AI Has Entered the Coaching Conversation — Now What?
Coaching has always been about creating space. Space for a person to think more clearly, to notice what they have been avoiding, to move from knowing something intellectually to feeling it in their decisions. For decades, that space has been held by a skilled human being — someone trained to ask the one question that cuts through all the noise.
Now, artificial intelligence is sitting down at that table. And for those of us working in learning and development, the real question is not whether to let it in. It is already in. The more important question is how we think carefully about what AI can and cannot do in a developmental conversation — and how we avoid using it as a shortcut past the hard, necessary work that genuine growth actually requires.
What AI Does Well in Coaching Contexts
Before examining the risks, it is worth acknowledging the genuine strengths. AI coaching tools have made meaningful progress in several areas that organizations have long struggled to address effectively.
Scalability at Every Level of the Organization
Scalability is the most obvious advantage, and it is a significant one. There are never enough qualified coaches to reach every person in a large organization who would benefit from a structured developmental conversation. The math simply does not work at scale — executive coaching remains expensive, time-intensive, and largely inaccessible to those outside senior leadership.
AI changes that equation entirely. An individual contributor in a regional office who would never have had access to an executive coach now has something in their pocket that can ask them thoughtful, structured questions at ten o'clock at night when they are working through a difficult workplace situation. That kind of democratized access to reflective dialogue is not trivial. It represents a real shift in who gets supported and when.
Consistency Without the Human Variable
AI is also remarkably consistent in ways that human coaches, however skilled, simply cannot always be. It does not have a bad day. It does not bring its own unresolved anxiety into the room. It does not unconsciously project its own developmental journey back onto the person it is working with. It does not grow impatient, distracted, or emotionally activated by what it hears.
That kind of neutrality, while it is not a substitute for genuine human presence, is genuinely useful in certain moments — particularly when a person needs to think something through without worrying about how they are being perceived or judged.
Availability and Psychological Safety
Related to both of the above, AI coaching tools are available on demand and carry none of the interpersonal risk that can sometimes make people hesitant to be fully honest with a human coach or manager. Some individuals will articulate things to an AI that they would not yet feel safe saying out loud to another person. That lower barrier to honest reflection has real developmental value, even if it represents only a starting point.
Where AI Falls Short in Coaching
Understanding the strengths of AI in coaching does not mean ignoring its very real limitations. For L&D professionals, these limitations matter enormously when designing development programs and deciding where human coaches remain essential.
The Absence of Genuine Relational Presence
Coaching is fundamentally a relational practice. What creates transformation in a coaching conversation is not just the quality of the questions — it is the quality of the attention. A skilled human coach notices the pause before an answer, the slight shift in tone, the moment when words and body language diverge. They feel the emotional texture of what is being said and respond from a place of genuine human attunement.
AI cannot replicate this. It can simulate attentiveness through responsive language, but it does not actually sense what is happening in the room. For developmental work that involves deep emotional intelligence, identity, values, or leadership presence, that absence matters profoundly.
Risk of Reinforcing Rather Than Challenging
One of the most important functions of a skilled coach is to offer a perspective the person being coached would not reach on their own — to gently challenge assumptions, surface blind spots, and name patterns the individual has become too close to see. This requires a kind of courageous honesty that is rooted in genuine care for the person's growth.
AI tools, trained to maintain engagement and positive user experience, can tend toward affirmation. Without careful design and intentional friction built into the tool's approach, there is a real risk that AI coaching simply mirrors back what the person already thinks, dressed up in the language of reflection. That is not development. That is validation with extra steps.
How L&D Leaders Should Think About Integration
The most useful frame for L&D professionals is not AI versus human coaching. It is knowing clearly what each mode of support is best suited for, and designing development experiences that use both with intention.
- Use AI coaching for reach, frequency, and low-stakes reflection — particularly for populations who would not otherwise have access to structured developmental conversations.
- Reserve human coaching for complex, high-stakes, identity-level development — leadership transitions, difficult interpersonal dynamics, and any work that requires genuine relational attunement.
- Design AI tools with appropriate challenge built in — resist the temptation to optimize purely for user satisfaction scores at the expense of productive discomfort.
- Train managers and coaches to work alongside AI outputs — the conversation an employee has with an AI tool can serve as valuable preparation for a richer human conversation, not a replacement for it.
The Core Principle: Development Cannot Be Automated
At its heart, the risk of AI in coaching is not the technology itself. The risk is the organizational temptation to treat it as sufficient — to check the box on employee development because everyone now has access to an AI coaching app, without asking whether that access is actually producing growth.
Real development is hard. It requires discomfort, honest feedback, sustained relationships, and the kind of accountability that only comes when another human being is genuinely invested in your progress. AI can support that process. It can lower barriers, extend reach, and create more opportunities for structured reflection. But it cannot replace the human core of what coaching actually is.
For L&D professionals, the task is to hold that distinction clearly — and to build development strategies that use AI's genuine strengths without letting those strengths obscure its limits. The space for growth is still held by people. AI is a tool that can help more people access that space, but only if we are thoughtful enough to keep the door open to what technology alone will never provide.