Q
A company has a multi-agent solution built in Microsoft Foundry. The solution grounds responses in internal HR documents and uses agent tools to call internal and external systems.
Security testing shows that prompt injection attacks can expose internal HR content and some external tool calls run without clear permission boundaries.
You need to recommend a solution that enforces agent governance by applying least-privilege access to the grounding data, restricting external tool usage, and preserving auditability.
What should you recommend?
Question Info
Choose the Best Option
Click any option to instantly check if you're correct.
Explanation
Objective:
3.4 Design responsible AI, security, governance, risk management, and compliance
What This Item Tests:
Design governance for agents
Additional Reading:
Governance and security for AI agents across the organization
Establish a single control plane for AI agents across the organization
Data governance and compliance
Manage your Copilot Studio projects, an overview
Rationale:
Using managed identities with Azure RBAC enforces least-privilege access for both grounding data and tool execution, while restricting external tools to explicitly approved integrations establishes clear governance and auditability boundaries. Monitoring controls, log storage changes, or model selection do not enforce permission boundaries or govern agent tool usage.
Share This Question
Challenge a friend or share with your study group.