Add guidance for controlling AI agent identity and permissions

[azarboon]

This change adds actionable guidance for controlling the permissions and autonomy of AI agents to address risks related to OWASP LLM06 Excessive Agency. The current article focuses on securing AI workloads, networks, and data paths, but does not sufficiently address AI agents that can autonomously invoke APIs, modify data, or trigger downstream actions. This creates a security gap in which agents may operate with excessive privileges, limited traceability, and insufficient oversight. This update introduces agent-specific controls to help mitigate these risks.

[prmerger-automator]

@azarboon : Thanks for your contribution! The author(s) and reviewer(s) have been notified to review your proposed change.

[learn-build-service-prod]

Learn Build status updates of commit dc9380e:

💡 Validation status: suggestions

File	Status	Preview URL	Details
docs/scenarios/ai/secure.md	💡Suggestion		Details

docs/scenarios/ai/secure.md

• Line 45, Column 390: [Suggestion: docs-link-absolute - See documentation] Absolute link 'https://learn.microsoft.com/entra/agent-id/identity-professional/microsoft-entra-agent-identities-for-ai-agents' will be broken in isolated environments. Replace with a relative link.
• Line 45, Column 551: [Suggestion: docs-link-absolute - See documentation] Absolute link 'https://learn.microsoft.com/azure/role-based-access-control/overview' will be broken in isolated environments. Replace with a relative link.
• Line 45, Column 757: [Suggestion: docs-link-absolute - See documentation] Absolute link 'https://learn.microsoft.com/agent-framework/tutorials/agents/function-tools-approvals?pivots=programming-language-csharp' will be broken in isolated environments. Replace with a relative link.
• Line 45, Column 1027: [Suggestion: docs-link-absolute - See documentation] Absolute link 'https://learn.microsoft.com/entra/identity/conditional-access/overview' will be broken in isolated environments. Replace with a relative link.
• Line 45, Column 1184: [Suggestion: docs-link-absolute - See documentation] Absolute link 'https://learn.microsoft.com/security/zero-trust/zero-trust-overview' will be broken in isolated environments. Replace with a relative link.

For more details, please refer to the build report.

Note: Your PR may contain errors or warnings or suggestions unrelated to the files you changed. This happens when external dependencies like GitHub alias, Microsoft alias, cross repo links are updated. Please use these instructions to resolve them.

[learn-build-service-prod]

Learn Build status updates of commit 80969f8:

💡 Validation status: suggestions

File	Status	Preview URL	Details
docs/scenarios/ai/secure.md	💡Suggestion		Details

docs/scenarios/ai/secure.md

• Line 45, Column 390: [Suggestion: docs-link-absolute - See documentation] Absolute link 'https://learn.microsoft.com/entra/agent-id/identity-professional/microsoft-entra-agent-identities-for-ai-agents' will be broken in isolated environments. Replace with a relative link.
• Line 45, Column 551: [Suggestion: docs-link-absolute - See documentation] Absolute link 'https://learn.microsoft.com/azure/role-based-access-control/overview' will be broken in isolated environments. Replace with a relative link.
• Line 45, Column 757: [Suggestion: docs-link-absolute - See documentation] Absolute link 'https://learn.microsoft.com/agent-framework/tutorials/agents/function-tools-approvals' will be broken in isolated environments. Replace with a relative link.
• Line 45, Column 992: [Suggestion: docs-link-absolute - See documentation] Absolute link 'https://learn.microsoft.com/entra/identity/conditional-access/overview' will be broken in isolated environments. Replace with a relative link.
• Line 45, Column 1149: [Suggestion: docs-link-absolute - See documentation] Absolute link 'https://learn.microsoft.com/security/zero-trust/zero-trust-overview' will be broken in isolated environments. Replace with a relative link.

For more details, please refer to the build report.

Note: Your PR may contain errors or warnings or suggestions unrelated to the files you changed. This happens when external dependencies like GitHub alias, Microsoft alias, cross repo links are updated. Please use these instructions to resolve them.

[azarboon]

I considered using relative links, but I was concerned about introducing incorrect paths. In this repo, referenced pages resolve under different prefixes such as /azure-docs and /entra-docs, while links in the article use /azure and /entra. Without a clear way to verify the correct relative path, I opted to use full URLs to avoid linking to the wrong location. These can be safely converted to relative links if preferred. This approach minimizes the risk of broken or misleading links.

[Copilot]

Pull request overview

This PR adds guidance for controlling AI agent permissions and autonomy to address OWASP LLM06 Excessive Agency risks. The addition expands the security documentation to cover AI agents that can autonomously invoke APIs and perform actions, filling a gap in the current guidance that focuses primarily on workloads, networks, and data paths.

Changes:

• Added a new point (#4) under "Secure AI resources" section with comprehensive guidance on controlling AI agent permissions and autonomy

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[ttorble]

@stephen-sumner

Can you review the proposed changes?

IMPORTANT: When the changes are ready for publication, adding a #sign-off comment is the best way to signal that the PR is ready for the review team to merge.

#label:"aq-pr-triaged"
@MicrosoftDocs/public-repo-pr-review-team

[azarboon]

@stephen-sumner can you please review this?

[stephen-sumner]

@azarboon - It's in the backlog for me to review. Thank you for your patience.