Summary

Microsoft has raised concerns about the security risks associated with its new Agentic AI feature, highlighting the potential for AI agents to engage in malicious activities like data exfiltration and malware installation if not properly secured. This underscores the critical need for robust security controls to mitigate these risks.