← All articles
Guide14 min readApril 5, 2026

The Best AI Prompts for Executives: 15 You'll Actually Use

Most 'AI prompt' roundups were written for marketers or freelancers. These 15 were built for executives who run teams, manage stakeholders, and don't have time to debug a bad output.

Most "AI prompt" roundups were written for marketers or freelancers. These 15 AI prompts for executives were built for a different context: senior professionals who run teams, manage stakeholders, and don't have time to debug a bad output. Every prompt here is paste-ready — fill in the [bracketed fields] and run it.

They're written for Claude but work with ChatGPT, Copilot, and Gemini.

What's covered: 15 prompts across four executive scenarios — meeting prep, strategic communication, decision-making, and presentations. Each prompt includes the situation it solves, a paste-ready prompt, a real output excerpt, and one line on why it works.

What's not covered: Prompts for content creation, social media, or general productivity. This list covers your actual executive work: leadership calls, board prep, stakeholder comms, and high-stakes decisions.

One amplifier before you start: Paste this at the opening of any conversation before running these prompts — it sharpens every output: "You are working with a senior executive. Your role is to challenge, clarify, and surface what's being missed — not to produce polished output for its own sake. Be direct. Never hedge when a clear answer is possible." The Role Calibration component in the Executive AI Toolkit takes this further with role-specific calibrations, but this version works with any prompt below.

Two executives collaborating at a table with laptop and notebook

The right prompt turns 10 minutes of prep into the most prepared person in the room

Scenario 1: Meeting Prep (Prompts 1–5)

The average executive spends 23 hours a week in meetings. Most of that time is recoverable. These five prompts cover the full pre-meeting stack: stakeholder intelligence, hard question prep, your opening statement, post-meeting extraction, and the one prompt you run when you have five minutes and no preparation at all.

For the full 10-minute workflow with prompts for every stage, see How to Prepare for Any Executive Meeting Using AI.

Prompt 1: Stakeholder pressure points

Situation: You're walking into a meeting with someone whose full agenda isn't visible to you.

You are a strategic advisor helping me prepare for an important meeting. I am [your name/role]. I'm meeting with [name], [their title] at [company/org]. The meeting topic is [topic]. Based on their role and the context below, what are their likely priorities, pressure points, and what they most need to get out of this meeting — even if they won't say it directly?

Context: [paste any relevant background — prior exchanges, project status, recent news about their team or org, known tensions]

Give me: (1) their likely #1 priority in this meeting, (2) two things they're worried about that they probably won't raise directly, (3) the outcome they'd consider a win, and (4) one thing I should avoid doing or saying.
"Their likely #1 priority is not the project timeline — it's demonstrating to their VP that they have control of the vendor relationship. They won't raise it directly, but the unresolved escalation from Q3 has put them in a politically uncomfortable position. A win for them looks like leaving this meeting with a concrete commitment they can report upward. Avoid: framing the delay as a process failure on their side — they're already under pressure on this."

Why it works: It forces the AI to reason about subtext — the unstated agenda — not just the stated meeting topic.

Prompt 2: Anticipate the five hardest questions

Situation: You're presenting a recommendation and need to know where it will be attacked before you're in the room.

I am presenting the following recommendation to [audience — e.g., executive leadership team / board / steering committee]: [paste your recommendation in 3–5 sentences].

You are a skeptical senior leader who does not agree with this recommendation. Generate the five hardest questions you would ask — the ones that expose the weakest assumptions, the data gaps, and the political complications. For each question, give me a one-paragraph answer I can use.
"Q: 'What happens to the current vendor contract if we proceed — and have we modelled the exit costs?' Answer: The current contract runs to [date] with a [X]-month termination clause. Early exit would cost approximately [range] depending on which deliverables are accepted as complete. We've built this into the business case as a one-time transition cost, offset within 14 months at the projected run-rate savings."

Why it works: You're not asking for validation. You're stress-testing. This is preparation that actually changes what you say in the room.

Prompt 3: Draft your opening and your ask

Situation: You need a tight, confident opening that frames the room before anyone else does.

I'm opening a [meeting type — e.g., steering committee / QBR / budget review] with [audience]. The goal of the meeting is [goal]. The decision I need from them is [specific decision or commitment].

Write me two things: (1) A 3–5 sentence opening statement that frames the meeting, establishes what we're deciding, and signals I'm in control of the agenda. (2) A single, clear ask I can deliver at the end — one sentence, no hedging.

