Ferment — Voice & Biometric Data Consent

Last updated: June 1, 2026

Before Ferment creates a voice from your recording, here’s exactly what happens and what you’re agreeing to.

What we collect. When you provide a voice, we create a voiceprint — a mathematical model of its unique vocal characteristics — from that recording. In many places this counts as biometric data.

Why. We use it for one thing: to build a voice model so you can generate audio in that voice inside Ferment. We don’t use it to identify you elsewhere, to profile or target you, or to infer anything about you.

Who processes it. Your recording and voiceprint are stored in our own cloud storage (on Nebius and Tigris), and we use a third-party voice-AI provider to create and run your voice model (named in our Privacy Policy). We choose a provider whose terms don’t claim the right to train its own models on your voice. The provider stores a voice model to generate speech; when you delete your voice or account, we delete our copy and instruct the provider to delete its model and the source audio. They’re an independent provider acting under their own terms. We don’t sell your voice data, and we don’t share it except as described in our Privacy Policy.

How long we keep it, and when we destroy it. We keep your recording and voiceprint until you delete the voice or your account, and in any case no longer than 3 years after your last interaction with Ferment (or, where Texas law applies, no later than 1 year after the purpose is fulfilled) — then we permanently destroy them. When you delete, we promptly remove them from active storage and delete them from backups on our normal backup cycle, and we instruct our voice-AI provider to delete its voice model and the source audio, except for limited records we must keep for legal reasons. Our full retention and destruction schedule is in the Privacy Policy under “Voice and biometric data.”

You’re in control. This consent is optional — the voice features just won’t work without it. You can withdraw it and delete your voice at any time in the app, which deletes the recording, the voiceprint, and the model.

You must be 18+ and have the right to the voice. You must be 18 or older to create a voice. Use your own voice — or someone else’s only if you have their express written consent and the rights to use it (for example, a voice actor you’ve hired). Don’t create a voice from a recording you found online, pulled from a video or social media, or took from a public figure, performer, creator, colleague, friend, family member, or client without their written consent. Never use the voice of anyone under 18, and don’t upload a recording that contains a child’s voice. We may ask you for proof of consent before or after we create the voice model — if you can’t provide it, we may refuse, suspend, or delete the voice and any output. You’re also responsible for disclosing synthetic or cloned-voice output where the law, a platform, or the context requires. See our Terms of Service, Section 10.

Your consent and release. This is a separate, optional biometric consent, independent of accepting our general Terms and Privacy Policy. By checking the box, you confirm you’ve read this notice, that a voiceprint is biometric data, that you’re acting voluntarily, and that checking the box is your electronic signature. Then: