{"investigation":{"slug":"mistral-ai","entity_name":"Mistral AI","trust_score":62,"severity_base":null,"score_modifier":0,"confidence":0.7,"status":"draft","content_type":"investigation","summary":"Mistral AI is a Paris-based artificial intelligence company, founded in April 2023 by former Google DeepMind and Meta researchers, known for its open-weight and commercial large language models and the Le Chat consumer chatbot. The company has grown rapidly with multibillion-euro funding rounds, but has also faced a data-privacy complaint in France, EU antitrust scrutiny over its Microsoft partnership, third-party research alleging serious safety failures in some multimodal models, a May 2026 supply-chain attack on its official Python package, and unverified dark-web claims of a source-code breach. There is no evidence of Mistral AI itself perpetrating fraud; adverse findings largely concern regulatory/privacy exposure, AI-safety benchmarking, and security incidents typical of a fast-growing AI infrastructure company. Separately, its name has reportedly been used by unrelated parties in domain-squatting and, per general crypto-scam patterns, could plausibly be exploited in impersonation-style token/airdrop scams, though no specific verified case naming Mistral AI was found.","sections":[{"content":"Mistral AI SAS was founded in Paris in April 2023 by Arthur Mensch (CEO, formerly of Google DeepMind), Guillaume Lample (Chief Science Officer, formerly of Meta), and Timothée Lacroix (CTO, formerly of Meta); the three met as students at École Polytechnique. The company has raised capital in several large rounds: a reported €2 billion investment around September 2025 valuing the company near €12 billion, following a $1.5 billion investment from ASML (which took an approximately 11% stake); an $830 million raise in March 2026 to fund datacenters near Paris and in Sweden; and, as reported by CNBC and Bloomberg, a further funding round led by Samsung announced around September 8, 2026, valuing the company at more than $23-24 billion. Reported 2026 annualized revenue was approximately $400 million, up from about $100 million in 2025. These are standard venture financings; no fraud or misrepresentation allegations regarding the funding process itself were found.","heading":"Corporate Structure and Funding","sources":[{"url":"https://www.cnbc.com/2026/09/08/mistral-ai-funding-valuation-samsung.html","name":"Mistral bags $24 billion valuation as Samsung leads funding (CNBC)","type":"news_article","credibility":1},{"url":"https://www.bloomberg.com/news/articles/2026-06-12/france-s-mistral-in-funding-talks-at-about-20-billion-valuation","name":"France's Mistral in Funding Talks at About €20 Billion Valuation (Bloomberg)","type":"news_article","credibility":1},{"url":"https://en.wikipedia.org/wiki/Mistral_AI","name":"Mistral AI (Wikipedia)","type":"other","credibility":2},{"url":"https://techcrunch.com/2026/07/04/what-is-mistral-ai-everything-to-know-about-the-openai-competitor/","name":"What is Mistral AI? Everything to know about the OpenAI competitor (TechCrunch)","type":"news_article","credibility":1}],"severity":"low"},{"content":"In February 2024, Microsoft announced a roughly €15 million ($16 million) investment in Mistral AI alongside a deal to distribute Mistral's models via Microsoft Azure, prompting concerns from EU lawmakers and the European Commission about concentration of power among large AI platform providers; the Commission stated it would analyze the agreements. The UK Competition and Markets Authority opened and then, within about a day, closed a merger-control inquiry in February 2024, concluding that Microsoft's investment did not confer sufficient rights or influence to qualify for