Mistral AI
AI API · mistral.ai · last read 2026-11-16
At a glance
- Is Mistral AI GDPR compliant?
- No tool is GDPR compliant on its own: compliance depends on what you process with it and how. What follows is what Mistral AI itself publishes, each answer linked to its source and the date we read it.
- Does Mistral AI have a data processing agreement (DPA)?
- Yes. Mistral AI publishes a data processing agreement at legal.mistral.ai. It describes how a new subprocessor is announced but states no notice period. It promises breach notification "without undue delay", with no number of hours. Source and date
- Which version of the Mistral AI DPA is current?
- On 2026-11-09, Mistral AI's DPA at legal.mistral.ai printed “Effective: July 27, 2026”. Source and date
- Does Mistral AI have a SOC 2 report?
- Mistral AI says a SOC 2 Type 2 report is available on request. In Mistral AI's words: “SOC 2 Type II - Letter of Attestation (Mistral AI).” We read this on 2026-10-05. Source and date
- Is Mistral AI ISO 27001 certified?
- Mistral AI states ISO/IEC 27001 certification (2022 edition, as Mistral AI names it), and says the certificate or audit report is available on request. In Mistral AI's words: “ISO/IEC 27001:2022 Certificate (Mistral AI).” We read this on 2026-10-05. Source and date
- Where is Mistral AI's trust center?
- Mistral AI publishes a page it calls “Trust Center” at trust.mistral.ai. Some of its documents sit behind a request or sign-in: the page says “Request access”. Its text appears only after JavaScript runs: a client without it, as most tooling is, gets an empty page. The page mentions SOC 2, ISO 27001, a penetration test and subprocessors. We read it on 2026-11-03. Source and date
- Does Mistral AI use your data to train AI models?
- For consumer plans (Vibe): Input and output used for training by default; user can opt out at any time. Feedback (thumbs up/down with comment) is used to improve models on every plan. For business plans (Vibe (Enterprise)): Opted out of training by default; an admin can opt in. Feedback (thumbs up/down with comment) is used to improve models on every plan. For the API (Mistral Studio (Free mode)): Input and output may be used for training; user can opt out at any time. For the API (Mistral Studio (Pay-as-you-go)): Customers can opt out of this processing at any time. Source and date
- What is Mistral AI's data retention for AI prompts and outputs?
- For every plan (Vibe): Conversations are kept until the user deletes them or deletes the account, typically followed by a short grace period. Source and date
- Does Mistral AI offer zero data retention?
- For every plan (Vibe Work and Chat): ZDR not available for Vibe Work or Chat on any subscription plan. For the API (Mistral Studio (API, pay-as-you-go)): ZDR only with pay-as-you-go and only for stateless API endpoints; you request it and Mistral approves or denies it at its discretion. Source and date
- Can Mistral AI process AI data in the EU?
- For every plan (Mistral (Vibe and APIs)): Data hosted in the EU by default; US hosting if the US API endpoint is used; some features may temporarily transfer data outside the EU (Enterprise can deactivate some of these features). Source and date
- How does Mistral AI transfer personal data out of the EU?
- Mistral AI's agreement relies on the EU standard contractual clauses. We found no entry for Mistral AI on the official Data Privacy Framework List. Source and date
- Does Mistral AI offer EU data residency?
- The EU is the default region for this vendor. Source and date
- Who are Mistral AI's subprocessors?
- Mistral AI lists 23 third-party subprocessors; 8 of them name at least one processing country outside the EEA. We last read the list on 2026-10-05. Source and date
- How do you report a security vulnerability to Mistral AI?
- Mistral AI publishes a security.txt at mistral.ai naming security@mistral.ai. Source and date
The same questions for any vendor, and the rule behind each: the vendor due diligence checklist.
Identity
| Legal entity | Mistral AI |
|---|---|
| HQ | FR |
| Parent company | — |
| Ultimate ownership | FR |
| EU contracting entity | — |
| LEI | 894500Q9JNDV9YXSOL64 |
| US CLOUD Act reach | No US company in the ownership chain on record |
| Source | legal.mistral.ai · 2026-10-08 |
Legal entity from the vendor's own terms. LEI, registered country and parent company from the Global LEI Index (GLEIF); ownership is left empty where GLEIF names no parent. The CLOUD Act row is arithmetic over those countries, not an assessment of risk.
