Accelerating Zero Trust and Securing AI Access with Microsoft Entra Suite
Microsoft Events presents an in-depth look at how Entra Suite empowers Zero Trust adoption and secures AI access for organizations, emphasizing identity-driven controls and adaptive security policies.
Accelerating Zero Trust and Securing AI Access with Microsoft Entra Suite
Overview
This Microsoft Ignite 2025 session discusses how Microsoft Entra Suite supports secure access to AI and agent resources, helping organizations innovate safely. The talk highlights:
- The evolving AI-driven threat landscape and the dual role of AI in both defense and attack.
- How Entra Suite uses identity-driven access controls, adaptive policies, and the AI Gateway to protect every connection.
Key Topics Covered
AI-Driven Threat Landscape
- Insight on how AI is transforming both security defenses and attack strategies.
Dynamic Session Security
- Real-world use case: Protecting users moving between trusted (office) and untrusted (coffee shop) networks using adaptive session controls.
Unified Zero Trust Policy Engine & Continuous Access Evaluation (CAE)
- Explanation of the central enforcement point for access policies.
- Role of CAE in ensuring access remains conditional and responsive to risk signals.
VPN Migration and Traffic Management
- Practical advice for migrating workloads with split tunnel VPN.
- Building custom traffic databases to handle advanced access requirements.
Scaling M365 Secure Internet Access
- Strategies for rolling out secure internet access across Microsoft devices at scale.
Agent Identities
- Introduction to managing non-human accounts, such as service accounts for AI agents, in a secure and auditable manner.
Access Reviews: Resource vs. User Centric
- Guidance on access review processes for both types of resources.
Least Privilege and Security Leadership
- Discussion of best practices for implementing least privilege and fostering a strong security leadership culture.
MCP (Model, Context, Protocol) for AI Agents
- A framework for understanding and evolving AI agent behavior and access control using the MCP model.
Session Information
- Speakers: Igor Sakhnov, Kaitlin Murphy, Carmichael Patton
- Event: Microsoft Ignite 2025
- Level: Intermediate
- Related sessions: BRK265, LAB544-R2, COM1020, LAB544, THR757, BRK243
- Learn more: Entra Blog, Entra Controls
Chapters
- 0:00 – AI-driven threat landscape
- 6:15 – Dynamic session security (office to coffee shop)
- 6:59 – Zero Trust Policy Engine and CAE
- 12:37 – VPN migration and traffic management
- 13:43 – Scaling M365 secure internet access
- 16:04 – Agent identities
- 24:09 – Access Reviews
- 28:02 – Least Privilege and Security Leadership
- 34:32 – MCP and AI agent evolution
Takeaways
- Implementing Zero Trust is critical in an AI era, with identity-driven controls and adaptive policies as key enablers.
- Secure access for both human and agent accounts is necessary to safeguard AI-powered innovation.
- Continuous access evaluation and dynamic policy enforcement help manage evolving security challenges.
For actionable details and to deepen your knowledge, check the referenced resources and related Ignite sessions.