review. Green Members of the European Parliament separately asked the Commission to examine whether Mistral was being used as a vehicle for Microsoft's lobbying interests, given that Mistral was negotiating the Microsoft partnership while simultaneously lobbying to soften provisions of the EU AI Act. As of the most recent reporting found (mid-2026), no concluded outcome of the European Commission's review was identified in public sources; Microsoft and Mistral expanded their partnership further in July 2026 with a multibillion-dollar agreement to support European AI infrastructure.","heading":"Microsoft Investment and Antitrust Scrutiny","sources":[{"url":"https://www.bloomberg.com/news/articles/2024-02-27/microsoft-s-mistral-ai-investment-to-be-examined-by-eu-watchdog","name":"Microsoft's Mistral AI Investment to Be Examined by EU Watchdog (Bloomberg)","type":"news_article","credibility":1},{"url":"https://techcrunch.com/2024/05/17/microsoft-dodges-uk-antitrust-scrutiny-over-its-mistral-ai-stake/","name":"Microsoft dodges UK antitrust scrutiny over its Mistral AI stake (TechCrunch)","type":"news_article","credibility":1},{"url":"https://aibusiness.com/companies/antitrust-regulator-drops-probe-into-microsoft-s-mistral-deal","name":"Antitrust Regulator Drops Probe Into Microsoft's Mistral Deal","type":"news_article","credibility":2},{"url":"https://www.mlex.com/mlex/articles/2191760","name":"Mistral AI, Microsoft talks overlapped with lobbying EU AI Act for looser rules (MLex)","type":"news_article","credibility":2},{"url":"https://news.microsoft.com/source/2026/07/21/microsoft-and-mistral-expand-strategic-partnership-to-give-enterprises-and-regulated-industries-frontier-ai-they-can-control/","name":"Microsoft and Mistral expand strategic partnership (Microsoft Source)","type":"official","credibility":2}],"severity":"medium"},{"content":"Mistral AI, through lobbyist Cédric O (France's former state secretary for digital), was reported to have led opposition, alongside other European AI startups, to the European Parliament's proposal for tiered regulation of foundation models during the negotiation of the EU AI Act. Corporate Europe Observatory characterized this as European startups being used as a \"Trojan horse\" to weaken the Act on behalf of larger technology interests. In July 2025, Mistral joined 55 other EU-based AI companies in a public letter urging the European Commission to pause and simplify parts of the AI Act. These are lobbying activities disclosed and reported by journalists and advocacy groups; they are not illegal, but are noted here as a governance/transparency concern raised by MEPs and watchdog organizations.","heading":"EU AI Act Lobbying","sources":[{"url":"https://corporateeurope.org/en/2024/03/trojan-horses-how-european-startups-teamed-big-tech-gut-ai-act","name":"Trojan horses: how European startups teamed up with Big Tech to gut the AI Act (Corporate Europe Observatory)","type":"research","credibility":2},{"url":"https://venturebeat.com/ai/the-eu-ai-act-passed-but-mistral-parties-on-the-ai-beat","name":"The EU AI Act passed, but Mistral parties on (VentureBeat)","type":"news_article","credibility":2},{"url":"https://www.mlex.com/mlex/articles/2191760","name":"Mistral AI, Microsoft talks overlapped with lobbying EU AI Act (MLex)","type":"news_article","credibility":2}],"severity":"low"},{"content":"In February 2025, a French lawyer specializing in data protection law, Jérémy Roche, filed a complaint with France's data protection authority (CNIL) alleging that Mistral AI's free-tier Le Chat chatbot used personal data to train AI models without offering free users a way to opt out, while paying subscribers had that option. Mistral AI disputed the characterization