Legal & Privacy FAQ

Voice data is special category data under Article 9 GDPR only when it is processed for the purpose of uniquely identifying a person. Voice cloning does not serve that purpose, it reproduces how a voice sounds. We therefore do not treat it as special category processing.

If you wish to use the voice cloning services provided by Gradium for the purpose of uniquely identifying an individual, you must ensure that such processing is lawful, for example, by obtaining the specific individual's consent for that identification purpose.

Yes. Our terms require the explicit consent of the voice owner, so that the person has agreed in advance to their voice being cloned. Collecting and documenting that consent is your responsibility.

It depends on your plan:

• Self-serve plans: your use of Gradium is governed by three documents: (i) our Terms of Service, which set out the conditions for using the Services; (ii) our Data Processing Agreement, which applies when we process personal data on your behalf and is incorporated in the Terms of Service; and (iii) our Privacy Policy, which explains how Gradium processes personal data as a controller. By using the Services, you accept these documents in their current version.
• Enterprise plans: your use of Gradium is governed by a dedicated contract, agreed with you and adapted to your needs. To discuss an Enterprise arrangement, contact us.

No. On self-serve plans, our DPA applies automatically when you use the Services, and no separate signature is required. On Enterprise plans, the DPA is included in your contract.

In our Trust Center at trust.gradium.ai. Go to the Security controls section and click Download. The document describes the technical and organisational measures we apply to protect your data. The version that applies is the one in force on the date you download it.

For any question, contact our privacy team.

Our subprocessors are the providers we use when Gradium processes personal data on your behalf, as a processor under our Data Processing Agreement. This covers the processing of the content you send through our Services on your own end users' data, and the infrastructure that runs the platform.

The up-to-date list is available in our Trust Center at trust.gradium.ai.

Yes. Gradium uses service providers for its own operations, where Gradium acts as a controller rather than on your behalf. These providers do not process the content you send through our Services, nor data about your end users.

For example, we use providers for authentication of users signing into Gradium Studio, payment processing, or email communications.

These providers are described in our Privacy Policy.

No. We do not use the content processed through our Services to train or improve our models. This covers your data and the data of your end users.

The only exceptions are content you have authorised us to use, and content you choose to share with us directly, for example to report an issue or help us improve our Services.

Yes. Voice samples you upload are used only to create and run your custom voice. They are not used to train our models without your prior authorisation.

Nothing is stored after the call, so nothing can be used for any other purpose. See the Zero Data Retention section below.

With Zero Data Retention enabled, we do not store the text and audio processed during your sessions. It is self-serve on paid plans, and you can enable it in Gradium Studio.

Zero Data Retention does not cover data that must persist for the service to work, such as account data.

For more details, see our dedicated page: EU data residency and Zero Data Retention.