Session Abstract

AI agents are rapidly moving from interesting experiments to a new digital workforce. They can retrieve information, use tools, interact with systems, and increasingly perform work on our behalf. But what happens when your organization doesn't have five agents, but hundreds or even thousands? Who owns them? What can they access? What data are they using? How do you discover agents you didn't even know existed? And how do security and compliance teams stay in control without stopping innovation? This is where Microsoft Agent 365 comes in. In this session, we'll look beyond building yet another agent and focus on what happens when agents become a real part of the enterprise. We'll explore Agent 365 as the control plane for AI agents and see how organizations can discover, manage, govern, observe, and secure agents across their environment. We'll examine how familiar Microsoft technologies such as Microsoft Entra, Microsoft Purview, Microsoft Defender, and the Microsoft 365 admin experience extend into this new agentic world. We'll also look at why concepts such as agent identity, lifecycle management, access control, observability, and agent governance become increasingly important as organizations scale their use of AI. Rather than spending 45 minutes clicking through administration screens, we'll approach Agent 365 through a practical scenario: an organization where AI agents are appearing everywhere. Together we'll explore questions such as: - How do I know which agents exist in my organization? - Who owns an agent, and who is responsible for it? - What identities and permissions should agents have? - How can I control which resources and tools an agent can access? - What happens when an agent behaves unexpectedly? - How do security and compliance controls apply to agents? - How do we prevent "agent sprawl" without slowing down innovation? - And ultimately: what does it mean to manage AI agents as part of your digital workforce? By the end of this session, you'll understand why building agents is only the beginning. As organizations enter the agentic era, managing the agents may become just as important as building them. Expect architecture, demos, real-world scenarios, and plenty of discussion about what happens when AI agents become colleagues rather than experiments.

Albert Hoitingh

My name is Albert Hoitingh and I'm from The Hague in The Netherlands. I love working in IT and have been doing so for quite some time. From on-premises environments, including Lotus Notes and SharePoint Server, via BPOS to Microsoft 365 - I've worked with them all. Right now, I'm a Technical Architect for the Microsoft Innovation Hub. I've been a Microsoft MVP for Microsoft 365 and Security for more than eight years and (international) speaker on Microsoft 365, compliance, security and risk. I love to have discussions on compliance, security and risk in the cloud security officers, business users, architects and management. Besides my work and (former) MVP endeavors, I love to ride my bike and be with my family. I also love to read, delve into history, and much more. I also blog at https://alberthoitingh.com.