and subsequently updated its terms of use, stating that free users could object to data processing by emailing privacy@mistral.ai. As of available reporting, the CNIL had not issued a public decision; potential outcomes for non-compliance under GDPR could range from a warning to a fine.","heading":"Data Privacy Complaint (CNIL)","sources":[{"url":"https://oecd.ai/en/incidents/2025-02-11-f208","name":"Mistral AI Faces Data Privacy Controversy Over Chatbot 'Le Chat' (OECD.AI)","type":"research","credibility":2},{"url":"https://www.usine-digitale.fr/article/mistral-ai-vise-par-une-plainte-pour-violation-des-donnees-personnelles.N2227401","name":"Mistral AI visé par une plainte pour violation des données personnelles (Usine Digitale)","type":"news_article","credibility":2},{"url":"https://wancore.fr/news/mistral-ai-dans-le-viseur-de-la-cnil/","name":"Mistral AI dans le viseur de la CNIL (Wancore)","type":"news_article","credibility":2},{"url":"https://legal.mistral.ai/terms/privacy-policy","name":"Mistral AI's Privacy Policy","type":"official","credibility":2}],"severity":"medium"},{"content":"A May 2025 report by AI-security research firm Enkrypt AI found that Mistral's Pixtral-Large (25.02) and Pixtral-12B vision-language models generated child sexual abuse material (CSAM)-adjacent content at a rate the researchers described as roughly 60 times higher than comparable models such as GPT-4o and Claude 3.7 Sonnet when tested with 500 CSAM-themed adversarial prompts, and produced chemical, biological, radiological and nuclear (CBRN) weapons information up to approximately 40 times more often than competitors across 200 CBRN-targeted prompts, with about two-thirds of harmful prompts successfully eliciting unsafe output using image-based \"typographic\" jailbreak techniques. These findings concern specific multimodal models and testing methodology from a single third-party research firm; Mistral's own response to the report was not confirmed in available sources. Separately, Mistral acknowledged upon release of its original Mistral 7B Instruct model that it lacked built-in moderation mechanisms and stated an intent to work with the community on guardrails. Other reporting (OECD.AI, April 2026) alleged that Le Chat had amplified state-sponsored disinformation in at least one documented incident.","heading":"AI Safety and Content Moderation Concerns","sources":[{"url":"https://www.bankinfosecurity.com/mistral-ai-models-fail-key-safety-tests-report-finds-a-28358","name":"Mistral AI Models Fail Key Safety Tests, Report Finds (BankInfoSecurity)","type":"research","credibility":2},{"url":"https://oecd.ai/en/incidents/2026-04-28-e382","name":"French AI Chatbot Mistral Amplifies State-Sponsored Disinformation (OECD.AI)","type":"research","credibility":2},{"url":"https://www.cryptopolitan.com/french-ai-startup-mistral-faces-backlash/","name":"French AI Startup Mistral Faces Backlash as New LLM Generates Harmful Content (Cryptopolitan)","type":"news_article","credibility":3}],"severity":"high"},{"content":"On May 11-12, 2026, Mistral AI's official \"mistralai\" Python package on PyPI (including its core SDK and Azure/GCP integration variants) was compromised as part of a broad supply-chain campaign researchers dubbed \"Mini Shai-Hulud,\" which affected over 170 packages and published more than 400 malicious versions across npm and PyPI within a roughly five-hour window. Microsoft Threat Intelligence identified that compromised version 2.4.6 of the mistralai package contained code that silently downloaded and executed a second-stage payload (a credential stealer, with a destructive branch capable of executing a filesystem-wiping command under certain geographic conditions) on Linux systems that imported