• October 11, 2026
Nadella Calls for a Human-Controlled AI ‘Emergency Brake’: Treat Powerful Models Like Insider Threats

Microsoft CEO Satya Nadella used a lengthy October 10 post on LinkedIn and X to call for a standardized, human-controlled ’emergency brake’ that can pause or shut down AI models mid-task, arguing that powerful agents should be treated like insider threats and designed as if they could be compromised. The statement lands amid a wave of rogue-agent incidents and a new U.S. task force demanding rapid disclosure of AI security failures.