[Microsoft CEO: Companies Should Treat AI as an Internal Security Risk, Ensuring AI Agents Can Be Paused or Shut Down at Any Time] Odaily Planet Daily News - Microsoft CEO Satya Nadella stated that companies should incorporate AI models into their internal security risk management systems. Even if the models themselves are not malicious, they may pose security risks due to errors or exploitation by attackers, especially when sensitive data access and critical operations are involved. Companies cannot solely rely on the security assurances of model providers. Nadella emphasized that AI models should be separated from the agent execution framework, with external systems controlling data access and tool invocation permissions. For critical operations, tamper-proof records that allow for human review should be retained. Additionally, companies should consider the possibility of model breaches during the deployment phase, ensuring authorized personnel can pause or shut down agents at any time, while continuously conducting security testing and incident disclosures. Furthermore, Microsoft announced this month that the Windows platform officially supports the MXC isolation tool, which can restrict agent access to files and networks through external policies.