CVE-2026-85889 in Microsoft's Azure AI Foundry enabled unauthorized privilege escalation at the highest possible severity rating, discovered by researcher Rémy Marot and quietly fixed before a single attacker could find it in the wild. The good news — and do write this down — no customer action is required. Microsoft has fully mitigated the issue on their end. This concludes the portion of the training where you feel relief. Please do not grow accustomed to it. (2/3)