Usage Policy

Mistral AI's Usage Policy, published 2026-04-08, governs all products accessible on Mistral AI's own platform or applications — including Le Chat and Mistral AI Studio — covering all users, generated content, and interactions on those platforms. The policy explicitly does not apply to Mistral AI products deployed on customer or partner infrastructure, nor to open-source Mistral AI models and products. Three guiding principles underpin it: that technology should "be as neutral as possible," that users should "be empowered to define and use available and robust controls adapted to you and your priorities," and that Mistral AI commits to transparency about its models, policies, and enforcement.

Not disclosed in this document.

Not disclosed in this document.

Not disclosed in this document.

The policy defines ten prohibited categories: illegal activities (including weapons development, drug trafficking, exploitation through human trafficking or sexual services, and intellectual property infringement); CSAM; non-consensual intimate imagery; hate and discrimination (including revisionism or denialism of the Holocaust or other genocides); violence and threats; self-harm promotion; fraud and scams (phishing, impersonation, financial scams); deliberate misinformation (including content undermining civic or political processes); privacy violations such as using someone's likeness without consent; unqualified professional advice in financial, legal, or medical domains; and security violations. Mistral AI states a "zero tolerance policy regarding CSAM," with mandatory law enforcement reporting and immediate account termination for any generation or attempt. Users are prohibited from creating malware, exploiting vulnerabilities, or circumventing "security protections and AI safety filters." Enforcement ranges from temporary suspension to permanent account termination and, where appropriate, reporting to relevant authorities.

The policy applies to all Mistral AI products accessed through the company's own platform or applications, covering individuals, organizations, and businesses alike. It explicitly does not govern deployments on customer or partner infrastructure, nor open-source Mistral AI models and products — a boundary stated in the opening scope section. Violations may be reported through the Help Center or directly at legal@mistral.ai, and Mistral AI states it "will investigate promptly."

Not disclosed in this document.

The policy carries a date of 2026-04-08 and provides no version history or changelog comparing it to a prior policy version. Mistral AI reserves the right to update the policy at any time on its website and invites users to consult it frequently to stay informed of changes.