the package. The malicious package versions were available for download for approximately three hours before removal. This was an attack on the software supply chain affecting Mistral's published package, rather than an intrusion into Mistral's own internal systems; the incident has been assigned CVE-2026-45321 (CVSS 9.6) as the campaign's entry point.","heading":"Supply Chain Attack on Official Package (May 2026)","sources":[{"url":"https://www.tomshardware.com/tech-industry/cyber-security/compromised-mistral-ai-and-tanstack-packages-may-have-exposed-github-cloud-and-ci-cd-credentials-in-mini-shai-hulud-malware-infection-supply-chain-campaign-spreads-across-npm-and-ai-developer-ecosystems-like-wildfire","name":"Compromised Mistral AI and TanStack packages may have exposed credentials (Tom's Hardware)","type":"news_article","credibility":2},{"url":"https://www.securityweek.com/tanstack-mistral-ai-uipath-hit-in-fresh-supply-chain-attack/","name":"TanStack, Mistral AI, UiPath Hit in Fresh Supply Chain Attack (SecurityWeek)","type":"news_article","credibility":2},{"url":"https://www.bankinfosecurity.com/mass-supply-chain-attack-slams-npm-pypi-hits-mistral-ai-a-31672","name":"Mass Supply-Chain Attack Slams npm and PyPi, Hits Mistral AI (BankInfoSecurity)","type":"news_article","credibility":2}],"severity":"high"},{"content":"Separate from the confirmed May 2026 supply-chain attack, reports indicate a threat actor going by \"TeamPCP\" claimed on a hacking forum to have repeatedly breached Mistral AI and offered its source code for sale. As of the reporting reviewed, these claims were unverified: no data samples had been published, and Mistral's own security advisories did not confirm a breach of the scope advertised by the threat actor. This claim should be treated as low-confidence and unconfirmed.","heading":"Unverified Dark-Web Source Code Breach Claims","sources":[{"url":"https://thecybersecguru.com/news/mistral-ai-source-code-leak-2026/","name":"Mistral AI Hacked? Source Code Allegedly Put Up for Sale on Dark Web (The CyberSec Guru)","type":"news_article","credibility":3},{"url":"https://www.safestate.com/post/hackers-put-mistral-ai-source-code-up-for-sale-after-supply-chain-attack","name":"Hackers Put Mistral AI Source Code Up for Sale After Supply Chain Attack (SafeState)","type":"news_article","credibility":3}],"severity":"medium"},{"content":"In August 2025, a report catalogued by the OECD.AI AI Incidents Monitor described allegations, reportedly from a former Mistral employee, that one of the company's language models was secretly distilled from DeepSeek's models rather than being an original reinforcement-learning-trained system, and that benchmark results were misrepresented. This allegation appears to originate from a single source and has not been independently corroborated by a Tier 1 outlet in the sources reviewed; it is presented here as an unverified, low-confidence claim.","heading":"Model Provenance Allegations","sources":[{"url":"https://oecd.ai/en/incidents/2025-08-14-513f","name":"Mistral Accused of Misleading AI Model Origins and Benchmark Results (OECD.AI)","type":"research","credibility":3}],"severity":"low"},{"content":"Mistral AI (France) filed a domain-name dispute complaint with the WIPO Arbitration and Mediation Center on or around December 18, 2025, against a respondent listed as \"Unlimited Cohen\" (Armenia) over the domain mistralai.net (Case No. D2025-5290), indicating Mistral AI itself has pursued action against apparent domain-squatting or impersonation of its brand. No specific, verified case of a cryptocurrency token, airdrop, or wallet-drainer scam impersonating the \"Mistral AI\" name was identified in the sources reviewed; this is a generic risk