AI Coaching
Data and Privacy in AI Coaching
Clients share real, sometimes sensitive things in a check-in. Handling that data well is part of the product. Practical principles for trustworthy AI coaching, and what to tell clients.
A good check-in invites honesty, which means clients share real and sometimes sensitive things. How you handle that data is not a back-office detail; it is part of the product and part of the trust that makes async coaching work.
Collect less, on purpose
The safest data is the data you never collect. Ask only for what genuinely improves the next session. Minimal collection reduces risk, simplifies compliance, and signals respect. If a field does not change what the client receives, drop it.
Keep the human in the loop on sensitive material
Automation should never be the final word on anything sensitive. A human-in-the-loop setup routes distress or risk signals to a person and keeps the coach validating delicate sessions. The system prepares; the human decides.
Be explicit about training and access
Two questions matter most to clients, even if they do not ask them: is my data used to train AI, and who can see it. A trustworthy setup keeps client data out of model training, restricts access, stores it securely, and uses it only to personalize that client’s experience. Confirm this with your provider, not just your own intentions.
Tell clients plainly
Trust comes from clarity, not silence. Tell clients that AI helps prepare personalized sessions, that you stay in the loop, what is collected and why, who can see it, and how to request deletion. Said plainly, this reassures far more than it worries, and it is the foundation the whole model rests on. (This is general guidance, not legal advice; follow the privacy laws and professional rules that apply to you.)
Frequently asked questions
Is it safe to share personal things in an async check-in? +
It can be, when the coach handles data responsibly: clear consent, minimal collection, secure storage, and a human in the loop on sensitive material. Clients should be told what is collected, why, who sees it, and how it is used.
Does the AI train on my clients' data? +
It should not, unless explicitly agreed. A trustworthy setup keeps client data out of model training and uses it only to personalize that client's sessions. Be explicit about this with both your provider and your clients.
What should I tell clients about AI and privacy? +
Plainly: that AI helps prepare personalized sessions, that you stay in the loop, what data is collected and why, who can see it, and how they can request deletion. Clarity builds the trust the whole model depends on.