board summary in 40 minutes instead of three hours. Here is how." Now you have everyone's attention, because you fixed a problem everyone was living with.
The boss has to go first or nothing moves. Before you ask anything of the team, spend two weeks using AI yourself, on your emails and your meeting prep. You will learn what it does well and where it slips, and you will be able to talk about it in your own words, not a consultant's.
One CEO I work with started by dictating his meeting notes to AI and having them restructured in under five minutes. Three weeks in, his head of sales asked him how he turned that around so fast. That is where the conversation starts. In a corridor, not in a presentation.
Find the ally, let the skeptic watch
You are not going to win over the whole leadership team at once, and that is fine. Look for the person ready to try: not necessarily the most technical, more likely the one who spends evenings compiling and formatting. Often the finance director or the head of HR. Save them two hours on their own prep and you have your first champion.
In one seven-person leadership team I worked with, we started with two volunteers. Six weeks later, the other five had asked to be trained. Nobody forced them, they could see the difference in the quality of the files.
Put AI into the meetings you already have
The mistake to avoid is a separate "AI committee", with its own meetings and its own slides, cut off from everything else. AI does not need its own line on the agenda. It belongs in the prep for every line.
When the CFO presents the forecast, they also show the scenarios they tested. When HR runs the headcount review, they have already cross-checked skills against open roles. None of that needs a budget or a reorganisation, just a new reflex in the same meetings.
Four questions that get a leadership team moving
When I run a diagnostic with a leadership team, I start with these.
"Who here uses AI at least once a week in their actual work?" Usually one or two hands go up. Sometimes none, even though everyone says it matters.
"Which part of preparing this meeting takes you the most time every month?" Now everyone has an answer, and it is almost always a task AI can shorten.
"Which piece of data must never leave this room?" The answer sketches your usage rules better than a ten-page policy.
"If a competitor prepared in two days what takes you two weeks, what would you do?" The silence after that one is worth more than any presentation.
What it looks like after six months
I am not going to promise you a revolution. In month one, everyone tests on their own files, and that is healthy. In months two and three, practices spread: the CFO shows how they prep, HR borrows from it, and meetings change pace because the files arrive ready.
By month six, AI is no longer a topic. It is a reflex. And the team that was hesitating before the summer is now asking the opposite question: "Can we push this further?"
What waiting costs
Using AI badly is a risk. Not starting is a bigger one. For France, Institut de l'Entreprise and McKinsey estimate that annual productivity growth could reach 3% by 2030 with fast AI adoption, against 0.3% with slow adoption (press release in French, 7 January 2025).
The good news is that you do not have to do everything. BCG found that the companies furthest ahead on AI focus on 3.5 use cases on average versus 6.1 for the rest, and expect 2.1 times the return (From Potential to Profit: Closing the AI Impact Gap, BCG, 15 January 2025). For a leadership team, meeting prep is a very good first use case. You do not need a strategic plan or a Big Four consultant to launch it, just one concrete result at the next meeting.
Frequently asked questions
How do you prepare a leadership team meeting with AI?
Give it the full board pack in a tool your company has approved and ask for one page per topic: the decision needed, the options, the numbers that matter and what is missing to decide. Have it list the likely objections. After the meeting, turn your notes into an action log. You review everything and check every number against the source document.
How do you get your leadership team to back AI?
Do not start with a budget. Start with a result. Take a real irritant from your leadership team, solve it with AI in under an hour, and put the result on the table at the next meeting. One real result persuades more than ten slides.
Do you need a Chief AI Officer to get started?
No. A job title is not a practice. What you need is a leader who uses AI themselves and sets the example. If nobody on the leadership team actually uses it, a Chief AI Officer will not change that.
Which AI use case should a leadership team start with?
Start with whatever eats time and annoys everyone: preparing the meeting itself. Board pack summaries, KPI consolidation, action logs, board prep. These tasks are visible and repetitive, and the payoff shows at the very first meeting.
Want to run this exercise with your whole leadership team, on your own files? Our programs are delivered in French today. If your team works in English, join the waitlist and you will hear first when an English programme opens.
Join the waitlist →
