Organization/organizations/mistral-ai
Mistral AI
Incidents by year
20241
202510
20267
Incident records18
Jun 2026French Treasury halts use of Alibaba's Qwen AI over biased outputsLe Monde.frLe Figaro.frLe Parisien+4Jun 2026Mistral AI fails to filter Russian propaganda in European studyHospodarske NovinyAktuality.skwww.sme.skMay 2026Cisco Study Reveals Major AI Models Vulnerable to Multi-Turn Adversarial AttacksSiliconANGLEDigitMay 2026Mistral AI Source Code Stolen in Data BreachFranceinfoTechNaduLe Figaro.fr+3Apr 2026Mistral's Le Chat chatbot repeats state-sponsored disinformation in auditFranceinfoLe Figaro.frFeb 2026Mistral AI Accused of Using Copyrighted Works to Train Large Language ModelClubic.comMediapartFeb 2026Pronunciation Bypass Jailbreak Affects Multiple AI Models0din.aiDec 2025Meta and Mistral AI cofounder accused of copyright infringement for AI trainingLes NumériquesClubic.comDec 2025Mistral Vibe CLI Shell Expansion Command Execution Vulnerabilitymindgard.aiDec 2025Mistral Vibe CLI trusts MCP configuration files with arbitrary commandsmindgard.aiNov 2025Poetry-based jailbreaks bypass AI safety guardrails across multiple modelsThe GuardianJul 2025Study finds AI models cannot fully forget personal dataABC Digitalwww.diariolibre.comJun 2025WormGPT variants hijack Grok and Mixtral for cybercrimeCyberScoopTechNaduMay 2025Mistral's Pixtral models generate harmful content at high rates, mental health risks identifiedThe HinduCointelegraphMar 2025AI Search Engines Fabricate News Citations, Undermining PublishersTechSpotDeveloppez.comfutura-sciences.com+1Mar 2025Multiple AI models vulnerable to "Police Interrogation Terminal" jailbreak0din.aiFeb 2025Mistral AI faces CNIL complaint over Le Chat free tier data practicesLe FigaroL'Éclaireur FnacOct 2024AI Resume Screening Shows Racial and Gender Bias in Three LLMsArs TechnicaTech XploreGeekWire+2