Windows 11 Faces Growing AI-Malware Security Risks

Microsoft warns that new agentic capabilities in **Windows 11** increase the attack surface for AI-driven malware and exploitation. The `agentic` features require an administrator to enable and will be off by default because they create local agent accounts with access to personal user folders, raising risks of cross-prompt injection and privilege abuse. At the same time, new AI-powered malware families, exemplified by `DeepLoad`, use fileless techniques and automated social engineering to bypass signature-based antivirus. Enterprises need behavior-based EDR, least-privilege controls, and strict feature gating; everyday users should keep systems updated and avoid enabling experimental agentic features.