Coaching Skills in the Age of AI!

A tool can ask the questions, but only a leader can let the answer change the plan.
Last week I told you about my own story with a leader who sat down and kindly asked, "What's getting in the way?" I answered honestly, and when the conversation ended, the plan stayed exactly the same. I did not share that story to point a finger at anyone. I have been on both sides of this situation, and I suspect a good many of us have been.
This week we are going to explore this coaching challenge with great empathy for anyone who recognizes themselves in it. When we use coaching skills only to advance the plan, they become a way to collect information and compliance while giving very little back. A leader can use every technique correctly, with open questions, attentive listening and a warm tone, and still leave someone feeling like a source of data. Most people experience this but don't always register or process it, so it begins to build into a growing sense of not mattering. They answered honestly, and nothing moved in what was expected, and they quietly adjusted how much honesty and effort they offered the next time.
Technology, and specifically AI, amplifies this conversation. An AI tool can already ask thoughtful, well-sequenced questions, and it will ask them more smoothly every year. If question-asking is the main skill a leader brings to a conversation, then it is the part of the role that is easiest to hand off.
What a tool cannot do is be changed by what it hears. It cannot decide that a deadline will move, that a project will wait, or that an expectation will shrink to something a person can actually carry.
That takes a human being with authority, empathy, and the willingness to use it, and the willingness is the piece most of us underestimate.
The conditions around us make this more pressing. The DDI Global Leadership Forecast 2025 found that leaders who build trust are 2.8 times less likely to meet resistance to AI, and that first-level leaders are about three times more worried about AI's impact than senior leaders are. The Leapsome 2025 HR Insights survey, which gathered responses from 1,000 HR decision-makers across the US, UK, Germany and the Netherlands, found that 73 percent say their organization plans workforce restructuring to integrate AI. Those same HR leaders report that employees resent AI being added to already packed workloads, which is HR's perception of how employees feel and not the employees' own account, though it matches what many of us hear in the hallway.
It is important to be careful with these numbers. They describe conditions: change is arriving fast, the leaders closest to the work are the most anxious, and trust appears to shape how change is received around AI use. They show association, and they do not prove that coaching skills alone fix any of it. However, it is an important reminder that the people in front of us are carrying a great deal, and that the moment someone answers a hard question honestly is a moment that matters.
None of this means the plan does not matter.
Leaders and managers feel real pressure to advance the plan, and they feel an equally real pull to care for the people doing the work. Both are legitimate, and anyone who has held both at once knows how heavy that can be. The truth is, people can usually tell when you are coaching to advance the plan or you are coaching to support the people. When we do the latter first, the former is received better.
It's not an either-or discussion, but rather a first-then discussion.
So here is a small experiment I would invite you to try, with curiosity and not with judgment. After your next one-on-one, name one thing that changed because of what you heard. It might be the plan itself, or a workload, or a deadline, or simply how you showed up in the room the following day. If you can identify something, you have let the conversation reach you, and the person across from you will probably feel that even if they never mention it. If nothing changed, it may have been information gathering, which is something most of us have done with good intentions and a full calendar.
This is part of why I built the C.L.E.A.R.™ Commitment Process the way I did. The Clear Conversation Model™ is designed to make room for the real answer to be said out loud, which is often the hardest part, since people need some evidence that honesty will be safe and worthwhile before they offer it. The Commitment Map™ then records what each person agrees to, including what the leader will do differently, so that the conversation has consequences on both sides. A leader's own commitment belongs on that page, because an answer that changes nothing for the person who asked is just a record of what was said.
I find some relief in all of this, even though it asks something of us.
A machine may soon ask better questions than most of us do, and that frees us to focus on the part that was always ours, which is to hear someone, let it matter, and act on it where we have the authority to do so.
People can tell when that is happening, and they tend to respond in kind, usually with more candor, more trust, and more willingness to move toward the plan when it is time to move.
Next week I want to look at what it costs when conversations leave people feeling like sources of data, and I will put some numbers to it. For now, I will leave you with a question to carry into your week.
In your last one-on-one, what did you change because of something you heard?




Comments