Clone an executive's voice with consent

Email an executive a one-use link so they record their own consent and voice sample, what they see, the five statuses, and how they withdraw.

An executive voice is a clone made only after the person records their own consent. You never handle their recording. They get an email from you with a one-use link, read what the copy is for, record one sentence of consent, then record or upload two minutes of speech. The clone is created only then.

Before you start: tell the executive the email is coming. The link works once and expires in 7 days.

​

Before you start

• The executive’s full name, role and work email.
• A free executive voice slot. A workspace holds 3 by default, counting invitations not yet answered. See Add a voice.
• Written authorisation for the programme. See Before you test anyone.

​

Steps: send the invitation

1. Open Voices and click Invite an executive

   On the Voices page click Invite an executive. A form opens, titled Invite an executive to record their consent.

2. Fill in who they are

   Enter Full name, Role and Work email. If your workspace has more than one voice engine, pick one under Voice engine.

3. Click Email the consent link

   You see “Invitation emailed to [name]. The link works once and expires in 7 days.” The invitation appears under Consent invitations with its status.

​

What the executive receives

The email subject is [Your company] asks for your consent to a voice clone for security training. It says your company runs phone-based security awareness tests on its own staff, names you as the person asking, and gives the link.

The link opens a page titled Consent to a copy of your voice. It takes about five minutes.

1. They read five points

   What this is, what it is used for, what is stored, that they can withdraw, and who to contact. They tick “I have read the five points above and I agree.”

2. They say one sentence

   Under 1. Say this sentence out loud they click Record the sentence and read: “My name is [name]. I agree that [company] may make a copy of my voice and use it in security awareness tests of its own staff. I know I can withdraw this at any time.” Re-record lets them try again.

3. They record two minutes of speech

   Under 2. Record about two minutes of speech they click Start recording and talk naturally, or click Upload a file. A live recording must reach at least 60 seconds; two minutes gives the best copy.

4. They submit

   They click Give my consent and submit. The page then reads Thank you, [name] and reminds them to keep the email so they can withdraw later. If they do not agree, they close the page and nothing is stored.

​

The five statuses

Status	What the Voices page shows	What it means
Sent	Invitation sent, waiting for the executive	The email went out. Shows the expiry date
Recording	The executive is submitting now	The link was opened and a submission started. A second submit is refused
Recorded	Consent recorded	The clone exists and can be used
Revoked	Consent withdrawn	The executive withdrew. The clone is deleted
Expired	Invitation expired	7 days passed. Send a new invitation

​

What is stored

Item	Why
The two minutes of speech	To build the clone
The recording of the consent sentence	As the record of what they agreed to
Name, role and email	To show whose voice it is

Questions about this data go to [e-mail].

​

Withdrawal

The executive opens the same link again. The page shows Your voice clone with the date they consented, and a button Withdraw my consent. After confirming, the clone is deleted at the voice service, the speech sample is deleted, and every test that uses the voice is stopped. Only the consent recording and the time of withdrawal are kept.

On your side the voice shows Withdrawn. Any campaign pinned to it is blocked and its page shows the reason: the person withdrew their consent and the clone cannot be used. Pick another voice on the attacker, or invite the executive again.

​

What blocks a launch

• An attacker whose voice has No consent or Withdrawn.
• A withdrawn voice on a campaign that is already scheduled: the campaign stops.

Launch readiness is checked on the campaign page and again on the server. See Authorise and launch.

​

Voices that are not an executive’s

For a colleague who is in the room, use Clone a voice on the Voices page and tick the consent box there. For an older clone that shows No consent, click Attest consent, type their name, tick the confirmation and click Record consent. An executive voice refuses this: the message reads “An executive voice needs the person to record their own consent.”

​

Problems

The invitation is still 'waiting' after a week

The link expired after 7 days. Send a new invitation; the old one does not count against your slots any more.

The executive says the link is unavailable

The page reads Link unavailable when the link was used, expired, or is malformed. Send a fresh invitation. Each link works once.

The executive cannot record

The page says “This browser cannot record audio here. Open the link in Safari or Chrome, or upload a file instead.” They can upload a recording made on their phone.

Invite an executive is greyed out

The workspace is at its executive voice limit, counting open invitations. Withdraw one, or write to [e-mail] to raise the limit.