[WEBINAR] - Protect the Agentic Enterprise: Can you detect and respond when an agent becomes a risk?

What happens when the agent itself becomes the risk?

As AI agents gain more autonomy, so does their ability to interact with tools, identities, data and business resources. A compromised, manipulated or misconfigured agent can use these capabilities in unintended ways.

The challenge is no longer just protecting the agent, but also understanding what it could reach if it were compromised, how to identify anomalous behavior and which controls exist to respond before an activity causes greater impact.

In Protect the Agentic Enterprise, we will explore how Microsoft Agent 365, Microsoft Defender and Intune extend security capabilities to the AI agent ecosystem, providing more context about where agents operate, what they interact with and which behaviors may represent a risk.

Learn how to move from visibility to detection, investigation and response, assessing an agent's blast radius and putting containment mechanisms in place to reduce exposure to potential threats.

What will you see during the session?

• IDENTIFY: Understand the context of each agent: the associated identities, tools, MCP servers and business resources it can reach, as well as its potential blast radius.
• DETECT: Explore how security and observability signals help you identify suspicious activity and investigate anomalous behavior across the agent ecosystem.
• RESPOND: Learn how to respond to risky activity and apply controls to contain agents, restrict their access and reduce their potential impact.