Tone: direct and senior. This is not a presentation to sit through — it's a working session they're participating in.
"Opening: 'We're here to make a decision on the Q3 vendor consolidation — not to revisit the analysis. I'll take you through the three options we've narrowed to, the recommendation, and what we need from this group to move. We have 45 minutes. I'll need a go/no-go by the end.' Ask: 'I need a decision on Option B today so procurement can execute before end of quarter.'"

Why it works: An executive who walks in with their opening and their ask doesn't get ambushed by the room's agenda.

Prompt 4: Extract action items from meeting notes

Situation: You have 30 minutes of rough notes and need clean, delegatable action items before you leave the room.

Below are notes from a [meeting type] on [date] with [attendees]. Extract all action items in this format: Owner | Action | Deadline | Context (one sentence). If a deadline wasn't stated, write "not specified." If ownership wasn't clear, flag it as "unassigned." Do not fabricate any detail — only extract what's in the notes. If an action item is ambiguous, flag it rather than interpreting it.

Notes: [paste]
"James | Send revised pricing model to Sarah | By end of week | Discussed in context of Q2 CAC payback targets — existing tier structure flagged as misaligned with current margins. | Sarah | Confirm data residency requirements with legal | Not specified | Required before vendor contract moves to procurement. | Unassigned | Follow up on Q3 escalation resolution status | Not specified | Action raised but no owner volunteered — may need a separate thread to resolve."

Why it works: The explicit instruction to flag gaps instead of filling them prevents the AI from inventing plausible-sounding but wrong details. That one instruction is the difference between a usable output and a liability. Note: if you're using Fireflies or Otter for your meetings, you can paste the transcript directly instead of manual notes — the prompt works on either input. (See Best AI Meeting Assistant for Executives for the full tool breakdown.)

Prompt 5: Five-minute prep when you have no time

Situation: You're walking into a meeting in five minutes with no preparation and you need to be the most prepared person in the room.

I have [X] minutes before a meeting with [name(s)], [their roles], about [topic]. Give me: (1) three questions I should ask to control the conversation, (2) one thing I should not say without knowing more, (3) the one outcome that would make this meeting genuinely successful.

Context: [paste whatever you have — a calendar invite, a prior email, a one-line brief, anything]
"Three questions: 'What's changed since the last update?' / 'Is the blocker technical, political, or budget?' / 'What would you need from me today to move this forward?' Don't say: 'I haven't had a chance to review the latest report' — it repositions you as reactive before you've said anything of substance. Success looks like: leaving with one clear next step that you own, not one that's left floating between three people."

Why it works: Five minutes of the right preparation beats 30 minutes of reading the wrong thing. This prompt finds the leverage point in whatever context you give it.

Want 100 prompts like these, organised by executive scenario?

The Executive AI Toolkit includes the full Prompt Library: 100 paste-ready prompts across eight categories — meetings, communication, decision-making, stakeholder management, negotiation, people leadership, personal brand, and more.

Get the Executive AI Toolkit — $67
Executive smiling confidently in a professional setting

AI-generated communication should sound like you wrote it — not like a chatbot did

Scenario 2: Strategic Communication (Prompts 6–10)

These five prompts cover the communications work that eats executive time without showing up on anyone's calendar: the update with bad news embedded in it, the all-hands that sounds like it went through legal, the message you've rewritten four times and still haven't sent.

Prompt 6: Stakeholder update that leads with the situation

Situation: A project has gone sideways and you need to update leadership without burying the lead or losing the room's confidence.

I need to write a [email / memo / Slack message] to [audience — e.g., CFO and COO] about [project or situation]. The situation is: [describe in 3–5 sentences — what happened, current status, what caused it].

Write a stakeholder update that: (1) leads with the situation and impact, not the explanation, (2) gives a clear timeline for resolution, (3) owns what's ownably ours without over-apologising, (4) ends with one specific ask — not a list of next steps. Tone: transparent and controlled. This should read like someone managing a difficult situation, not being managed by it.
"We have a four-week delay on the Phase 2 rollout. Root cause: the data migration dependency we flagged in January was underestimated in scope — by us. Impact: the Q3 go-live moves to 7 November. What we're doing: I've pulled in two additional engineers and am personally reviewing each sprint gate. What I need from you: approval to extend the integration vendor contract by six weeks. I'll send the request through procurement today if I have your go-ahead."

Why it works: The prompt structure forces conclusion-first writing. That's what senior audiences actually want — they don't want to read to the end to find out what happened.

Prompt 7: Calibrate AI to write in your voice

Situation: AI-generated communication sounds generic. This prompt makes it sound like you.

You are a writing assistant calibrated to my communication style. Below are three messages I've written. Study the tone, sentence length, level of directness, how I handle caveats, how I open and close, and any patterns you notice.

