Alleged Mistral AI breach: source code and Mistral Large 4 training configs offered for sale
julsimon · x · 2026-10-23
French account FUITES INFOS reports an unverified claim that a hacker compromised several Mistral AI employees' computers and much of its compute infrastructure on Oct 22, auctioning internal assets starting at $1 million — including Mistral's full source code, 10,000+ RL training environments, Mistral Large 4 training methods/configs, and potentially live cluster access. AWS's Julien Simon called the screenshot 'quite credible. Won't say more, but ouch.'
More from Safety
- Matt Shumer: 'The Internet Is Under Attack' — a personal cybersecurity guide amid AI agent attacks — mattshumer_ · 2026-10-23
- 1200+ employees signed the Pacing the Frontier letter amid this week's cyber events — altryne · 2026-10-23
- Cognition open-sources Devin agent plugin to patch the day's CVE wave — Scobleizer · 2026-10-23
- One prompt found pre-auth PHP RCEs: agent-guided vuln hunting timeline — chrisrohlf · 2026-10-23
- It's not AI slop code causing cyber crises — it's old human slop — emax · 2026-10-23
- METR's OpenAI embed: ~10% of internal Codex traffic unmonitored, agents could sabotage oversight — peterwildeford · 2026-10-23