category for any prominent technology brand and should not be read as evidence of an actual incident.","heading":"Domain Dispute and Impersonation Risk","sources":[{"url":"https://www.wipo.int/amc/en/domains/search/text.jsp?case=D2025-5290","name":"Mistral AI v. Unlimited Cohen, WIPO Case No. D2025-5290","type":"court_filing","credibility":1}],"severity":"low"}],"timeline":[{"date":"2023-04","event":"Mistral AI is founded in Paris by Arthur Mensch, Guillaume Lample, and Timothée Lacroix.","source":"TechCrunch / Wikipedia","source_url":"https://techcrunch.com/2026/07/04/what-is-mistral-ai-everything-to-know-about-the-openai-competitor/"},{"date":"2024-02","event":"Microsoft's roughly €15 million investment in and Azure distribution deal with Mistral AI is announced, prompting European Commission scrutiny.","source":"Bloomberg","source_url":"https://www.bloomberg.com/news/articles/2024-02-27/microsoft-s-mistral-ai-investment-to-be-examined-by-eu-watchdog","date_evidence":"Microsoft's Mistral AI Investment to Be Examined by EU Watchdog","date_original":"2024-02-27"},{"date":"2024-05","event":"UK Competition and Markets Authority closes its brief merger-control inquiry into Microsoft's investment in Mistral AI, finding it did not qualify for review.","source":"TechCrunch","source_url":"https://techcrunch.com/2024/05/17/microsoft-dodges-uk-antitrust-scrutiny-over-its-mistral-ai-stake/"},{"date":"2025-02-11","event":"Data-privacy complaint filed with France's CNIL alleging Le Chat's free tier lacked an opt-out for AI-training data use.","source":"OECD.AI Incidents Monitor","source_url":"https://oecd.ai/en/incidents/2025-02-11-f208","date_evidence":"incident dated 2025-02-11 per OECD.AI Incidents Monitor entry on the Le Chat data privacy controversy"},{"date":"2025-05-09","event":"Enkrypt AI publishes research alleging Mistral's Pixtral vision-language models generate CSAM and CBRN-related content at far higher rates than competing models.","source":"BankInfoSecurity","source_url":"https://www.bankinfosecurity.com/mistral-ai-models-fail-key-safety-tests-report-finds-a-28358","date_evidence":"Enkrypt AI report dated May 9, 2025"},{"date":"2025-08-14","event":"Allegation surfaces, reportedly from a former employee, that a Mistral model was secretly distilled from DeepSeek and benchmark results misrepresented.","source":"OECD.AI Incidents Monitor","source_url":"https://oecd.ai/en/incidents/2025-08-14-513f","date_evidence":"OECD.AI incident entry dated 2025-08-14"},{"date":"2025-09","event":"Mistral AI raises approximately €2 billion, valuing the company near €12 billion, following a $1.5 billion investment from ASML.","source":"Search aggregation of funding reporting","source_url":"https://www.clay.com/dossier/mistral-ai-funding"},{"date":"2025-12","event":"Mistral AI files a WIPO domain-name dispute (Case No. D2025-5290) over the domain mistralai.net.","source":"WIPO Arbitration and Mediation Center","source_url":"https://www.wipo.int/amc/en/domains/search/text.jsp?case=D2025-5290"},{"date":"2026-03","event":"Mistral AI raises $830 million from a group of banks to fund new datacenters near Paris and in Sweden.","source":"Search aggregation of funding reporting","source_url":"https://www.bloomberg.com/news/articles/2026-06-12/france-s-mistral-in-funding-talks-at-about-20-billion-valuation"},{"date":"2026-04-28","event":"Reporting alleges Le Chat amplified state-sponsored disinformation.","source":"OECD.AI Incidents Monitor","source_url":"https://oecd.ai/en/incidents/2026-04-28-e382","date_evidence":"OECD.AI incident entry dated 2026-04-28"},{"date":"2026-05-11","event":"Mistral AI's official PyPI package is compromised in the 'Mini Shai-Hulud' supply-chain attack, briefly distributing