Mistral AI generally does not use your business data to train its AI, but there are five exceptions, including when you use Labs Models, submit Feedback, or use free or certain paid subscriptions without opting out. For those exceptions, you grant a permanent, irrevocable license to use your data for training.
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
The license granted for training purposes under the listed exceptions is perpetual and irrevocable, meaning that once data falls within one of these categories, Mistral AI's right to use it for training cannot be withdrawn even if the Customer later terminates the agreement.
Commercial customers whose data is submitted through Labs Models, flagged by moderation, or associated with Feedback may have that Customer Data and Output used for AI model training under a license that cannot be revoked, affecting organizations with confidentiality or data minimization obligations.
How other platforms handle this
If you would like to opt out of the disclosure of your personal information for purposes that could be considered "sales" for those third parties' own commercial purposes, or "sharing" or processing for purposes of targeted advertising, please visit the following link, which is also available in the...
California law gives residents the right to know what personal information we collect, use, share or sell; to delete personal information under certain circumstances; to opt-out of the sale or sharing of their personal information; to correct inaccurate personal information; to limit the use and dis...
T-Mobile collects Customer Proprietary Network Information (CPNI), which is information about the quantity, technical configuration, type, destination, location, and amount of use of your service. T-Mobile may use your CPNI within its family of companies for the purpose of providing wireless telecom...
Monitoring
Mistral AI has changed this document before.
Receive same-day alerts, structured change summaries, and monitoring for up to 10 platforms.
"Mistral AI will not use Customer Data or Outputs to train its artificial intelligence models except (a) for Customer Data or Outputs used or generated in connection with the use Mistral AI Products under a free subscription, Le Chat Pro or Le Chat Teams, where Customer has not opted-out of training, (b) when Customer or an End User provides Feedback to Mistral AI, (c) when Customer Data our Outputs are flagged as part of Mistral AI's automated moderation or reported by a user under the Additional Terms, (d) as otherwise may be provided in an Order Form or (e) when Customer uses Labs Models. Customer grants Mistral AI a perpetual, irrevocable, worldwide, non-exclusive, non-transferable (except as permitted in Section 14.2 (Assignment)), royalty-free, fully-paid license (with the right to sublicense to our service providers) to use Customer Data and Outputs solely as provided in the preceding sentence to train Mistral AI's artificial intelligence models.— Excerpt from Mistral AI's Mistral AI Commercial Terms
(1) REGULATORY LANDSCAPE: This provision engages GDPR, particularly regarding lawful basis for processing personal data for AI training purposes, and the requirement for specific, informed consent where consent is relied upon. The CNIL and EU supervisory authorities are the primary enforcement bodies. The perpetual, irrevocable nature of the training license may create tension with GDPR rights to erasure and data portability if personal data is embedded in Customer Data used for training. The EU AI Act's requirements around transparency and data governance for AI system providers are also engaged. (2) GOVERNANCE EXPOSURE: High. The five-category carve-out structure means that organizations must actively manage their product usage patterns, opt-out configurations, and Feedback workflows to avoid unintended training data contributions. The moderation flagging carve-out is particularly noteworthy because it is triggered by Mistral AI's automated systems rather than Customer action, and the resulting training license is irrevocable. (3) JURISDICTION FLAGS: EU/EEA organizations face heightened exposure given GDPR's strict requirements on processing personal data for AI training and data subject rights. California-based customers should evaluate CCPA implications if personal data of California residents is included in Customer Data that falls within a training carve-out. Organizations in regulated sectors (healthcare, financial services, legal) should assess whether Customer Data submitted to the platform may contain regulated information that triggers additional compliance obligations. (4) CONTRACT AND VENDOR IMPLICATIONS: Procurement teams should verify whether their Order Forms include provisions that expand or restrict the default training carve-outs, particularly carve-out (d). The perpetual, irrevocable license grant means standard contract termination and data deletion provisions will not fully unwind this right, which may affect vendor risk assessments and data processing agreements with downstream clients. (5) COMPLIANCE CONSIDERATIONS: Compliance teams should map all Mistral AI product usage against the five training carve-out categories, implement controls to prevent personal or confidential data from being submitted through channels that trigger training licenses, review whether End User consents obtained by the Customer are sufficient to cover the training use cases, and assess whether the Data Processing Addendum addresses the scope of training-related processing.
Full compliance analysis
Regulatory citations, enforcement risk, and due diligence action items.
Free: track 1 platform + weekly digest. Watcher: 10 platforms + same-day alerts. No credit card required.
Professional Governance Intelligence
Need to monitor specific governance provisions?
Professional includes provision-level monitoring, governance timelines, regulatory mapping, and audit-ready analysis.
Built from archived source documents, structured governance mappings, and historical version tracking.
The license granted for training purposes under the listed exceptions is perpetual and irrevocable, meaning that once data falls within one of these categories, Mistral AI's right to use it for training cannot be withdrawn even if the Customer later terminates the agreement.
Commercial customers whose data is submitted through Labs Models, flagged by moderation, or associated with Feedback may have that Customer Data and Output used for AI model training under a license that cannot be revoked, affecting organizations with confidentiality or data minimization obligations.
No. ConductAtlas is an independent monitoring service. We are not affiliated with, endorsed by, or sponsored by Mistral AI.