I’ve ghostwritten 54+ books for executives, and the communication problems I see in manuscripts are the same ones that plague their companies. The issue is rarely word choice or grammar. It’s that executives don’t know what they want to say, and no AI tool fixes that. Leaders get sold on the idea that the words are the hard part, and it’s a con. The hard part is deciding what you believe and saying it to people who depend on you, and that’s exactly the part these articles tell you to hand off.
What is the real communication problem for leaders?
Most executive communication fails for one of three reasons: the leader hasn’t clarified their own thinking, they’re avoiding the hard part of the message, or they’re performing authority instead of exercising it.
Context is everything in executive communication, and AI has none of yours.Share on X
AI makes all three worse.
When a CEO doesn’t know what they think about a layoff, asking ChatGPT to “draft an internal memo addressing layoffs with sensitivity and clarity” produces something that sounds thoughtful without containing any actual thought. The memo lands in inboxes, employees read it, and they feel exactly what they felt before: that leadership is hiding behind polished language instead of being straight with them.
Employees can smell hollow corporate communication. They’ve been reading it their entire careers. I write in other people’s voices for a living, and a ghostwritten message only works when the person whose name is on it has decided what they mean. Running an undecided message through AI gets the hollow version into inboxes faster, and every hollow memo spends a little of the trust you’ll need the next time something goes wrong.
Where Does AI Help Leadership Communication?
AI is useful for executives in the same way it’s useful for writers: as a tool that handles mechanical tasks so you can focus on the work that requires judgment.
Summarizing meeting notes so you can write a focused follow-up instead of reconstructing the conversation from memory. Organizing survey responses into themes so you can see patterns before crafting your response. Pressure-testing a draft by asking what objections employees might raise, then deciding which ones to address. Catching inconsistencies between what you said in Monday’s email and what you’re about to say in Friday’s town hall.
None of this is authorship. It’s preparation. The executive still has to decide what to say and stand behind it.
Where Does AI Fail Leadership Communication?
AI can’t tell you what you believe. It can’t read the political dynamics of your leadership team. It doesn’t know your VP of Engineering will read “we’re exploring restructuring” as a direct threat to her department, or that your CFO will hear “innovative approaches to efficiency” as code for the budget cuts he’s been fighting against for six months. Context is everything in executive communication, and AI has none of yours.
AI also can’t do the thing that matters most in a crisis: be honest when honesty is uncomfortable. The best crisis communication I’ve seen from executives was blunt, specific, and clearly written by a human who was willing to own the situation. The worst was smooth, empathetic-sounding, and obviously generated by someone who wanted to appear caring without committing to anything.
| Task | Who should handle it | Why |
|---|---|---|
| Summarizing meeting notes | AI | Frees you to write a focused follow-up |
| Organizing survey responses into themes | AI | Shows patterns before you craft a response |
| Pressure-testing a draft for objections | AI, then you decide | You choose which objections to address |
| Catching inconsistencies between messages | AI | Flags conflicts between Monday’s email and Friday’s town hall |
| Deciding what you believe | You | No tool fixes unclear thinking; it decorates it |
| Crisis communication | You | It must be honest, blunt, and owned by a human |
| Performance reviews | You | Specific feedback needs someone who was in the room |
| Explaining layoffs | You | People deserve the reason in your own words |
The Performance Review Problem
Using AI to draft performance reviews is especially risky. A review is one of the few moments where an employee gets direct, personal feedback from their leader. If that feedback reads like it was generated from a template, the employee knows it, and whatever trust existed between them erodes.
The sentence you’re most tempted to hand off to AI is almost always the sentence that matters most.
“Draft feedback for an employee who met goals but needs to improve teamwork” produces generic advice about collaboration and communication skills. What the employee needs to hear is that they talked over three colleagues in the Q3 planning meeting and that two people have asked not to be assigned to their projects. That specificity requires a human who was in the room and cares enough to say the hard thing.
If a leader can’t be bothered to write that paragraph personally, I’d question whether they should be giving the review at all. People remember for years how their boss told them the hard thing, and a template tells them exactly how much they mattered.
What should executives do about AI in leadership development?
From the podcast
Rob Fenstermaker draws the line between the two things people mistake for each other, on Leaders and Their Stories:
A warrior is never somebody who’s out to fight those who are in front of him. A warrior is a man who takes his stand to protect those who are behind him.The leaders who use these tools well are pointed the same direction: at the people behind them.
Clarify your thinking before you write anything. If you can’t explain your message in two sentences to a trusted colleague, you’re not ready to communicate it to the company. No tool fixes unclear thinking. It just decorates it.
Write the hard part yourself. The sentence you’re most tempted to hand off to AI is almost always the sentence that matters most.
If you’re announcing layoffs, the employees being laid off deserve to hear the reason in your own words, and they’ll know the difference. I keep the rest of this together in my Leadership Hub.
Use AI for preparation and mechanics, and keep voice and judgment for yourself. Let it organize your notes, check your facts, and flag inconsistencies. Then write the message yourself, or work with a human who can capture your voice and push back when your thinking is fuzzy.
Accept that good communication takes time. The executives who communicate well are more willing to tolerate discomfort, think through implications and say what they mean. The ones who reach for whatever sounds safe are protecting themselves at their employees’ expense, a poor way to lead.
AI handles the preparation; the leader decides what to say and writes the hard part in their own words.
My take
Hand AI your notes and fact checks, and keep the hard sentences for yourself. A performance review that reads like a template tells the employee exactly how much they mattered. If a leader can’t be bothered to write that paragraph, I’d question whether they should give the review at all. Layoffs work the same way. People deserve the reason in your own words. And if you can’t explain the message in two sentences, no tool will fix it. It’ll only decorate it.
The Bigger Picture
The rush to automate executive communication reflects a misunderstanding of what communication is for. Its job is building trust, not producing polished text, and trust comes from specificity, honesty, and the willingness to be accountable for what you say.
AI can help you get there faster by handling the tasks that don’t require your judgment. It can’t replace the judgment itself. Leaders who understand this distinction will communicate better than leaders who hand their hardest conversations to a prompt.
Handing your hardest conversations to a prompt is a leadership failure, and calling it efficiency doesn’t change that. Leaders owe their people their own judgment in their own words. A leader who won’t give them that shouldn’t be surprised when they stop believing anything that comes out of the corner office.
If your communication challenges go deeper than word choice, if you’re trying to articulate a vision, document a legacy, or tell a story that matters, that’s a conversation worth having.
Leadership communication is where outsourced judgment shows first. The Death of Thinking covers what handing the thinking to tools costs and why you feel the loss only when you need what’s gone.
