Voice & Biometric Data Notice
Recall — AI Memorial
Effective date: August 2, 2026 Provider: Stefano Anania · support@stefanoanania.com · 525 N Sycamore Ave, Los Angeles, CA 90036, USA
This notice is the written disclosure and retention schedule required by biometric privacy laws including the Illinois Biometric Information Privacy Act (740 ILCS 14), the Texas Capture or Use of Biometric Identifier Act (Tex. Bus. & Com. Code § 503.001), and Washington's My Health My Data Act. It forms part of the Privacy Policy and the Terms of Use.
Ticking the confirmation box in the app before uploading a voice sample is your written release under BIPA.
1. What is collected
When you set up a voice for a loved one, Recall collects:
- A voice recording — audio you record in the app, or an audio or video file you choose from your device. The app analyses it and uploads only the best segment, up to ten seconds, not the whole file.
- A voiceprint — the mathematical voice model our provider derives from that segment. This is biometric information under the laws named above.
- The name you gave the voice, and the language it speaks.
2. Why it is collected
For one purpose only: to synthesise speech in that voice inside the Recall app, at your request.
It is not used for identification, authentication, verification, security, surveillance, advertising, profiling, research, or model training. It is not used to build any product other than the conversation you asked for.
3. Consent, and why we ask for it the way we do
The person whose voice is being recreated has died and cannot consent. That is why the obligation falls to you.
Before any sample is uploaded, you must confirm in the app that you are the legal next of kin of the deceased, or otherwise hold explicit legal authority to use their voice. This is not a formality:
- It is enforced on our servers. An upload that does not carry the confirmation is rejected.
- We record that you confirmed and when, stored against the voice it authorises. We record no audio and no biometric material in that record.
- If a family member later disputes a voice, that record is what we have.
You may not create a voice for a living person. See section 5 of the Terms of Use.
4. Who receives it
| Recipient | What they get | Role |
|---|---|---|
| Cartesia (privacy policy) | The audio segment, the voice name, the language. Stores the derived voice model. | Our processor, under terms prohibiting use of your data for model training |
| Nobody else | — | — |
We do not sell, lease, trade, or otherwise profit from voice or biometric data. We do not disclose it to any other party except where compelled by valid legal process, and we will tell you if that happens unless we are prohibited from doing so.
The original recording stays on your device. Our own servers do not keep a copy of it.
5. Retention schedule
| Item | Where | Kept until |
|---|---|---|
| The original recording | Your device only | You delete the loved one's profile, or delete the app |
| The voiceprint / voice model | Cartesia | You delete the loved one's profile, which sends a deletion request to Cartesia |
| The authorisation record (date + voice name + anonymous identifier) | Our database | The voice model is deleted |
Backstop. If you never ask us to delete a voice, we will delete it no later than three years after your last use of that voice, or when the purpose for collecting it has been satisfied, whichever comes first — the standard BIPA requires.
6. How to delete a voice
In the app: delete the loved one's profile. This removes the local sample, and sends a deletion request to Cartesia for the voice model. A voice can only be deleted by the account that created it.
By email: write to support@stefanoanania.com.
If you are next of kin and did not create the voice: if someone has created a voice of your deceased relative without authority, write to support@stefanoanania.com with enough detail to identify the voice and your relationship to the deceased. We will investigate and delete it where the claim is substantiated, and we may terminate the account that created it. You do not need to be a Recall user to make this request.
7. Security
The voice sample is transmitted over TLS to our backend, and from there to Cartesia. The account credentials for our voice provider are held in Google Secret Manager and never leave the server — the app itself can only ever obtain a short-lived, synthesis-only token, which cannot create or delete voices. We protect biometric information using at least the reasonable standard of care used for other confidential and sensitive information, and in a manner at least as protective as we use for our own confidential information.
8. Your rights
Depending on where you live you may have the right to know what biometric information we hold, to obtain a copy, to have it deleted, to withdraw consent, and — in Illinois — a private right of action for violations of BIPA. Write to support@stefanoanania.com; see section 9 of the Privacy Policy for how we handle requests.
Questions: Stefano Anania, 525 N Sycamore Ave, Los Angeles, CA 90036, USA, support@stefanoanania.com