A question came out of an ICF panel in Paris recently that’s worth sitting with: not what AI can do for coaching, but what it should do.
It’s a small shift in wording. The implications are not small.
The panel was part of the ICF Global Summit, and what stood out wasn’t the technology itself. It was that every panelist had some direct stake in shaping how the profession responds to it. Committees, boards, working groups — people actively trying to get ahead of the question rather than react to it. That matters. The coaching profession has an opportunity to participate in defining those boundaries, and these conversations are part of how that happens.
The “can” question tends to dominate. AI can draft content, synthesize research, support session prep, automate administrative work. It can do a lot, and it’s getting faster. But capability has never been a sufficient reason to proceed. Coaches know this. The entire practice is built on the idea that having something to say isn’t the same as knowing when to say it, or whether to say it at all.
The “should” question is where things get harder and more interesting. It asks about values, not just functions. It asks who’s accountable when something goes wrong, and what gets lost when a human step is removed. It asks whether efficiency is always the right goal in a profession grounded in presence and relationship.
That doesn’t mean resistance. It means discernment. Subject matter expertise still matters. A final human eye on AI-generated content isn’t optional. And the profession’s willingness to set guardrails proactively, rather than scramble to catch up after the fact, will shape what coaching looks like on the other side of this moment.
The Paris conversation was a reminder that those guardrails don’t appear on their own. They get built by people who care enough to show up and argue for them.
What do you think AI should do in the coaching space? The conversation is worth having.
