Legal
Voice data policy
Last updated October 4, 2026
Most of the assessment is spoken. This page says exactly what happens to your voice, in the order it happens.
You choose to speak
Every voice question has a typed alternative. Nothing is recorded until you press the record button, and you can re-record or switch to typing on any question. Your browser asks for microphone permission the first time.
The recording is transcribed
The clip is uploaded to a private bucket in our database provider's storage in Frankfurt (EU), reachable only by our server. It is sent once to a speech-to-text provider, in your language, to produce a transcript.
The audio is deleted
As soon as a valid transcript exists, the server deletes the clip. A daily job deletes any clip older than a few days that slipped through (for example when a transcription failed and you typed instead). Nobody at Instinct AI or your employer listens to recordings; there is no player in the product.
What the transcript is used for
The transcript text, with your other answers, is what the scoring model reads and what your plan, exercises and playbooks are written from. Your employer's admins can read the transcript of your answers in your session detail; they cannot hear your voice.
What is never done
Audio is never sent to the scoring or writing models, never used to identify you by voice, never used to train anyone's models and never kept past transcription.
Your choices
Type instead of speaking on any question. Ask us to delete your transcripts and answers at any time (your employer's report keeps aggregate figures only).
Questions about this document: hello@enable.instinctai.io.