Speakers' consent
Draft: awaiting approval
This summary follows the consent form, which is itself a draft. It is not legal advice. The open questions below are for the operator to answer before it is published.
- The form (version 1) is a draft of 2026-10-08: confirm its wording, then this summary.
- Withdrawal: the form says 'by writing to the organiser'; this summary points to the site's removal form as well. Confirm both routes.
- The fallback voice for a speaker who does not agree: the form says the organiser's own voice. Confirm, or name the voices used.
Some dubs on LinguaForge use a computer-made voice that copies the speaker's own voice. That is done only when the speaker has signed a consent form. This page sums up what the form says. The signed forms are kept privately by the operator and are never published.
What the speaker is told
- The talk is translated into other languages and dubbed with a computer-made voice that copies the speaker's voice.
- To dub faster, the work is shared out to volunteers' computers. Each one receives a short clip of the voice (8 to 14 seconds), kept encrypted and deleted when the job ends, and the translated text of the talk, in parts. Volunteers never receive the recording of the talk.
- Volunteers accept an agreement not to keep, copy or reuse the voice or the text.
- The finished dub is published with the talk, labelled as an AI-generated voice, with a link to the original.
What the speaker chooses
- A copy of their voice may be used by volunteers known to the operator only;
- or also by any registered volunteer;
- or not at all. The talk is then dubbed in a different voice.
The consent can have an end date, or none.
Withdrawing consent
A speaker can withdraw consent at any time, through the removal form or by contacting the operator. No new jobs with their voice are sent out after that. Dubs already published stay published unless the speaker asks for them to be taken down.
How the voice clip is handled is also described in the privacy notice.