Permission is part of the voice—not an afterthought.

Before recording or uploading

Tell the speaker that AI will analyze the recording to reproduce vocal characteristics. Explain what kinds of text may be generated, who will hear it, whether files may be published or monetized, how the reference voice will be stored, and how to request deletion.

Keep a record of consent appropriate to the risk and project. Written consent is strongly recommended for public, paid, organizational, or ongoing use. A parent or lawful guardian must provide consent where required for a minor, and high-risk use involving minors is not permitted.

Uses that are never permitted

• Pretending to be another person in calls, messages, authentication, or transactions
• Creating false endorsements, evidence, news, political statements, or emergency instructions
• Harassment, sexual content involving a non-consenting person, exploitation, extortion, or reputational harm
• Cloning deceased people without appropriate rights and sensitive consideration for family and audience
• Circumventing a platform’s disclosure, safety, identity, or copyright controls

Disclosure and downstream use

Clearly disclose synthetic or AI-generated speech whenever a reasonable listener might believe a real person made the recording, especially in public, commercial, educational, journalistic, political, or customer-facing contexts. Do not remove provenance or disclosure information added by another service.

If you share a generated file, ensure recipients understand any limits on editing, redistribution, or reuse. Consent to create one project is not blanket permission for every future project.

Withdrawal and deletion

If a speaker withdraws consent, stop future generation and distribution within your control, delete the saved voice from SuraYomi, and address previously shared files where reasonably possible. SuraYomi can remove a stored voice or suspend an account when credible misuse is reported.

Report a concern

Send concerns to [email protected] with the relevant account or file information, why you believe the voice was unauthorized, and any supporting evidence. Do not send sensitive identity documents unless requested through a secure follow-up. We will review credible reports and may restrict access while investigating.