Microsoft CEO Calls for Emergency Brake Mechanism for Advanced AI Models

According to a report by Golden Finance, on October 11, Microsoft (MSFT.O) CEO Satya Nadella stated that companies should treat powerful artificial intelligence models as potential internal threats, assume these models may have been compromised, and establish an "emergency brake" mechanism to prevent autonomous intelligent agent models from going out of control. On Saturday, Nadella posted on social media: "We must assume that models have already been compromised and implement restrictions from the very beginning. Think of this as an emergency brake device. Authorized personnel should always be able to pause or shut down the model during task execution." Nadella's safety recommendations include: avoiding reliance on a single model for critical decision-making; retaining immutable records of intelligent agent behavior; and subjecting systems to independent audits. He also called for the disclosure of major failures or security vulnerabilities, urging companies to share the causes and details of incidents so that other organizations can strengthen their security measures.