Companies are increasingly adopting AI agents – from Copilot scenarios to complex, multi-agent workflows. As autonomy grows, so do the demands for transparency, traceability, and compliance: Who triggered which action, with what data, and under which policies?

Agent 365 is Microsoft’s central control and governance platform for AI agents in the enterprise. It enables organizations to uniformly register, observe, govern, and secure all AI agents – regardless of whether they were built with Microsoft tools, Copilot Studio, Foundry, or external platforms.

This session shows how Agent 365 and Microsoft Purview together make every AI interaction visible, auditable, and governed. Guiding principle: Agent 365: Enable observability at every layer of the AI stack. From the model and prompt layer through tool calls to agent actions and data flows, every layer of the AI stack becomes observable and logged.