Microsoft CEO Nadella warns AI models could be compromised

Satya Nadella said companies should treat powerful AI models as internal threats, assuming models may be compromised. He urged “emergency brake” mechanisms to pause or shut down autonomous agents, advised against relying on a single model for critical decisions, and called for immutable logs, independent audits, and mandatory disclosure of major failures or security vulnerabilities (including causes and details).