Mistral AI data residency
| Region | EU data location | Minimum plan | Source | Observed |
|---|---|---|---|---|
| EU | EU by default | — | help.mistral.ai | 2026-10-05 |
EU: The document states that by default, data is hosted in the European Union. Only Enterprise customers can switch off some features that move data outside the EU, but EU hosting itself is not tied to any plan or cost.
The same question for every vendor: where SaaS vendors store EU data.
Mistral AI's subprocessor list names Microsoft and Google for hosting or infrastructure. Which SaaS vendors run on which cloud, and the sovereign offerings of each: sovereign clouds.
Subprocessor chain
23 third parties · 2 of the vendor's own group companies · 8 with a country outside the EEA · all read 2026-10-05
| Subprocessor | Country | Purpose |
|---|---|---|
| Backblaze Inc. | NL | Cloud infrastructure |
| Blackforest Labs Inc. | US | Image Generation |
| Brave Software, Inc. | US | Web Search |
| Cloudflare Inc. | — | Security & Traffic Routing (CDN) |
| CoreWeave Inc. | EU | Inference Provider |
| CrowdStrike Inc. | EU, US | Security |
| Foundrylabs, Inc. | FR, BE, NL | Code Interpreter (E2B.dev) |
| Functional Software, Inc. (Sentry) | EU | Error handling |
| Get Lago Corp. | IE | Billing |
| Google LLC | NL, BE, US | Cloud Infrastructure |
| Gradual Inc. | US | Mistral Partner Academy |
| Intercom Inc. | IE | User support |
| Kong Inc. | EU | Security (API) |
| Megaport Inc. | — | Cloud infrastructure (Latitude.sh) |
| Microsoft Inc. | SE, NO | Cloud Infrastructure |
| Ory Corp. | BE, DE | User Authentication |
| OVH Cloud SAS | DE, US | Cloud Infrastructure |
| Plus Five Five, Inc. (Resend) | IE | Email communications |
| Scaleway SAS | FR | Cloud Infrastructure |
| Stripe Inc. | US | Payment and Billing |
| The Constant Company, LLC | — | Cloud Infrastructure (Vultr) |
| Twilio Inc. | US | Phone number verification |
| Wiz Inc. | EU | Security |
Mistral AI's own group companies listed as subprocessors (2)
| Company | Country | Purpose |
|---|---|---|
| Mistral AI Affiliates | — | — |
| Mistral Compute | FR | Cloud infrastructure |
The count outside the EEA is arithmetic over the country column, not a judgement about lawfulness. A transfer outside the EEA can be perfectly lawful; whether this one is depends on your use, not on the vendor.
Every subprocessor in the registry, and which vendors name it: subprocessors.
Processing terms
From the vendor's data processing agreement · read 2026-10-09 · legal.mistral.ai
| Term | What the agreement says |
|---|---|
| Transfer mechanism | SCC module 4 |
| Notice of new subprocessors | Automatic email notification of any addition or replacement of a Subprocessor, for customers who subscribe to updates on the Trust Center Subprocessor page |
| Notice of new subprocessors | Reasonable notice of changes to the Subprocessor list before engaging the new Subprocessor; customers subscribed to the Trust Center Subprocessor page get automatic email notification |
| If you object to one | Written objection to privacy@mistral.ai within 10 days of notice, on reasonable data-protection grounds (otherwise deemed approved); parties negotiate in good faith; if no resolution, Mistral AI may terminate the Agreement or the affected products |
| Obligations passed to subprocessors | Written agreement with each Subprocessor requiring protection of Personal Data to the same standards as the DPA |
| Breach notification | without undue delay after becoming aware |
| Deletion when the service ends | Delete or return all Personal Data after the end of the provision of the Mistral AI Products, under Mistral AI's deletion policies and procedures; the data is no longer accessible once a 30-day period after termination of the Customer's access ends |
| Getting your data back | Mistral AI will delete or return all Personal Data to Customer after the end of the provision of the Mistral AI Products; the data is no longer accessible once a 30-day period after termination of access ends |
| Requests from data subjects | commercially reasonable assistance on Customer's request; requests sent directly to Mistral AI are transferred to Customer and not answered without Customer's prior consent unless required by law |
| Audit | Customer audit: customer bears all audit costs; audit must not impair Mistral AI products or operations; Mistral AI helps where commercially reasonable; both parties get the report; findings are Mistral AI's confidential information |
| Audit | Documents and information on written request; if those are insufficient, one on-site audit per year with at least 90 days' written notice, by a jointly selected independent auditor, at the customer's sole cost |
| Processing on your instructions | only on Customer's documented lawful instructions, including as set out in the DPA or Agreement, unless required by law (Customer informed promptly unless prohibited) |
| Confidentiality of staff | all authorised persons, including team members and Subprocessors, subject to a duty of confidentiality |
| Security measures | Technical and organizational measures listed in Exhibit 2 of the DPA |
Terms as the agreement states them. Where it is silent on a term, there is no row; that is not a statement that the vendor does not offer it.
