AI in Coaching: Just Because We Can, Does It Mean We Should?

 

AI is becoming part of coaching practice remarkably quickly. Coaches are using it to prepare for sessions, support their reflection, produce transcripts and summaries, create resources and help with the business side of coaching.

And there are some genuinely useful ways to use it.

But the question I’ve become increasingly interested in isn’t “What can AI do?”

It’s “How do I want to use it as a coach?”

We explored this recently in Cath’s Coaching Circle when we spent a session looking at AI and coaching ethics. Rather than looking at a list of rules about what coaches should and shouldn’t do we talked about our own practice and some of the decisions we’re already making, sometimes without really noticing we’re making them.

 

A Compass Rather Than a Rule Book

One of the things that struck me when I was preparing our Coaching Circle session was that AI is changing far too quickly for us to rely on a simple list of dos and don’ts. There will always be new tools and new situations that we haven’t encountered before.

So I created the AI Coaching Compass as a simple way of helping coaches pause and think before using AI in their practice.

It isn’t an ICF framework, although I’ve developed it with the ICF guidance and our ethical responsibilities as coaches very much in mind. It’s a set of five questions we can return to when the answer isn’t immediately obvious:

Is my judgement still at the centre?
Am I using AI to support my thinking, or am I starting to hand over decisions that belong with me as the coach?

Am I protecting client confidentiality?
What information am I sharing, where is it going and do I understand how it will be handled?

Would this client be identifiable?
Have I genuinely anonymised the information, or could the combination of details still reveal who the client is?

Am I being transparent?
Does my client know when and how I’m using AI in relation to our coaching?

Is this serving the client?
Perhaps the most important question of all. Am I using AI because it genuinely contributes to the coaching, or simply because it’s quicker or more convenient for me?

I think of the Compass as exactly that: a way of finding our bearings when we’re navigating something new. It doesn’t make the ethical decision for us. It helps us slow down long enough to make one consciously.

One area where these questions become particularly relevant is the increasing use of AI to produce transcripts of coaching sessions.

 

AI Transcripts – Useful, but…

I can absolutely see the attraction.

A transcript can be incredibly useful for reflective practice, mentor coaching and reviewing our coaching against the ICF competencies. AI transcription tools make something that was once quite laborious remarkably easy.

But we’re giving an AI tool access to the entire coaching conversation.

So before we press record, there are questions to consider.

Does the client know AI is being used?

Have they agreed to the recording and how it will be used?

Do we know what happens to their data, where it’s stored and whether the platform is appropriate for confidential information?

And once we have the transcript, there are still choices to make about what we do with it.

 

Taking Out the Name Isn’t Enough

Suppose we want to use that transcript with another AI tool to reflect on our coaching.

We remove the client’s name.

Is it now anonymous?

Not necessarily.

A transcript might contain names of colleagues or family members, organisations, job titles, locations, dates, particular events or specific numbers such as headcount or the value of a deal.

There may also be sensitive personal information relating to health, ethnicity, religion, sexuality or political views.

And sometimes the risk isn’t one particular detail- it’s the combination of several.

Imagine describing someone as:

“A female Finance Director in a pharmaceutical company in Manchester whose organisation has just announced 200 redundancies.”

There’s no name there, but could someone work out who she is?

That’s why I think a much more useful question than “Have I removed their name?” is:

“Could someone identify this person from what I’ve left in?”

 

Open AI Changes the Conversation

There’s also an important distinction between using an appropriate AI transcription service with the necessary safeguards and taking the resulting transcript and putting it into an open generative AI platform.

They’re two separate decisions.

If it’s identifiable or sensitive client data, it shouldn’t be going into open AI.

We need to understand the tools we’re using and their data policies, as well as our responsibilities around confidentiality and data protection. And ultimately, whatever AI does with the information we give it, the accountability for choosing to use it remains with us.

 

And What About the ICF Code of Ethics?

For those of us who are ICF credentialed, this isn’t a separate conversation from coaching ethics simply because the technology is new.

Our existing ethical responsibilities still apply when we’re using AI.

We still need to protect confidentiality, be clear with clients about how their information is being used and take responsibility for the choices we make. Using an AI tool doesn’t hand any of that responsibility over to the technology.

For me, that’s an important distinction. We don’t need a completely different set of ethics for AI. We need to apply the same care and professional judgement to a new and rapidly changing context.

And because the technology is changing so quickly, I think that requires us to keep asking questions rather than deciding we’ve got our AI practice sorted once and for all.

 

What Happens to Us as Coaches?

There’s another part of this conversation that fascinates me just as much as the ethics.

What happens to us as coaches as AI becomes increasingly capable?

If I routinely ask AI to suggest my questions, am I still developing my ability to listen and respond in the moment?

If AI analyses my transcripts and tells me what I missed, does that deepen my reflective practice or could I gradually become dependent on it?

 

I don’t think there’s one simple answer.

Which is why, alongside “Is it appropriate to use AI here?”, I think there’s another question worth asking:

“What happens to me as a coach if I do?”

Because AI can generate a coaching question in seconds…

But it still can’t wonder.

 

So, as AI becomes more capable, what part of your coaching are you determined to keep distinctly human?

Until next time,

Cath

x