Mistral now trains on user input by default, except on enterprise tier

Mistral has changed its default training process to incorporate user input, except for enterprise tier customers. This means that, by default, user-generated content such as conversations, documents, and other provided material could be used to train Mistral's models. However, users retain full control over this data usage and can opt-out of the training programs at any time.

The opt-out option varies depending on the specific service or platform being used, with detailed instructions available for users in their respective settings.

In the case of enterprise tier customers, Mistral opts them out of training by default. These customers have the ability to opt-in to the training process via the Admin panel, specifically by disabling the "Allow your interactions to be used to train our models" toggle within the Privacy section. This opt-in toggle is managed at the administrative level.

For users accessing Mistral through iOS or Android, the process involves adjusting the Data & Account Controls. By navigating to the Data & Account Controls panel under the Account section, users can deselect the Enable data sharing checkbox to opt out of Mistral's training program. This allows users to maintain control over their data and ensure it is not used for model training.

It's important to note that users have the right to opt out at any time, and the opt-out process does not automatically apply to both the Mistral Studio and related API services. These services have separate opt-out toggles, which must be configured individually by the user.