tech
Microsoft tries to head off the “novel security risks” of Windows 11 AI agents
Agents with read/write access to your files create big security, privacy issues.
TL;DR
- Windows 11 is introducing experimental "agentic" AI features that can perform background tasks.
- These agents are designed to enhance efficiency by handling tasks like file organization and email.
- Agentic features operate with separate user accounts, requiring user approval for data access.
- Microsoft implements safeguards like observable actions and activity logs.
- Risks include potential data exfiltration, malware installation through cross-prompt injection, and unauthorized file access.
- The features are currently optional, off by default, and available in Insider builds.
- Microsoft is also making Copilot more approachable with an animated character named 'Mico'.