A French artificial intelligence company that develops and provides large language models and AI services through its platform. The company offers various AI models that users can access via API or chat interfaces for text generation, analysis, and other language-related tasks. Their policies are relevant to consumers as they govern data handling practices, usage restrictions, and terms for accessing AI services that may process personal or sensitive information.
Higher = more protective of you, the user/customer — a deliberately user-protective lens, not a universal “good vs bad” score. Computed algorithmically from Mistral AI’s cited governance stances; no dimension we haven’t confidently read is ever counted against the score. Free, reproducible, and never purchasable.
The license grants Mistral AI and its delegates and subcontractors broad rights to use user data in connection with its products, without additional compensation to the user.
Mistral AI has a contractual license to use Customer Data and Outputs for operational product purposes, but the explicit exclusion of model training limits that use and provides the Customer a define…
The prohibition extends beyond a user's own illegal conduct to cover encouraging, promoting, offering, or soliciting illegal activity by others, making the scope of restricted use broad.
The cap limits the maximum financial recovery a user can obtain from Mistral AI regardless of the nature or severity of the claim.
The Customer has an affirmative right to be defended and compensated by Mistral AI if its use of the products exposes it to third-party intellectual property claims.
This document describes Mistral AI's Mistral Medium 3.5 model: what it can do, how to access it, and what it costs. The model weights are publicly available under a Modified …
This document is a data processing agreement that sets out how Mistral AI handles personal data it processes on a business customer's behalf. Mistral AI is only permitted to use …
These terms set out the rules Customers must follow when using Mistral AI's products, including audio tools and web-search features. You are responsible for reporting serious problems within 48 hours, …
This document sets out the rules for businesses that pay to use Mistral AI's products. Every payment is final—no refunds, cancellations, or credits—and if something goes wrong, the most either …
This is the rules-of-use document for Mistral AI Products, specifying what users are and are not allowed to do. Key prohibitions include generating illegal, fraudulent, deceptive, hateful, or rights-infringing content, …
These terms govern your use of Mistral AI's products and set the rules for billing, data use, and disputes. If you are on a free, Vibe Pro, or Vibe Student …
This document establishes Mistral AI's data processing practices for personal data collected through its products, including Le Chat and Mistral AI Studio. The policy authorizes the use of chat prompts …
Mistral AI revised its Commercial Terms effective September 25, 2026, narrowing the circumstances under which customers can object to updates. Previously, customers could object to any material update within 30 …
View change record →Mistral AI revised its Commercial Terms effective August 5, 2026, making three material changes to how customer data is handled. First, the definition of 'Customer Data' now explicitly includes 'access …
View change record →Mistral AI revised its consumer Terms of Service effective August 5, 2026, making several operational changes to its product scope and API access rules. The company removed language stating that …
View change record →Mistral AI updated its Terms of Service effective May 28, 2026, with changes affecting how third-party service integrations work and which products are covered. The company replaced references to 'Le …
View change record →Get the weekly coverage letter
Companies change their terms without announcing it. We capture governance documents, compare each new version with the last, and publish what changed, with the source text behind it. One email a week on what we found and what we could not reach.
ConductAtlas tracks 8 Mistral AI documents including terms of service, privacy policy, and other governance documents. Every document is captured daily with cryptographic verification.
Mistral AI has made 9 policy changes in the past 12 months across the documents ConductAtlas tracks, including 1 classified as high severity.
ConductAtlas has classified 40 provisions across Mistral AI's tracked documents.
Yes. Monitor subscribers ($4.99/month) can add Mistral AI to their watchlist and receive an email alert after the daily check that finds a change to a tracked document. We check daily where we can, and show you where we can't.