Sceptical executives are rarely against AI itself. They are against wasted money, reputational risk and being sold hype. The fastest route to buy-in is to stop asking for a big AI investment and start asking for a small, capped experiment with clear success criteria, run on a real problem they already care about.
Understand what the sceptic is really worried about
Before you pitch, find out which objection you are actually facing. In our experience it is usually one of these:
| What they say | What they often mean | What answers it |
|---|---|---|
| "It's just hype." | "I've seen tech fads waste money before." | Evidence from a small test in our own business |
| "It's too risky." | "I don't want a data breach or embarrassing error." | Clear data rules, approved tools, human review |
| "We can't afford it." | "I don't believe the return." | Capped pilot cost with measurable targets |
| "Our people won't use it." | "Change programmes usually fail here." | A training plan and named champions |
| "Not now." | "I have bigger priorities." | Link the pilot to one of those priorities |
Five tactics that work
1. Tie it to their problem, not to AI
Do not lead with "we need an AI strategy". Lead with "proposals take our senior people two days each; I think we can halve that". Executives fund solutions to problems they recognise.
2. Ask for less than they expect
A request for a large programme invites a large debate. A request for a 90-day pilot with a handful of licences and a training budget is easy to approve and easy to stop.
3. Show, do not tell
Run a live demonstration on one of their own tasks: summarising a board paper, drafting a reply to a difficult email, analysing a report. A sceptic who sees a useful result on their own work in five minutes often changes position faster than after any slide deck.
4. Answer risk before they raise it
Come with a one-page safe-use plan: which tools, what data is excluded, where human review happens, who is accountable. Sceptics relax when they see someone has thought about what could go wrong.
5. Define success and an exit
State the baseline, the target and the date you will report back. Make it explicit that if the target is missed, you will stop. This is the single most persuasive thing you can say to a cautious executive.
The one-page proposal
Your proposal should fit on a page:
- Problem: one sentence, with a number.
- Pilot: which team, which task, how long.
- Cost: licences, training, people's time.
- Safeguards: data rules and human checks.
- Success measure: baseline, target and review date.
- Ask: the specific approval you need today.
After the pilot
Report back on the date you promised, with the numbers, the problems you hit and what you recommend next. Credibility from one honest report is what unlocks the second, bigger budget. For the board-level conversation that follows, see how to talk to your board about AI investment, and for a deeper financial case, how to build a business case for AI.
Building credible proposals, running pilots and reporting results are central to our Leading AI Innovation & Organisational Transformation course, which finishes with you presenting a real AI plan for your own team.
Frequently asked questions
What if the executive has had a bad experience with AI before?
Acknowledge it directly and ask what went wrong. Then show how your proposal avoids that specific failure, usually through tighter scope, better training or clearer measurement.
Should I bring in an outside expert to persuade leadership?
It can help, but evidence from your own organisation is usually more persuasive than any external voice. A small internal pilot often beats a consultant's presentation.
How do I handle an executive who simply says no?
Ask what evidence would change their mind, then go and get it at minimal cost, for example by showing individual time savings from approved tools already available.