Insights · The Growth Coach HK
AI Can Tell a Leader What to Do. It Can't Tell Them Who to Be.
21 October 2025
Introduction
There is a version of AI-assisted leadership that I find genuinely useful and a version I find genuinely concerning.
The useful version: AI helps leaders think more clearly, prepare more thoroughly, communicate more effectively, and make better-informed decisions. It is a thinking tool that augments the leader's capability.
The concerning version: leaders use AI to generate not just their communications and analysis but their leadership positions — what to prioritize, how to respond to a difficult team dynamic, what stance to take on a cultural question. AI as the source of the leadership, not just the support for it.
The difference matters because leadership is not primarily a knowledge problem.
Main Insight
Leadership is fundamentally a character problem. Not in a moral sense, but in the sense of: who are you, and how does who you are show up in the way you lead?
The decisions that most define leadership quality are the ones that involve genuine trade-offs — where the right answer is not obvious, where different stakeholders have legitimate competing interests, where the leader's values and priorities are the determining variable. In those moments, what matters is not what AI says a leader should do. It is what the leader actually believes, what they are willing to stand for, and how they behave when the answer is not provided for them.
AI can generate a leadership response to almost any situation. That response may even be good advice. But following AI-generated leadership guidance is not the same as leading. It is executing someone else's — or something else's — judgment.
Common Mistakes
One mistake is using AI to generate positions on genuinely contested leadership questions — questions about culture, people, values, and direction — rather than using AI to sharpen thinking about positions already arrived at through the leader's own judgment.
Another mistake is treating the discomfort of hard leadership calls as a problem to be solved with AI rather than as a signal that genuine judgment is required. The discomfort usually means the decision matters and the trade-offs are real. That is where the leader needs to show up, not step back.
A third mistake is not distinguishing between AI assistance in preparation and AI assistance as a substitute for the decision. Both involve AI. One builds on the leader's thinking. The other replaces it.
Framework: Where AI Belongs in Leadership Development
Preparation and reflection — Use AI to prepare for hard decisions, surface considerations, stress-test positions. Use AI to reflect after significant decisions — what happened, what to adjust, what to do differently next time.
Communication — Use AI to communicate decisions that the leader has genuinely reached through their own judgment — more clearly, more quickly, more comprehensively.
Learning inputs — Use AI to access perspectives, research, and frameworks that inform the leader's development. These are inputs to the leader's thinking, not substitutes for it.
Not as a source of positions — The leader's stance on a genuinely contested question — about their culture, their people, their direction — must come from the leader's own judgment, values, and accountability. AI can inform that thinking. It cannot own it.
Practical Lessons
The leaders I have seen grow most significantly over the past year are the ones who use AI as a sophisticated preparation tool — and who are then fully present and personally accountable in the moments that matter.
They do not outsource the hard calls. They use AI to be better prepared for them.
And the development that comes from making those calls — from genuinely owning the difficult moments — compounds in ways that AI-assisted positions do not. Judgment improves through consequences, feedback, and the accumulated weight of real decisions made by a real person with real stakes.
Leaders who consistently outsource the hard part of decision-making to AI may produce good outputs. They will develop less as leaders.
Conclusion
AI can tell a leader what to do in almost any situation.
It cannot develop their judgment, their integrity, their clarity of values, or their self-knowledge. Those develop through experience, reflection, honest feedback, and the accumulated weight of genuinely owning the hard calls.
The leader who builds those things — with AI as a tool in the process — will lead differently from the leader who uses AI as a substitute for that development. And over time, the difference will be visible.
FAQs
What can AI not do for leadership development?
AI cannot develop the judgment, integrity, self-awareness, and clarity of values that constitute leadership character. Those develop through experience, reflection, honest feedback, and the accumulated weight of genuinely owning hard decisions. AI can support preparation and reflection around those experiences — but not replace the experiences themselves.
Is it wrong to use AI to help with leadership decisions?
Using AI to prepare for decisions, surface considerations, and stress-test reasoning is legitimate and valuable. Using AI as the source of your position on genuinely contested leadership questions — where your values and judgment are the determining variable — means executing someone else's judgment rather than leading.
How does leadership character develop in an AI era?
The same way it has always developed: through the accumulated weight of making genuine decisions, experiencing the consequences, reflecting honestly on what happened, and adjusting. AI can accelerate the preparation and reflection around those experiences. It cannot replace the experience of genuinely owning the hard calls.
This site requires JavaScript for the full experience. Email us or enable JavaScript to continue.