[Paste 3 emails or messages you've written — don't edit them, use them as-is]

Now write [the message you need] in the same style. Do not make it more formal, more hedged, or longer than my examples. Match the rhythm exactly.

Real output excerpt — before/after from this prompt, using a senior director's own emails as training input:

Before (uncalibrated AI draft): "I wanted to take a moment to connect regarding the upcoming budget review and share some initial thoughts on how we might structure our approach to the finance conversation. I believe there are a few key considerations worth discussing in advance."

After (calibrated to their voice): "Quick note before Thursday — the finance conversation is going to go sideways if we don't agree on our position beforehand. Can we get 20 minutes this week? I'll come to you."

Why it works: Voice calibration is a one-time setup that changes every piece of writing you produce afterward. The Role Calibration pack in the Executive AI Toolkit takes this further — role-specific system prompts for six executive archetypes, each tuned to how that role actually communicates under pressure.

Prompt 8: Message delivering a difficult decision

Situation: You need to say no, rescind a commitment, or announce something people won't like — without triggering a political incident.

I need to write a message to [recipient or audience] delivering the following decision: [state the decision clearly in one sentence]. They are likely to react with [anticipated emotion or objection — e.g., frustration / feeling blindsided / concern about their team].

Write a message that: (1) is honest about the decision without softening it into ambiguity, (2) briefly explains the reasoning without over-justifying, (3) acknowledges the impact without being patronising, (4) ends with a clear next step. No corporate language. No passive voice.
"I've made the decision to consolidate the two product teams under a single lead. I know this isn't the outcome you were expecting, and I wanted to tell you directly before anything circulates. The reasoning is straightforward: we need one decision-maker on the roadmap for the next six months and the current structure isn't giving us that. I'll have 1:1s with everyone affected this week. If you want to talk before then, send me a message and I'll make time."

Why it works: Most executives either over-explain difficult decisions — which signals doubt — or under-explain them, which signals arrogance. This prompt finds the middle.

Prompt 9: Turn an email thread into a decision brief

Situation: A thread has been circling for days. You need to pull it into something a senior leader can act on in two minutes.

Below is an email thread on [topic]. Turn it into a decision brief with this structure: (1) Situation — what's the issue and why it needs a decision now, (2) Options — no more than three, each with a one-sentence summary of the trade-off, (3) Recommendation — the clearest path forward with a one-line rationale, (4) Decision needed — the specific question being put to the decision-maker.

Use only what's in the thread — stated positions and facts. Do not add opinions. Flag if a clear recommendation can't be drawn from the thread as written.

Thread: [paste]
"Situation: The Q4 budget reallocation has been in discussion since 14 March with no resolution. A decision is needed this week to hit the procurement deadline. Options: (A) Approve the full $240K — highest ROI projection, requires CFO sign-off. (B) Approve $150K with phased release — lower risk, delays by 6 weeks. (C) Defer to Q1 — no current-quarter impact, loses the vendor rate lock. Recommendation: Option B — balances risk and timeline. Decision needed: Does the CFO approve $150K release for procurement by Friday?"

Why it works: Turns a political quagmire into a clean decision document. The instruction to flag unclear consensus prevents the AI from fabricating agreement that doesn't exist.

Prompt 10: All-hands message that doesn't sound scripted

Situation: You need to communicate something significant to your full team — without it reading like a press release.

I'm writing an all-hands message to my team of [X] people about [topic — e.g., org restructure / Q2 results / new strategic direction]. Key points I need to land: [list 3–5 in bullet form]. What I want to avoid: corporate-speak, passive voice, anything that sounds like it went through legal. My team is [describe in one sentence — e.g., experienced and direct / geographically distributed, mixed seniority].

Write a message that sounds like a senior leader talking to people they respect, not a communications department writing to a headcount.
"We're restructuring the APAC team — and I want to explain why before you hear it through the grapevine. The short version: we're moving from geography-based to product-line-based reporting. Some reporting lines change; no roles are being eliminated. I'll be having 1:1s with everyone affected this week. If you have questions before then, send them directly to me — I'll answer what I can."

Why it works: Describing the team as "people they respect" and explicitly naming what to avoid — rather than just saying "write naturally" — changes the register of the output. The AI stops performing communication and starts doing it.

Executive analyzing data on screen

The place where AI earns its keep isn't in writing — it's in thinking

Scenario 3: Decision-Making & Analysis (Prompts 11–13)

The place where AI earns its keep for senior professionals isn't in writing — it's in thinking. These prompts treat the AI as a thinking partner: running pre-mortems, surfacing blind spots, structuring the choices you've been carrying around in your head.

