When people ask whether I am worried about artificial intelligence (AI) replacing coaches, my answer is not defensive. It is considered. There are parts of coaching that I believe AI should replace.
Not the relational depth. Not the complex human work. But the transactional, formulaic layers that have dominated both organizational coaching and much of the wider coaching industry for years.
This is not only about corporate performance conversations. It is also about the pervasive reliance on models such as GROW and the rigid interpretation of non-directive coaching as the gold standard. Structured goal setting. Predictable questioning sequences. Conversations that orbit around “What do you want?” and “What will you do next?” without ever asking why the same patterns keep repeating.
If that is the level at which we are operating, AI will outperform many coaches, and rightly so.
The Essence of Coaching
Earlier in my career I worked in business transformation. We did not preserve tasks simply because people were used to doing them. We automated repeatable processes. We offshored what could be systematized. Then we upskilled the workforce into more complex, judgment-based roles. We called them knowledge workers. Their value lay in interpretation, contextual awareness, and decision-making under uncertainty.
Coaching now stands in a similar moment.
If a client needs a structured space to clarify a goal, identify options, and commit to action, an AI tool can facilitate that with impressive consistency. It can prompt reflection, generate alternatives, and track accountability. For straightforward performance progression or early stage thinking, that may be sufficient.
The difficulty arises when we pretend that this is the essence of coaching.
For decades, the industry has elevated non-directive questioning as a moral high ground. The coach asks, the client answers. The coach withholds opinion. The client finds their own solution. On paper, it sounds empowering. In practice, I have spoken to hundreds of clients who describe simply telling the coach what they believed was expected. They articulated insight fluently. They left with action points. They returned weeks later with the same emotional reactions and the same behavioral loops.
Because the pattern was never addressed.
GROW-style coaching, particularly when delivered mechanistically, works at the level of cognition. It assumes that once clarity is achieved, change will follow. It rarely interrogates the physiological state of the person in front of you. It does not examine the origin of the belief that drives the behavior. It does not consider that the client may say, “I don’t know,” not because they lack intelligence but because their nervous system is overwhelmed.
AI is perfectly capable of facilitating cognitive clarity. It can ask, “What is your goal?” and “What are your options?” It can challenge inconsistencies in reasoning. It can provide structure and repetition at scale. If coaching remains confined to that terrain, technology will do it faster and cheaper.
The question for the profession is whether that terrain is sufficient.
Rewriting the Rules of Coaching
Over the past few years, I have been deliberately rewriting the rules of coaching in my own work and in the way we train coaches. I blend solution-focused therapeutic thinking with both non-directive and directive coaching styles. I meet the client where they are, not where a model insists they should be.
If someone is clear, regulated, and resourced, non-directive exploration can be powerful. If someone is stuck in a defensive pattern they cannot see, a more directive intervention may be necessary. If a belief was formed in a moment of stress years ago and is still driving behavior, we address that. If the nervous system is dysregulated, we work there first, because insight without regulation rarely leads to sustained change.
This is not therapy in disguise, nor is it advice giving dressed up as coaching. It is an integrated, psychologically informed approach that recognizes that human beings are shaped by past experiences, present pressures, and future aspirations all at once. We cannot compartmentalize them and expect coherence.
AI cannot yet sit in that complexity.
It cannot sense the subtle tightening of someone’s jaw when they speak about a colleague. It cannot recognize when humor is masking shame. It cannot co-regulate in real time when a leader edges toward overwhelm. It cannot decide, with ethical discernment, when to challenge firmly and when to slow the pace.
In organizations, AI can and should support structured performance coaching at scale. It can help managers hold consistent conversations, analyze development data, and prompt reflection between sessions. That frees capacity. It allows learning and development budgets to be used more strategically.
But the human coaches within those systems must then raise their capability. They need to understand nervous system responses, identity formation, relational dynamics, and power. They need to be able to move beyond scripts and hold a room where tension, ambition, and fear coexist.
Outside organizations, the same principle applies. If independent coaches are offering little more than structured goal questioning, clients inevitably will compare that experience to what technology can provide. The value of a human coach must lie in depth, discernment, and integration.
Deepening the Profession
Saying, “Yes,” to AI is not about diminishing coaching. It is about clarifying it.
We have an opportunity to allow technology to absorb the transactional layers and to elevate the human craft. To move beyond rigid allegiance to any single model. To blend approaches intelligently. To work with the whole person rather than just their quarterly objectives.
In transformation work, automation did not eliminate human contribution. It demanded more of it. It required people to think differently, to operate at a higher level of skill, and to take responsibility for complexity rather than routine.
Coaching is at that same threshold.
If we release the formulaic, GROW-driven, purely cognitive layers to AI, we create space to deepen our profession. We stop defending tradition and start building capability. We become more human, not less.
That is why I say, “Yes,” to AI in coaching.
