40 Questions Executive Coacheses Should Ask Before Trusting AI
When ChatGPT generates a personality profile before your first conversation with a client, or when BetterUp scores their leadership readiness, you face a choice about whether to use that output. The questions you ask before accepting an AI insight determine whether you remain the primary observer of your client or become a delivery mechanism for someone else's algorithm.
These are suggestions. Use the ones that fit your situation.
Questions about AI-generated client profiles and personality assessments
1Does this AI personality profile come from interaction with my client, or from assumptions made about them based on their role title, company size, or industry?
2If I walk into this coaching conversation already holding this profile in mind, what questions will I not ask because I think I already know the answer?
3Have I verified whether this assessment tool has seen my specific client's actual behaviour, or am I accepting a probabilistic guess about someone like them?
4What does the AI tool define as leadership strength, and does that definition match what actually creates results in this client's organisation and role?
5If I present this profile to my client, will they recognise themselves in it, or will they spend the session defending themselves against a characterisation they don't accept?
6Can I name the specific behaviours I've observed that would justify this profile, independent of what the tool told me?
7Does this assessment measure what I actually need to know about this person's readiness for their next role, or does it measure what's convenient to quantify?
8What interpersonal dynamics has this tool never seen? What does it not measure at all?
9If this profile turns out to be wrong, how much coaching time will I lose correcting it before we get to real work?
10Would I be more curious about this client if I had no AI profile at all?
Questions about your own coaching judgment when AI offers insights first
11When I use Claude or ChatGPT to prepare for a session, am I building my own insight into the client or outsourcing it?
12Can I articulate what I notice about this client without looking at what an AI tool noticed first?
13If I always run a coaching challenge through an AI tool before my session, what am I not learning to recognise on my own?
14How often do I disagree with an AI recommendation, and what do I do when I disagree?
15Am I using AI tools to speed up my thinking, or to replace my thinking?
16What is one pattern I've learned to spot in a client that no AI tool would catch?
17If my coaching effectiveness depends on tools like Humu or CoachHub AI generating my insights, what happens when I work with a client those tools haven't assessed?
18Do my clients perceive me as the expert in their development, or as someone interpreting what their AI assessment told them?
19When a client says something contradictory to what an AI assessment found, do I trust their lived experience or the tool?
20What intuition about a client have I ignored because an AI output suggested something different?
Questions about assessment tools and what they actually measure
21What behaviour does this assessment tool reward, and is that behaviour actually what this client's organisation values?
22If the assessment scores my client as low in some dimension, does that score come from self-report or from observation of actual behaviour with their team?
23Can this tool distinguish between someone who lacks a skill and someone who doesn't use it in their particular role?
24Does this assessment account for cultural or contextual differences in how this client shows up, or does it compare them against a single standard?
25What happens to the score if my client was tired when they took it, or in a bad mood, or thinking about something else?
26Who designed this assessment, and what kind of leader did they have in mind when they built it?
27If two different assessment tools score my client differently on the same dimension, which one should I trust and why?
28Does this tool measure stable traits or current patterns that could shift in a different environment?
29How does this assessment distinguish between genuine development and someone who has learned to answer the questions the way the algorithm expects?
30What critical thing about this client's capability or impact would be missing from this score?
Questions about the coaching relationship and reflective space
31If my client processes every challenge through an AI tool before our session, what reflective work are they not doing?
32When I offer my observations in a coaching conversation, am I offering the client something the AI tools they're already using cannot offer?
33Does this client come to coaching for clarity or for confirmation that an AI tool was right about them?
34If I present an AI insight during a session, am I creating space for the client's own discovery or substituting my interpretation for their wisdom?
35What question about themselves is this client not asking because they've asked ChatGPT instead?
36How much of my value to this client depends on having access to the same tools they do, and how much depends on something else?
37When a client shares something difficult, am I the first person they've told, or have they already workshopped it with an AI?
38Is this client becoming more self-aware through our coaching, or more dependent on external tools to interpret themselves?
39What would this conversation be like if neither of us had access to any AI assessment or insight?
40If my role is to notice what AI tools cannot notice, what am I actually noticing that the tools miss?
How to use these questions
Before your session, write down three observations about your client based only on what you remember. Then check what the AI tools say. The gap between your instinct and the algorithm is where your judgment still lives.
When a client mentions an AI insight about themselves, ask them what they think about it, then wait for the real answer. Most people give the AI's answer first and their own answer second.
Test your assessment tools on yourself. Spend a week using Humu or BetterUp AI the way your clients do. Notice what it gets wrong about you and what it misses entirely.
Keep a record of when you trusted an AI output and when you trusted your own judgment instead. Review it quarterly. You'll see where your judgment has atrophied and where it's still sharp.
In your coaching conversation, commit to one session per quarter where you use zero AI tools, zero assessments, zero prepared insights. Notice what you discover without them and whether you coach differently.