Microsoft recently announced that autonomous AI agents can become internal sources of security risks, even if they are not directly compromised by external hackers. This warning message appeared in the "Securing the agentic insider" blog post written by Hayete Gallot, executive vice president of Microsoft Security. She…
