Why the coach won't say which AI it uses

What this does

Ask the coach “what AI are you?” or “which model is this?” and it will not answer. It stays in character and moves the conversation back to your trading. This is written into its instructions on purpose, and it is the one question it is built to sidestep.

Why it does that

VEXA is meant to be one coach with one voice. A coach that stops mid-conversation to describe its own infrastructure is a different product, and the answer is of no use to someone trying to work out why they keep sizing up after a loss.

The same instruction stops it discussing its own prompt, the names of the systems it reads from, or where your data sits. It is a consistency rule rather than secrecy.

Where the answer is published

The provider is named in the privacy policy, which is the document to read or to cite. The AI consent dialog inside the apps describes the same provider as “a secure, US-based third-party AI service” without naming it.

If you need the answer for a compliance review, a vendor assessment or a data-protection question, the privacy policy is the authority, and support can point you at the current subprocessor detail.

For what actually gets sent, feature by feature, see What data VEXA sends to its AI service.

What you should see

A reply that acknowledges the question and returns to coaching, for example an offer to look at your week instead. The deflection is brief. You should not see a refusal, an error, or a claim that VEXA is human.

If it doesn’t work

It gave a name. That contradicts its instructions. Report it to support with the wording, because it means something changed.

It claims to be a person. It should not. Report that as well.

You asked about the AI and lost the thread of your session. Ask your original question again. The deflection does not clear its memory of what you were discussing.