Edge Studio

You may clone only your own voice, or the voice of someone who has clearly agreed. This page says how we check that.

What we require

  1. Recorded consent. The speaker reads a short statement aloud, set by us for the language, which says they agree to the clone being made and used. The recording is uploaded with the speaker's name.
  2. Verification. We check the recording against the voice sample. We store the statement, the recording, your account, time, IP address and device details as proof.
  3. Your promise. You confirm you are the speaker or have their written permission, and that they are an adult.
  4. No consent record, no clone. Professional clones need the same consent and are not in the free trial.

Names we block

We keep a blocklist of public figures and others. Cloning a listed name is refused. See Prohibited voices.

What a clone can be used for

Only the uses the speaker agreed to, within Acceptable Use. Output carries an inaudible watermark and "AI-generated" metadata.

Withdrawing consent and deleting a voice

  • The speaker can withdraw at any time by email to [email protected] with the subject "Voice removal", or the account owner can delete the voice in the product.
  • We stop new generations with that voice immediately and delete the voice model and the source recording within 14 days. We keep the consent record (not the voice) for 6 years to prove what happened.
  • Audio already generated and downloaded cannot be recalled. We ask the owner to take down public uses and we act on reports.
  • If someone used your voice without consent, tell us. We investigate fast, remove the voice, and may ban the account.

Voice in the library

If a voice is shared in the voice library, the speaker also needs to accept Voice library terms.

Minors and the deceased

We do not clone voices of anyone under 18, and do not clone a deceased person's voice unless their estate has consented in writing.

Contact

[email protected]