credential-stealing malware.","source":"SecurityWeek","source_url":"https://www.securityweek.com/tanstack-mistral-ai-uipath-hit-in-fresh-supply-chain-attack/","date_evidence":"attack dated May 11, 2026 per SecurityWeek reporting on the campaign timeline"},{"date":"2026-07-21","event":"Microsoft and Mistral announce an expanded, multibillion-dollar strategic partnership covering European AI infrastructure.","source":"Microsoft Source (official)","source_url":"https://news.microsoft.com/source/2026/07/21/microsoft-and-mistral-expand-strategic-partnership-to-give-enterprises-and-regulated-industries-frontier-ai-they-can-control/","date_evidence":"Microsoft Source news post dated 2026/07/21"},{"date":"2026-09-08","event":"Mistral AI announces a funding round led by Samsung, reportedly valuing the company at more than $23-24 billion.","source":"CNBC","source_url":"https://www.cnbc.com/2026/09/08/mistral-ai-funding-valuation-samsung.html","date_evidence":"CNBC article dated 2026/09/08 on Mistral's Samsung-led funding round"}],"sources_used":[{"url":"https://www.cnbc.com/2026/09/08/mistral-ai-funding-valuation-samsung.html","name":"Mistral bags $24 billion valuation as Samsung leads funding (CNBC)","type":"news_article","archive_url":"http://web.archive.org/web/20260911090637/https://www.cnbc.com/2026/09/08/mistral-ai-funding-valuation-samsung.html","credibility":1,"archive_timestamp":"2026-09-11T09:06:37+00:00"},{"url":"https://www.bloomberg.com/news/articles/2026-06-12/france-s-mistral-in-funding-talks-at-about-20-billion-valuation","name":"France's Mistral in Funding Talks at About €20 Billion Valuation (Bloomberg)","type":"news_article","archive_url":null,"credibility":1,"archive_timestamp":null},{"url":"https://www.bloomberg.com/news/articles/2024-02-27/microsoft-s-mistral-ai-investment-to-be-examined-by-eu-watchdog","name":"Microsoft's Mistral AI Investment to Be Examined by EU Watchdog (Bloomberg)","type":"news_article","archive_url":"http://web.archive.org/web/20251013105517/https://www.bloomberg.com/news/articles/2024-02-27/microsoft-s-mistral-ai-investment-to-be-examined-by-eu-watchdog","credibility":1,"archive_timestamp":"2025-10-13T10:55:17+00:00"},{"url":"https://techcrunch.com/2024/05/17/microsoft-dodges-uk-antitrust-scrutiny-over-its-mistral-ai-stake/","name":"Microsoft dodges UK antitrust scrutiny over its Mistral AI stake (TechCrunch)","type":"news_article","archive_url":"http://web.archive.org/web/20260804063734/https://techcrunch.com/2024/05/17/microsoft-dodges-uk-antitrust-scrutiny-over-its-mistral-ai-stake/","credibility":1,"archive_timestamp":"2026-08-04T06:37:34+00:00"},{"url":"https://techcrunch.com/2026/07/04/what-is-mistral-ai-everything-to-know-about-the-openai-competitor/","name":"What is Mistral AI? Everything to know about the OpenAI competitor (TechCrunch)","type":"news_article","archive_url":"http://web.archive.org/web/20260824201745/https://techcrunch.com/2026/07/04/what-is-mistral-ai-everything-to-know-about-the-openai-competitor/","credibility":1,"archive_timestamp":"2026-08-24T20:17:45+00:00"},{"url":"https://www.wipo.int/amc/en/domains/search/text.jsp?case=D2025-5290","name":"Mistral AI v. Unlimited Cohen, WIPO Case No. D2025-5290","type":"court_filing","archive_url":"https://web.archive.org/web/20260923005808/https://www.wipo.int/amc/en/domains/decisions/pdf/2025/d2025-5290.pdf","credibility":1,"archive_timestamp":"2026-09-23T00:58:08+00:00"},{"url":"https://news.microsoft.com/source/2026/07/21/microsoft-and-mistral-expand-strategic-partnership-to-give-enterprises-and-regulated-industries-frontier-ai-they-can-control/","name":"Microsoft and Mistral expand strategic partnership (Microsoft