Policy Scope Exclusion for Non-Platform Deployments

What it is

Mistral AI's usage rules only cover products you access directly through Mistral AI's own platform; if a business built their own app using Mistral AI's technology or an open-source model, this policy does not apply to that product.

Why it matters (compliance & governance perspective)

Users interacting with AI products built on Mistral AI's models outside the company's own platform are not protected or governed by this policy, and should look to the deploying organization's own policies instead.

Consumer impact (what this means for users)

If you are using an AI product from a third party that is powered by Mistral AI's technology, the protections and restrictions in this policy do not apply to your interaction, which means there may be no equivalent governance framework unless the third party has established one independently.

Institutional analysis (Compliance & governance intelligence)

(1) REGULATORY LANDSCAPE: This scope limitation is relevant to EU AI Act obligations, which distinguish between AI developers, deployers, and distributors; organizations deploying Mistral AI models on their own infrastructure become responsible for compliance with applicable AI regulations in their deployment context. The carve-out for open-source models engages the EU AI Act's treatment of open-source AI, where reduced obligations may apply under certain conditions. (2) GOVERNANCE EXPOSURE: Medium. For enterprise customers deploying Mistral AI on their own infrastructure, this clause confirms the absence of Mistral AI's usage governance framework, which means the enterprise must independently establish and enforce an acceptable use policy governing that deployment. (3) JURISDICTION FLAGS: In the EU, deployers of AI systems bear independent obligations under the EU AI Act regardless of whether an upstream provider's acceptable use policy applies; this scope exclusion does not reduce those obligations. In the U.S., the absence of a governing acceptable use policy for infrastructure deployments may affect how the FTC evaluates unfair or deceptive practices claims against third-party deployers. (4) CONTRACT AND VENDOR IMPLICATIONS: Procurement teams should note that the Mistral AI commercial terms of service, data processing addendum, and additional product terms (referenced in the site navigation) govern customer-infrastructure deployments separately; the absence of this Usage Policy's coverage does not mean there is no contractual framework, but that a different framework applies. Due diligence should confirm which governance documents apply to the specific deployment model. (5) COMPLIANCE CONSIDERATIONS: Organizations deploying Mistral AI on customer infrastructure should establish their own acceptable use policies and ensure they address the categories covered in this document (CSAM, professional advice, misinformation, etc.) as part of their AI governance framework, since this policy will not cover those deployments.

Provision details

Other risks in this policy

• CSAM Zero Tolerance and Mandatory Reporting high
• Professional Advice Prohibition medium
• Security Violations and AI Safety Filter Circumvention medium
• Unilateral Policy Update Right low
• Enforcement Consequences: Suspension and Termination medium
• Privacy Violation Prohibition medium