One important note before you use these: AI does not have access to your organisation's data, live financials, or current market information unless you paste it in. If a prompt produces specific numbers or contract references you didn't provide, treat them as illustrative — verify against your actual sources before acting on them. The value here is in the structure and challenge, not in AI-generated facts.

Prompt 11: Pre-mortem a major decision

Situation: You've made a tentative call and want to identify every way it could fail before you commit.

I have made the following decision: [state the decision in 2–3 sentences]. Assume it is 18 months from now and this decision has failed badly. Describe in specific detail: (1) the three most likely failure modes — not generic ones, (2) what the early warning signs were that we ignored, (3) what we should have done differently before committing.

Context: [paste relevant background — team size, budget, timeline, known constraints, stakeholder dynamics]

Do not hedge. Argue that the failure was foreseeable and preventable.
"Failure mode 1: The technology dependency on the third-party API was never validated at scale. The proof of concept used mocked data, not live API calls — we never ran a load test above 500 concurrent users, and the rate limits made the core workflow unusable at production volume. Early warning sign: the vendor's SLA was in the small print of the contract; no one from technical reviewed it before sign-off. What we should have done: required a technical validation milestone before the board commitment, not after."

Why it works: Framing the prompt as "the decision already failed" removes the AI's tendency toward balance. You're asking for attack, not analysis.

Prompt 12: Surface what you're missing

Situation: You've been close to a problem for weeks and you know you have blind spots — you just don't know which ones.

I'm working on [decision or strategy — describe in 3–5 sentences]. I've been close to this for [X weeks/months] and I'm aware I have blind spots. Here's my current thinking: [paste your analysis or recommendation].

You are a smart external advisor who has not been involved in this. Your job is not to validate my thinking. Tell me: (1) the assumption I'm most confident about that is most likely to be wrong, (2) the stakeholder or perspective I've probably underweighted, (3) the question I haven't asked that I should have asked first.

Be direct. This is not a peer review — it's a challenge.
"The assumption you're most confident about that's most likely to be wrong: that the operations team will adopt the new system without significant resistance. You've modelled for technical integration; you haven't modelled for behavioural change at the team lead level — the people who will be measured on metrics your new process doesn't capture. Underweighted perspective: middle management in the affected region; their buy-in determines whether this works in practice or just on paper. Question you haven't asked: 'What does success look like for someone who doesn't want this to succeed?'"

Why it works: "This is not a peer review — it's a challenge" is a permissions grant. Without it, the AI hedges. With it, it attacks. One phrase changes the entire output.

Prompt 13: Compare two options with a structured trade-off

Situation: You have two paths and need a clean framework to present the choice upward — or to make the decision yourself.

I'm deciding between the following options for [situation]:

Option A: [describe in 3–5 sentences]
Option B: [describe in 3–5 sentences]

Evaluate them across: (1) speed to value, (2) risk level and type, (3) resource requirement, (4) reversibility — how hard is it to undo if we're wrong, (5) strategic fit with [your stated priority or constraint].

Give me a comparison table, then a one-paragraph recommendation with a clear rationale. If what I've given you isn't sufficient for a confident recommendation, tell me what you'd need to know before you could give one.
"Recommendation: Option A has lower resource requirements and faster time to value, but it's not reversible — if the market assumption is wrong, you're committed to a path that's expensive to exit. Option B takes 18 weeks versus 12, but preserves optionality. Given that you've described the board as risk-averse and the market conditions as uncertain, Option B is the stronger recommendation. The six-week delay is less consequential than the asymmetry in reversibility."

Why it works: The final instruction — "tell me what you'd need to know" — prevents false confidence. If your brief is thin, you find out before you act on a bad recommendation.

Conference room presentation setup

Get the argument right before you open the slide tool

Scenario 4: Presentations & Board Prep (Prompts 14–15)

These two prompts cover the presentation work that matters most: the narrative before the slides, and the questions you haven't prepared for.

For the full 30-minute workflow — brief to finished deck — see How to Build an Executive Presentation with AI in 30 Minutes.

Prompt 14: Build a narrative spine before you open a slide tool

Situation: You have raw material for a presentation but no structure. You need the argument sequenced before anyone sees a deck.

I need to build the narrative for a [type of presentation — e.g., board update / QBR / investment case]. My audience is [audience]. The decision I'm asking for is [decision].

Here is the raw information I have: [paste bullet points, data, context — it doesn't need to be structured]

Build me a narrative spine using this structure: (1) Situation — what is true right now, (2) Complication — why the status quo is no longer acceptable, (3) Recommendation — what I'm proposing and why, (4) Risks and objections — the three strongest arguments against my recommendation and a one-line response to each, (5) The ask — what I need from this room today.

