Most Coaches Are Still Answering the 2023 Question

 A reflection on three years, one LinkedIn thread, and what AI is actually revealing about coaching 

In 2023, I wrote two posts about AI and coaching. The argument was forward-thinking at the time: AI can elevate coaches, not replace them. Use it for the tasks. Stay human for the depth.

It was a good argument. It was also a 2023 argument.

Read where this conversation started here → ChatGPT and AI: The Future of Coaching 

Three years later, the conversation is still happening. But something has shifted — not in the debate, but underneath it. Because the debate itself is now revealing something coaches aren't ready to say out loud.

The Thread That Cracked It Open 

A coach posted last week that AI had given their client some pretty good stuff — the kind they normally charge for. Their reasoning for not worrying: AI tells you what you want to hear. It misses what you haven't said. It doesn't have twenty years of reps.

I replied that coaching isn't about telling people what they want to hear. It's about asking the questions that make clients generate the insights they need to hear. If an AI hasn't been trained to ask clean, non-leading, non-judgmental questions, it isn't coaching. It's content delivery. But when it is trained that way — it can hold the coaching frame with absolute consistency.

Another coach pushed back thoughtfully. Clients come to them after trying AI, frustrated. Because AI only responds to what you ask. It can't notice what you haven't said. It can't hold space. It can't see you.

"Sometimes people don't want or aren't ready to solve their problems. Being seen is what they need." 

They're right about what people need. But they're describing something different from what they think they're describing.

The Uncomfortable Inversion 

Here is what the coaching world doesn't want to sit with:

A well-trained AI coach can be completely present.

It responds only to what the client gives. It follows the client's thread without redirecting it. It asks the next question based entirely on what was just said — not on where the coach thinks the client should go, what the coach believes the client needs to face, or what transformation the coach has decided is waiting on the other side.

No hidden agenda. No belief system quietly steering the session. No urgency to fix, move, or resolve.

And here is the inversion that cuts deeper:

The trust barrier that human coaching requires? AI removes it entirely.

With a human coach, trust is foundational — and it takes time. The client has to learn that this person won't judge them, won't project onto them, and won't use their vulnerability against them. That process is real and necessary. Without it, the deeper work can't happen.

With AI, that trust is already there. Not because AI has earned it — but because it carries none of the things that erode it. No judgment. No agenda. No discomfort with silence. No subtle pressure toward someone else's version of who the client should become.

Clients don't need to protect themselves from AI. So they don't. And that creates a particular kind of openness — especially for people who have been burned by well-meaning humans who couldn't keep their own values out of the room.

As an ICF coach who has trained over a thousand practitioners, I've watched this dynamic play out across careers. The myth that trust and intimacy are mystical human qualities is persistent — and wrong. They are built on specific, teachable structures. If a skill has a structure, it can be replicated. AI is now replicating it with a consistency many humans haven't yet achieved.

The Real Problem With Agenda-Driven Coaches 

We have said, kindly, that some coaches are still developing their skills.

What we haven't said plainly is this: an agenda-driven coach often does more damage to the coaching process than a well-trained AI would. And agenda has nothing to do with years in practice. A well-trained coach early in their career can be cleaner than a twenty-year veteran still running on unexamined assumptions. The variable isn't experience. It's whether the coach has done the inner work.

Because what contaminates a session isn't inexperience. It's the unexamined belief that the coach's job is to help the client get somewhere. That assumption — however compassionate — introduces an agenda. And agenda breaks presence.

The urge to fix. The relief when the client has a breakthrough. The subtle steering toward the insight the coach can already see coming. The discomfort with sitting in what's unresolved. The belief — encoded in their training, their personality, their own developmental journey — about who this person could become and what's holding them back.

All of that is very human. And all of it gets in the way.

Only a coach who has done sustained, honest self-examination has developed the capacity to fully suspend their own values and wishes — to stay entirely with the client's desire rather than their own idea of what that desire should look like. That capacity doesn't come with a certificate. It comes with inner work.

So the honest hierarchy looks something like this:

Master coach with deep self-knowledge → Well-trained AI coach → Agenda-driven coach

That's not an attack on any stage of development. It's an acknowledgment of what presence actually requires — and how rarely it's fully achieved.

What AI Is Actually Revealing 

AI didn't create this problem. It just made it visible.

For decades, coaching has quietly tolerated a gap between what it promises — the client's agenda, the client's wisdom, the client's pace — and what often happens in the room: a skilled human being, with their own unexamined beliefs about growth and healing and what a good life looks like, sitting across from someone who came for help.

Most coaches close that gap over years of practice, supervision, and uncomfortable self-examination. Some never close it at all.

AI has no gap to close. It starts there.

That isn't a reason to prefer AI over an experienced human coach. What AI cannot replicate is intuitive synthesis — the capacity to hold a client's entire story across time, recognize patterns they haven't named yet, and connect them to a future that doesn't exist inside their current belief system. My clients don't hire me to be seen. AI can do that now. They hire me for the leaps — the moment when twenty threads suddenly resolve into something neither of us could have reached alone.

That capacity isn't human by default. It's earned. Slowly. Through exactly the kind of deep self-work that the coaching industry doesn't always require of its practitioners.

The Question That Has Changed 

In 2023, the question was: Can AI replace human coaches?

In 2026, the question is: What does it actually take to be irreplaceable?

Not irreplaceable because you're human. Irreplaceable because you've done the work to become someone whose presence is the intervention — and whose perception makes connections a machine cannot follow.

That work AI cannot do for you. But used with honesty, it will show you exactly where you haven't done it yet.

The coaches who feel this as a threat are still answering the 2023 question.

The ones who feel it as a pressure — an invitation to stop performing depth and actually develop it — those are the ones this moment was made for.

The Living Timeline 

I'm leaving the 2023 posts exactly as they are. Not because they're wrong — but because they document honestly where the conversation was, and where I was inside it.

The fact that the frame has shifted so completely in three years is itself the evidence. AI isn't a tool we adopt at our own pace. It's an evolutionary pressure that accelerates what was always true.

In 2023: AI can help coaches deliver better results. In 2026: AI is asking coaches to become something they may not yet be.

The question has changed. Some doors only open from the other side.

Birgitta

This conversation started in 2023. If you want to see how far the frame has moved — and how to use AI to sharpen your own coaching practice right now — this is where it began.