This notice explains how Citra Labs handles faces and voices. It supplements our Privacy Policy, which covers everything else.
Only when you give us a real person's face or voice — uploading a presenter photo, taking a selfie in the setup wizard, or using a photo that contains people. It does not apply to our preset presenters: those faces are generated by a model and belong to no one. It does not apply to product shots, backdrops, screen recordings or text.
The photograph, video or audio you provide, and the facial and voice characteristics our models derive from it in order to animate a presenter. Under Malaysia's Personal Data Protection Act 2010, as amended in 2024, this is sensitive personal data.
One thing: generating the video, image or audio you asked for. We also run an automated check on an uploaded portrait to tell you whether it will animate well, and automated safety screening to detect misuse.
We do not use your face or voice, or anything generated from them, to train, fine-tune or improve any AI model. We do not sell this data, use it for advertising, or share it with anyone except the processing partners in section 5. If we ever want to use it for model improvement, we will ask you separately, with an unticked box, and you will be free to say no without losing access to anything.
Generating video is computationally heavy, so this data is processed by GPU cloud providers running our own models, and by third-party AI providers for specific steps such as the portrait suitability check and certain image operations. These partners operate in the United States, Singapore and, for some services, the People's Republic of China. Each is engaged on terms requiring them to process the data only for us and not to use it to train their own models. We keep a register of these partners, their commitments and the source of each; it is available on written request through our contact form.
Uploaded portraits, recordings and the presenters built from them stay in your account until you delete them or close your account, after which they are removed from active storage. Deleting a presentation deletes the portrait it was built from.
We ask for your consent separately, before the first time you give us a face or voice — the general terms you accept at sign-up do not cover this. You can withdraw that consent at any time through our contact form. Withdrawal stops any new face- or voice-based generation on your account. It does not undo videos already made, and it does not affect the rest of the service.
If you upload media of anyone other than yourself, you must have their informed consent first. Our Terms of Service make this a binding warranty, and our Acceptable Use (Terms §3) rules prohibit deceptive deepfakes. We may ask for evidence of consent and remove content without it.
You may ask what biometric data we hold, ask us to correct it, or ask us to delete it. Use our contact form and choose “Privacy or data request”. If you are not satisfied with our response you may complain to Malaysia's Personal Data Protection Department (pdp.gov.my).
This platform is for adults. Do not upload media featuring anyone under 18. If we find a child's face or voice on the platform, we delete it.