Keep each section to 2–4 sentences. This is not the deck — it's the thinking behind the deck.
"Situation: We're delivering against plan on revenue but 12% behind on margin. Complication: At current trajectory, we hit the threshold that triggers a covenant review in Q3 — six months earlier than projected. Recommendation: Accelerate the cost consolidation programme by one quarter, prioritising the three initiatives with highest margin impact and lowest execution risk. Risks: (1) Execution speed — we've completed the design phase, so the risk is delivery, not design. (2) Team morale — framing this as an investment in financial resilience, not a cost-cut, is accurate and matters. (3) Client visibility — none of the affected programmes are client-facing. The ask: Approval today to begin the accelerated timeline."

Why it works: Situation-Complication-Recommendation is the structure that senior audiences can follow at speed because it answers the question they're always silently asking: "so what, and what do I need to do about it?" Getting this right before you open PowerPoint means your slides reinforce an argument instead of trying to become one.

Prompt 15: Generate the ten questions you haven't prepared for

Situation: You've prepared for the questions you expect. This prompt generates the ones you haven't thought of.

I'm presenting the following recommendation to [audience]: [paste your narrative spine or recommendation in 150–300 words].

You are the most skeptical person in that room. You do not support this recommendation. Generate ten questions — the ones that expose the weakest assumptions, the political complications, and the data I haven't shown. Rank them by how dangerous they are to my position if I can't answer them well.

For the top three, give me a draft answer I can use.
"Most dangerous: 'You're showing margin pressure in Q3, but your growth forecast assumes expansion into two new markets — how do you reconcile those?' Draft answer: 'The market expansion revenue doesn't hit P&L until Q4 at the earliest. The Q3 margin work is what funds the expansion — we're not running these timelines in parallel, we're sequencing them. The slide on page 8 shows the waterfall.' Second most dangerous: 'Who else has seen this analysis? Have Finance signed off on the margin figure?' Draft answer: 'The CFO's office reviewed the model last Thursday. The margin figure is their number, not ours.'"

Why it works: Ranking by danger forces prioritisation. The top three with draft answers means you walk in having already rehearsed the hardest questions. Even if you only prepare five of the ten, you'll be more prepared than anyone else in that room. This is the prompt that turns preparation into performance.

For a deeper board-specific prep workflow — including how to build your narrative, stress-test with AI, and prep live Q&A responses — see the upcoming article on AI for Board Prep.

The Executive AI Toolkit includes 100 prompts in the same format — across all 8 executive categories.

This article covers four: meetings, communication, decision-making, and presentations. The full Prompt Library adds 85 more covering people management, stakeholder management, negotiation, and personal brand — each built to the same standard: paste-ready, real scenarios, output that doesn't need to be rewritten before it's used.

$67. One purchase. No subscription.

Get the Executive AI Toolkit — $67

Where to Go Next

If the meeting prep prompts were useful: How to Prepare for Any Executive Meeting Using AI walks through the full 10-minute workflow — from context dump to opening statement — with prompts for every stage.

If the presentation prompts were useful: How to Build an Executive Presentation with AI in 30 Minutes covers the full process from brief to finished deck, including a worked example with a VP of Operations presenting cost reduction to a board.

For the tools that run best with these prompts: The Executive AI Stack in 2026 — what to use, when, and in what combination.

Free guide + weekly newsletter

Get Started with AI in One Day — Free

Subscribe and get our free 15-page starter guide instantly. Then weekly AI workflows, honest tool takes, and strategies for senior professionals. No fluff. Unsubscribe any time.

No spam. Unsubscribe anytime.

Keep reading

Guide10 min read

How to Prepare for Any Executive Meeting Using AI (a 10-Minute Workflow)

Real preparation — the kind that changes how a meeting goes — used to take hours. This workflow collapses it into 10 minutes using Claude.

Mar 28, 2026Read more →
Guide6 min read

The 10 AI Prompts Every Executive Should Know

From briefing synthesis to stakeholder communication — prompts that actually save hours every week.

Mar 5, 2026Read more →
Guide9 min read

How to Build an Executive Presentation with AI in 30 Minutes

Most AI presentation tools optimise for speed and aesthetics. This workflow builds narrative structure first — so your deck moves a decision, not just fills a room.

Mar 31, 2026Read more →
Guide15 min read

The Executive's Complete Guide to AI in 2026: What to Use, How to Start, What to Skip

Most executives don't have a time problem. They have a preparation problem. This guide covers what AI can actually do for you, which tools are worth paying for, which workflows to automate first, and why most executives quit before they see results.

Apr 9, 2026Read more →