Skip to content
humperdinck.
HomeStudioDownload

04 / RESPONSIBLE VOICE

AI Voice & Consent Policy.

Voice tools should be used with permission, transparency and respect for the people behind the recordings.

Pre-launch information · Updated November 7, 2026

LEGAL / INFORMATION

PrivacyTermsSoftware licenseAI voiceRefundsCookiesCopyrightLicenses
Pre-launch information. This page explains current practices and planned policies. Certain commercial, account and community features are not yet available; remaining legal and operational details will be published before those features launch.
01

Consent is required

Only upload, clone, imitate, generate or distribute an identifiable person's voice when you have a lawful basis and the necessary permission, or when your use is otherwise lawful. A recording being publicly available does not by itself grant voice-cloning rights.

02

Never misrepresent identity

Do not present synthetic audio as an authentic statement, endorsement, call or performance by another person when it is not. Do not use voices to impersonate government officials, employers, relatives, financial institutions or other people for deception, harassment or fraud.

03

Prohibited harmful uses

Do not create nonconsensual intimate or exploitative voice content, facilitate scams or coercion, evade biometric voice checks, infringe rights, or knowingly spread deceptive audio in contexts where it risks significant harm.

04

Responsible publication

Clearly label synthetic content where disclosure is required or necessary to avoid misleading an audience. Obtain permission for commercial endorsement uses. Respect performer, publicity, privacy, copyright and platform rules.

05

Community voice library — not yet launched

Community voice publishing is planned only for registered adults aged 18 or older, and is private-by-default and opt-in for each individual voice. A publisher's age eligibility does not establish legal rights or consent to use any speaker's voice, including a minor's voice. Creating, saving or previewing a voice does not publish it. A community submission may contain only the reusable voice package, a generated voice preview sample, minimal public metadata and the rights-and-consent attestation. Original source recordings, unrelated narration scripts and project files must not be silently packaged, uploaded or published. The publication workflow must present a preview of the selected content and obtain explicit submission confirmation. A generated sample still requires appropriate speaker rights and consent. Before submitting a selected voice for publication, users must explicitly confirm they have the rights and permissions needed to share it, including consent from an identifiable speaker when required. Every new community submission will remain nonpublic and pending review until approved. The planned review includes automated checks and an approval decision; automated checks are not proof of consent or ownership, and Humperdinck may require further evidence, reject or remove a submission. By choosing to publish a voice, its creator will grant other users a non-exclusive permission to use that voice for lawful personal and commercial narration and creative projects, including monetized videos, podcasts and audiobooks. The permission covers use of the voice to generate and distribute audio outputs, not resale, redistribution, sublicensing or repackaging of the voice files, weights, embeddings or voice package itself. Publishing does not transfer ownership of any rights the creator holds or expand permissions beyond those the creator lawfully possesses. No permission is granted to falsely claim endorsement or impersonate an identifiable person. Third-party model and source-recording restrictions continue to apply. Creators can request that their published voice be removed from Humperdinck-hosted listings and files. Removal prevents new distribution by Humperdinck after processing but cannot recall copies already downloaded or shared outside Humperdinck. Removing a voice from Humperdinck stops new downloads, but does not automatically revoke licenses already granted to users who lawfully downloaded it. Those users may continue using their copies to create lawful personal and commercial narration and other projects under the original license. This continuing permission is subject to applicable law and exceptions for infringement, missing rights or consent, fraud, and other legal requirements. Humperdinck cannot guarantee deletion of copies already downloaded, and removal of a hosted listing does not validate an otherwise unlawful copy. Community voices that have been approved and published will remain subject to user reports, investigation, restriction or removal for policy or rights violations. A reporting channel, review queue, moderation/removal procedures and appeals process must be implemented before public uploads begin; this document does not claim those systems are operational.

06

Report misuse — publication required

A private abuse-report channel and documented rights-verification, removal and review procedures will be provided before community uploads open. For copyright allegations, see the Copyright & DMCA page.

See the Privacy Policy and Terms of Service for related information.

humperdinck.A local-first voice studio.
FeaturesStudioDownloadPrivacyTermsSoftware licenseAI voiceRefundsCookiesCopyrightLicenses
© 2026 Humperdinck · Pre-launch information