VEX never reads a client behind their back. Here is exactly how that works, in plain language you can forward to any client.
Every client is told, before the reading begins, that completing it shares the result with their coach. No one discovers after the fact that a coach was watching.
When a client finishes, they see their own result first. Sharing it with their coach requires their explicit yes at that reveal. If they decline, the coach sees nothing.
What the coach receives is a brief: where the client is, what moves them, what to open with. The client's raw answers stay with the client. They are theirs, and theirs to delete.
The brief is AI-generated from what the client chose to share. Clients are told this. Coaches can say it plainly to clients because it is written down here, on a page anyone can read.
Any aggregate or team view is drawn only from clients who consented, and only when five or more have. Below that floor, no group read exists, so no individual can be singled out by subtraction.
Consent is not a one-time gate. A client can withdraw sharing, and their data drops out of what the coach and any group view can see.
Why this matters for ICF coaches. The 2025 ICF Code of Ethics requires coaches to disclose their use of AI to clients. VEX is built so that disclosure has already happened before you see a word: the client is told AI generates the brief, and nothing reaches you without their recorded yes. You are not retrofitting an ethics conversation. The tool had it first.