Which version applies
On 2026-11-09, the agreement printed: Effective: July 27, 2026
at legal.mistral.ai
In effect from 2026-07-27
This version is in our copies from 2026-10-09.
These terms side by side for every vendor: data processing agreements compared. What vendors give you for a transfer impact assessment: transfer impact assessments.
AI and your data
What Mistral AI's own pages state about the data you send its AI products, plan by plan. Compared with other vendors on AI vendors and your data.
| Plans | Term | What the page says | Source | Observed |
|---|---|---|---|---|
| All plans Vibe | Data retention | Until deleted Conversations are kept until the user deletes them or deletes the account, typically followed by a short grace period | help.mistral.ai | 2026-10-28 |
| All plans Vibe Work and Chat | Zero data retention | Not offered ZDR not available for Vibe Work or Chat on any subscription plan | help.mistral.ai | 2026-10-28 |
| All plans Mistral (Vibe and APIs) | EU processing | EU storage only Data hosted in the EU by default; US hosting if the US API endpoint is used; some features may temporarily transfer data outside the EU (Enterprise can deactivate some of these features) | help.mistral.ai | 2026-10-28 |
| Consumer Vibe | Training on your data | Used unless you opt out Input and output used for training by default; user can opt out at any time. Feedback (thumbs up/down with comment) is used to improve models on every plan | help.mistral.ai | 2026-10-28 |
| Business Vibe (Enterprise) | Training on your data | Not used unless you opt in Opted out of training by default; an admin can opt in. Feedback (thumbs up/down with comment) is used to improve models on every plan | help.mistral.ai | 2026-10-28 |
| API Mistral Studio (Free mode) | Training on your data | Used unless you opt out Input and output may be used for training; user can opt out at any time | help.mistral.ai | 2026-10-28 |
| API Mistral Studio (Pay-as-you-go) | Training on your data | Used unless you opt out Customers can opt out of this processing at any time | help.mistral.ai | 2026-10-28 |
| API Mistral Studio (API, pay-as-you-go) | Zero data retention | On approval ZDR only with pay-as-you-go and only for stateless API endpoints; you request it and Mistral approves or denies it at its discretion | help.mistral.ai | 2026-10-28 |
Assurance
Does Mistral AI have a SOC 2 report?
Mistral AI says a SOC 2 Type 2 report is available on request. We read this on 2026-10-05.
- Available on request “SOC 2 Type II - Letter of Attestation (Mistral AI)”
Source: trust.mistral.ai, 2026-10-05
Of the 114 vendors in this registry whose own pages mention SOC 2, 57 say the report is available on request, 1 links to the report itself, and 56 state it without saying how to get it. The type is the one the vendor names.
Is Mistral AI ISO 27001 certified?
Mistral AI states ISO/IEC 27001 certification (2022 edition, as Mistral AI names it), and says the certificate or audit report is available on request. We read this on 2026-10-05.
- Available on request “ISO/IEC 27001:2022 Certificate (Mistral AI)”
Source: trust.mistral.ai, 2026-10-05
On the same list Mistral AI also names ISO/IEC 27701. Each standard's page says how it relates to ISO/IEC 27001.
