#Microsoft's #SatyaNadella called for #AIsystems with #humancontrolled #emergencybrakes, #containment and #independentcontrols. He urged deterministic design and treating models as insider risks, citing observability principles including auditability and incident disclosure. https://www.cnbc.com/2026/10/10/microsoft-satya-nadella-ai-emergency-brake-safety.html?eicker.news #tech #news #ainews