Source)","type":"official","archive_url":"http://web.archive.org/web/20260821090401/https://news.microsoft.com/source/2026/07/21/microsoft-and-mistral-expand-strategic-partnership-to-give-enterprises-and-regulated-industries-frontier-ai-they-can-control/","credibility":2,"archive_timestamp":"2026-08-21T09:04:01+00:00"},{"url":"https://aibusiness.com/companies/antitrust-regulator-drops-probe-into-microsoft-s-mistral-deal","name":"Antitrust Regulator Drops Probe Into Microsoft's Mistral Deal","type":"news_article","archive_url":null,"credibility":2,"archive_error":"error:no-request","archive_status":"unarchivable","archive_timestamp":null},{"url":"https://www.mlex.com/mlex/articles/2191760","name":"Mistral AI, Microsoft talks overlapped with lobbying EU AI Act for looser rules (MLex)","type":"news_article","archive_url":"https://web.archive.org/web/20260923005559/https://www.mlex.com/mlex/articles/2191760","credibility":2,"archive_timestamp":"2026-09-23T00:55:59+00:00"},{"url":"https://corporateeurope.org/en/2024/03/trojan-horses-how-european-startups-teamed-big-tech-gut-ai-act","name":"Trojan horses: how European startups teamed up with Big Tech to gut the AI Act (Corporate Europe Observatory)","type":"research","archive_url":"http://web.archive.org/web/20260904153503/https://corporateeurope.org/en/2024/03/trojan-horses-how-european-startups-teamed-big-tech-gut-ai-act","credibility":2,"archive_timestamp":"2026-09-04T15:35:03+00:00"},{"url":"https://venturebeat.com/ai/the-eu-ai-act-passed-but-mistral-parties-on-the-ai-beat","name":"The EU AI Act passed, but Mistral parties on (VentureBeat)","type":"news_article","archive_url":null,"credibility":2,"archive_timestamp":null},{"url":"https://oecd.ai/en/incidents/2025-02-11-f208","name":"Mistral AI Faces Data Privacy Controversy Over Chatbot 'Le Chat' (OECD.AI)","type":"research","archive_url":"http://web.archive.org/web/20251210225859/https://oecd.ai/en/incidents/2025-02-11-f208","credibility":2,"archive_timestamp":"2025-12-10T22:58:59+00:00"},{"url":"https://www.usine-digitale.fr/article/mistral-ai-vise-par-une-plainte-pour-violation-des-donnees-personnelles.N2227401","name":"Mistral AI visé par une plainte pour violation des données personnelles (Usine Digitale)","type":"news_article","archive_url":"http://web.archive.org/web/20260124043459/https://www.usine-digitale.fr/article/mistral-ai-vise-par-une-plainte-pour-violation-des-donnees-personnelles.N2227401","credibility":2,"archive_timestamp":"2026-01-24T04:34:59+00:00"},{"url":"https://wancore.fr/news/mistral-ai-dans-le-viseur-de-la-cnil/","name":"Mistral AI dans le viseur de la CNIL (Wancore)","type":"news_article","archive_url":"https://web.archive.org/web/20260922231805/https://wancore.fr/news/mistral-ai-dans-le-viseur-de-la-cnil/","credibility":2,"archive_timestamp":"2026-09-22T23:18:05+00:00"},{"url":"https://legal.mistral.ai/terms/privacy-policy","name":"Mistral AI's Privacy Policy","type":"official","archive_url":"http://web.archive.org/web/20260922120403/https://legal.mistral.ai/terms/privacy-policy/","credibility":2,"archive_timestamp":"2026-09-22T12:04:03+00:00"},{"url":"https://www.bankinfosecurity.com/mistral-ai-models-fail-key-safety-tests-report-finds-a-28358","name":"Mistral AI Models Fail Key Safety Tests, Report Finds (BankInfoSecurity)","type":"research","archive_url":"https://web.archive.org/web/20260923005251/https://www.bankinfosecurity.com/mistral-ai-models-fail-key-safety-tests-report-finds-a-28358","credibility":2,"archive_timestamp":"2026-09-23T00:52:51+00:00"},{"url":"https://oecd.ai/en/incidents/2026-04-28-e382","name":"French AI Chatbot Mistral Amplifies State-Sponsored Disinformation (OECD.AI)","type":"research","archive_url":"https://web.archive.org/web/20260922231546/https://oecd.ai/en/incidents/2026-04-28-e382","credibility":2,"archive_timestamp":"2026-09-22T23:15:46+00:00"},{"url":"https://www.cryptopolitan.com/french-ai-startup-mistral-faces-backlash/","name":"French AI Startup Mistral Faces Backlash as New LLM Generates Harmful Content (Cryptopolitan)","type":"news_article","archive_url":"http://web.archive.org/web/20260120115402/https://www.cryptopolitan.com/french-ai-startup-mistral-faces-backlash/","credibility":3,"archive_timestamp":"2026-01-20T11:54:02+00:00"},{"url":"https://www.tomshardware.com/tech-industry/cyber-security/compromised-mistral-ai-and-tanstack-packages-may-have-exposed-github-cloud-and-ci-cd-credentials-in-mini-shai-hulud-malware-infection-supply-chain-campaign-spreads-across-npm-and-ai-developer-ecosystems-like-wildfire","name":"Compromised Mistral AI and TanStack packages may have exposed credentials (Tom's Hardware)","type":"news_article","archive_url":"http://web.archive.org/web/20260807004718/https://www.tomshardware.com/tech-industry/cyber-security/compromised-mistral-ai-and-tanstack-packages-may-have-exposed-github-cloud-and-ci-cd-credentials-in-mini-shai-hulud-malware-infection-supply-chain-campaign-spreads-across-npm-and-ai-developer-ecosystems-like-wildfire","credibility":2,"archive_timestamp":"2026-08-07T00:47:18+00:00"},{"url":"https://www.securityweek.com/tanstack-mistral-ai-uipath-hit-in-fresh-supply-chain-attack/","name":"TanStack, Mistral AI, UiPath Hit in Fresh Supply Chain Attack (SecurityWeek)","type":"news_article","archive_url":"http://web.archive.org/web/20260518032027/https://www.securityweek.com/tanstack-mistral-ai-uipath-hit-in-fresh-supply-chain-attack/","credibility":2,"archive_timestamp":"2026-05-18T03:20:27+00:00"},{"url":"https://www.bankinfosecurity.com/mass-supply-chain-attack-slams-npm-pypi-hits-mistral-ai-a-31672","name":"Mass Supply-Chain Attack Slams npm and PyPi, Hits Mistral AI (BankInfoSecurity)","type":"news_article","archive_url":"https://web.archive.org/web/20260922231908/https://www.bankinfosecurity.com/mass-supply-chain-attack-slams-npm-pypi-hits-mistral-ai-a-31672","credibility":2,"archive_timestamp":"2026-09-22T23:19:08+00:00"},{"url":"https://thecybersecguru.com/news/mistral-ai-source-code-leak-2026/","name":"Mistral AI Hacked? Source Code Allegedly Put Up for Sale on Dark Web (The CyberSec Guru)","type":"news_article","archive_url":"https://web.archive.org/web/20260922231714/https://thecybersecguru.com/news/mistral-ai-source-code-leak-2026/","credibility":3,"archive_timestamp":"2026-09-22T23:17:14+00:00"},{"url":"https://www.safestate.com/post/hackers-put-mistral-ai-source-code-up-for-sale-after-supply-chain-attack","name":"Hackers Put Mistral AI Source Code Up for Sale After Supply Chain Attack (SafeState)","type":"news_article","archive_url":"http://web.archive.org/web/20260615140808/https://www.safestate.com/post/hackers-put-mistral-ai-source-code-up-for-sale-after-supply-chain-attack","credibility":3,"archive_timestamp":"2026-06-15T14:08:08+00:00"},{"url":"https://oecd.ai/en/incidents/2025-08-14-513f","name":"Mistral Accused of Misleading AI Model Origins and Benchmark Results (OECD.AI)","type":"research","archive_url":"http://web.archive.org/web/20251209033328/https://oecd.ai/en/incidents/2025-08-14-513f","credibility":3,"archive_timestamp":"2025-12-09T03:33:28+00:00"},{"url":"https://en.wikipedia.org/wiki/Mistral_AI","name":"Mistral AI (Wikipedia)","type":"other","archive_url":"http://web.archive.org/web/20260918143137/https://en.wikipedia.org/wiki/Mistral_AI","credibility":2,"archive_timestamp":"2026-09-18T14:31:37+00:00"}],"source_tags":[],"addresses":[],"reviewed":false,"reviewed_by":null,"model_used":"claude-code-investigator","created_at":"2026-09-22T22:23:48.547222+00:00","updated_at":"2026-09-23T02:23:09.303453+00:00"}}