Creator consent
Voice & likeness consent
Draft for legal review · Updated 21 August 2026
Separate permissions
The interview records three separate grants: recording/transcription, likeness use and voice use. After Dark text, After Dark voice and generated video require additional grants; one permission never silently authorises another.
What is captured
The browser records a private low-bandwidth audio track and a private video track. Audio is transcribed for persona training. Video is retained as a likeness source. Capture status and upload progress must remain visible throughout.
Automatic training
Completing a consented likeness interview automatically trains the creator's AI. Creators can add further interviews and training sources at any time, take the AI offline immediately, revoke consent, and delete stored likeness sources from their account.
Revocation
The creator may revoke consent and request deletion. Publication and future model use must stop immediately; source files and derived models then enter the permanent deletion workflow. Audit records may be retained only as necessary to prove the revocation was honoured.
No third-party cloning
By consenting, the creator confirms they are the represented adult and have authority over all uploaded media. Consent cannot be granted for a celebrity, partner, employee, customer or any other third party without independently verified authority.