When you click thumbs up or thumbs down on a Mistral AI response, you are granting Mistral AI permission to use not just your rating but also the full conversation that prompted it for AI training — regardless of your other training opt-out settings.
This analysis describes what Mistral AI's agreement states, permits, or reserves. It does not constitute a legal determination about enforceability. Regulatory applicability and practical outcomes may vary by jurisdiction, enforcement context, and individual circumstances. Read our methodology
This provision establishes a data usage pathway distinct from primary service delivery, allowing the company to incorporate user feedback and conversational data into model development and product iteration processes. It creates a mechanism by which interactive user engagement with the service becomes training data for system improvement.
The updated terms state that Mistral AI Studio and APIs are now limited to business customers only. Previously, consumers outside the EEA could use APIs for personal needs and access Mistral AI Studio. Under the revised agreement, personal account holders retain access to Vibe and other products, but can no longer access APIs or Studio through personal accounts. Consumers who currently use these features will lose access unless they convert to a business account or transition to an alternative service.
View change record →The updated terms authorize Mistral AI to execute a broader range of actions on your behalf when you connect third-party services, including data modification, deletion, transmission, and invocation of external systems. Previously, the terms limited Mistral's authority to retrieve, process, and store data; the revised language now permits executing, modifying, deleting, invoking, and transmitting data through third-party services. Mistral's liability disclaimer has been expanded to explicitly exclude responsibility for data corruption, deletion, unauthorized disclosure, or other actions performed by third-party services on your request. You are now explicitly responsible for all actions you request Mistral AI Products to perform through third-party integrations, including how those third parties handle the data involved.
View change record →Removal of standalone feedback provision may indicate integration into broader data usage policies, eliminating explicit separate acknowledgment for feedback-based training.
View full change record →Clicking any feedback button on Le Chat or other Mistral AI Products creates a separate, unconditional grant allowing Mistral AI to use your full conversation — including your prompts — for AI model training, even if you have opted out of general training.
How other platforms handle this
your credit card or other payment information is collected and processed by our third-party payment processors...the payment information that you provide through the Services is transmitted directly to our payment processor.
If you do not timely cancel your subscription, your subscription will be renewed at the full price as indicated when the purchase was made, without any additional action by you, and you authorize us to charge your payment method for these amounts.
You also authorize us to retry any failed authorizations. We may use data provided to us by our partners to determine when to schedule such retries.
"For certain Mistral AI Products, such as Le Chat, you have the option to provide feedback, such as by selecting "thumbs up" or "thumbs down" ("Feedback"). By providing Feedback, you acknowledge and agree that Mistral AI may use your Feedback, along with the associated Input and Output data, for the purpose of training our artificial intelligence models and improving the Mistral AI Products.Excerpt from Mistral AI's Terms of Service
(1) REGULATORY FRAMEWORK: This provision implicates CCPA §1798.100 (right to know what personal information is collected and how it is used), FTC Act Section 5 (deceptive practices — users may not understand that a thumbs …
Enforcement risk, jurisdiction flags, contract triggers, and due diligence action items.
Get the research letter
Companies change their terms quietly. We read every version and catch what actually changed. One email a week on the changes that matter and what they mean.
This provision establishes a data usage pathway distinct from primary service delivery, allowing the company to incorporate user feedback and conversational data into model development and product iteration processes. It creates a mechanism by which interactive user engagement with the service becomes training data for system improvement.
Clicking any feedback button on Le Chat or other Mistral AI Products creates a separate, unconditional grant allowing Mistral AI to use your full conversation — including your prompts — for AI model training, even if you have opted out of general training.
ConductAtlas has identified this type of provision across 231 platforms. See the full comparison.
No. ConductAtlas is an independent monitoring service. We are not affiliated with, endorsed by, or sponsored by Mistral AI.