Of the 122 vendors in this registry whose own pages state ISO/IEC 27001, 16 link to the certificate, 43 say it is available on request, and 63 state it without saying how to get it. A certificate names the certification body, the scope and the expiry date: those are the three things to check.
| Scheme | Status | Scope | Source | Observed |
|---|---|---|---|---|
| ISO/IEC 27001:2022 · ISO27001 | Available on request | Mistral AI | trust.mistral.ai | 2026-10-05 |
| ISO/IEC 27701:2019 · ISO27701 | Available on request | Mistral AI | trust.mistral.ai | 2026-10-05 |
| SOC 2 Type II · SOC2T2 | Available on request | Mistral AI | trust.mistral.ai | 2026-10-05 |
Claimed means the vendor asserts the scheme on its own page and offers nothing further. Available on request means the report or certificate exists but sits behind a request, a login or an NDA. Neither is a certificate we have seen. That is a distinction an auditor makes, so we make it too.
SecNumCloud, C5, ENS and HDS for every vendor, with the pages for ISO standards, SOC 2 and TISAX: certifications compared.
Official registers
| Register | Offering, as the register names it | Status | Level | Valid until | Observed |
|---|---|---|---|---|---|
| Global LEI Index (GLEIF) | MISTRAL AI · MISTRAL AI | ISSUED | FR | 2027-01-15 | 2026-10-08 |
| GPAI Code of Practice signatories (European Commission) | General-Purpose AI Code of Practice (AI Act), as a provider of general-purpose AI models · Mistral AI | signatory | — | — | 2026-11-09 |
- ACN qualified cloud (Italy): checked 2026-11-16, no entry under Mistral AI
- ISO 27001 on the basis of IT-Grundschutz (BSI, Germany): checked 2026-11-16, no entry under Mistral AI
- CISPE Data Protection Code of Conduct (GDPR Art. 40): checked 2026-11-16, no entry under Mistral AI
- DORA critical ICT third-party providers (ESAs): checked 2026-11-16, no entry under Mistral AI
- Data Privacy Framework List (US): checked 2026-11-16, no entry under Mistral AI
- ENS certificates of conformity (CCN, Spain): checked 2026-11-16, no entry under Mistral AI
- EU Cloud Code of Conduct (GDPR Art. 40): checked 2026-11-16, no entry under Mistral AI
- FedRAMP Marketplace (US): checked 2026-11-16, no entry under Mistral AI
- GPAI Code of Practice signatories (European Commission): checked 2026-11-16, 1 entry under Mistral AI
- SecNumCloud qualified services (ANSSI, France): checked 2026-11-16, no entry under Mistral AI
What the authority publishes, not what the vendor claims. An entry covers the offering it names and nothing else: a FedRAMP authorisation for a government edition says nothing about the commercial service. No entry is not a finding about security - most vendors never apply - and an entry filed under a name not listed above would not have been found.
What an entry on the Data Privacy Framework List covers, and which vendors hold one: EU-US Data Privacy Framework.
Support window
Not yet collected. This row will say not declared only once we have read a lifecycle or security page and found no period stated — see support windows.
Vulnerability reporting: security@mistral.ai (security.txt, expires 2027-05-05; read 2026-10-17)
How other vendors publish theirs: security.txt across the registry.
Documents
Where is Mistral AI's trust center?
Mistral AI publishes a page it calls “Trust Center” at trust.mistral.ai. Some of its documents sit behind a request or sign-in. Its text appears only after JavaScript runs: a client without it, as most tooling is, gets an empty page. The page mentions SOC 2, ISO 27001, a penetration test and subprocessors. We read it on 2026-11-03. Open trust.mistral.ai
- Name “Welcome to Mistral AI's Trust Center!”
- Request step “Request access”
Source: trust.mistral.ai, 2026-11-03. Read from the page's text, without a model; every vendor's trust center.
| Document | Disclosure | Transcription | HTTP | Observed |
|---|---|---|---|---|
| ai | Public | full | 200 | 2026-10-28 |
| ai-2 | Public | full | 200 | 2026-10-28 |
| ai-3 | Public | full | 200 | 2026-10-28 |
| certifications | Public | full | 200 | 2026-10-05 |
| dpa | Public | full | 200 | 2026-10-09 |
| residency | Public | full | 200 | 2026-10-05 |
| subprocessors | JavaScript only | full | 200 | 2026-10-05 |
History
Nothing has moved since we started watching this vendor on 2026-10-28. Changes appear here only after a human has reviewed them.
Every reviewed change across the registry, newest first: vendor changes.