Voice Clone Consent

Version 0.1 (draft) · Effective: {{EFFECTIVE_DATE}} · Last updated: 24 October 2026 · Consent text version: vc-2026-10-v0.1

In short (plain language)

• You can clone only your own voice, and only if you are 18 or older.
• Before we create a clone you read and accept the consent below, then read a random sentence aloud, live. This proves the real person is present and agrees — it stops people cloning someone from a video or a phone call.
• Your clone is private to your workspace. We never put it in the public library.
• We do not train AI models on your voice.
• You can delete your clone at any time. It stops working immediately and we delete your recordings.

A. The consent you accept in Voice Lab

This is the exact text shown before you create a Voice Clone. You must tick all five boxes to continue.

Create a voice clone — your consent

Trinetra AI Labs Private Limited ("DamruVani") will use the voice recordings you provide in this step — your voice sample(s) and a short live recording of a sentence we show you — to check that the voice is yours and to create a private voice model that can generate speech in your voice inside your workspace.

☐ 1. This is my own voice. The recordings I provide are of my voice only. They are not of anyone else, not of a famous person, and not taken from another AI service.

☐ 2. I am 18 or older.

☐ 3. I consent to DamruVani processing my voice recordings, creating a voice model from them, and storing them and the model until I delete this voice or my account, as described in the Voice Clone Consent and the Privacy Policy. I understand that my voice is personal data and may be biometric information.

☐ 4. I understand how my clone may be used. Everything generated with it is labelled as AI-generated. During the beta it may be used only for non-commercial purposes. I will not use it to deceive anyone, and I will follow the Acceptable Use Policy.

☐ 5. I understand the checks. DamruVani will compare my live recording with the sentence shown and with my sample, may compare my voice with its library voices and known public figures, and its staff may listen to my recordings to review this clone for safety.

DamruVani will not use these recordings to train or improve any AI model.

I can withdraw this consent at any time by deleting the voice.

[Continue] [Cancel]

We save: which boxes you ticked, the consent text version, the date and time, your IP address and browser, and the live recording. [LAWYER: confirm this record satisfies "consent in writing" under SPDI Rule 5(1) and "free, specific, informed, unconditional and unambiguous" consent under DPDP §6(1)]

The consent is shown in your chosen UI language (English, ગુજરાતી, हिन्दी). [VERIFY: translations reviewed before launch] [LAWYER: which language version prevails]

B. Who can create a clone

• Only people 18 or older, on a plan that includes Voice Lab (during the beta: testers we approve).
• Only from your own voice. Cloning another person's voice — including a family member, a colleague, a client or a voice actor — is not allowed in Voice Lab, even with their permission. [OWNER: a separate contract-based flow for professional voice actors may come later]
• Never a child's voice. If our checks suggest a recording may be of a child, the clone is held for manual review or refused.
• Not the voice of a well-known person (politician, film star, singer, sportsperson, religious figure), and no "soundalike" of one.

C. The challenge sentence — what it is and why we use it

1. When you start, we show a random sentence of about 8–10 words in your language. If speech recognition does not yet support your language, we use another supported language you can read. A new sentence is generated for every attempt.
2. You read it aloud and record it live in the browser or app, within 2 minutes. Uploading a file is not accepted for this step.
3. We automatically check that (a) the recording says the sentence we showed, using our speech-to-text, and (b) the voice in it matches the voice in your sample.
4. We also check that your sample does not match one of our Library Voices (so previews cannot be used to copy them), and that the voice name or description does not match our list of well-known Indian public figures.
5. Some clones — including every clone created in the early beta — are reviewed by a person on our team before or after they are activated. Reviewers can play the recordings but cannot download them, and every access is logged.
6. If a check fails, the clone is not created. There is a daily limit on attempts.

Why: a random sentence spoken live cannot easily be faked from an existing recording. It shows that the real speaker is present and agrees to the clone, which protects people from having their voice copied without permission.

D. What we store and for how long

Swipe sideways to see the whole table →

Your clone is never shared with other users, never used for anyone else, and never added to the public library. Your recordings are not used for model training.

E. Withdrawing consent (revocation)

You can withdraw at any time: Voice Lab → your voice → Delete voice, or delete your account.

Swipe sideways to see the whole table →

F. If someone cloned your voice without permission

Report it at /legal/grievance or with the Report button on any audio. Choose "Impersonation / my voice used without consent". We will disable the voice and related content while we investigate, and act within the timelines on the grievance page. If you are a victim of fraud, also report it at cybercrime.gov.in or call 1930. [VERIFY: helpline number current]

G. Library voice actors

Professional voice actors and team members who record DamruVani's Library Voices sign a separate written contract covering use, payment, exclusivity, training permission (always a separate opt-in) and withdrawal. This page does not apply to them.