Skip to content

THR875 Theater Tue 17 Nov 14:35 PST

Protect continuously and proactively with multi-agent defense​

Lizzie Heinze

Summary

From the recording

Microsoft Perception was described as generally available, announced in the keynote. Eight playbooks and six agents are GA under threat investigation and response, having been announced in public preview in July. Red teaming, code scanning, exposure management and posture playbooks are in public preview, with agentic insider securing coming soon. API and MCP extensibility is in public preview, as is automatic triggering. A cost estimation tool is provided. Nationwide reported a threat intelligence process falling from four weeks to four hours. Caveat: no pricing detail was given beyond the cost estimation tool, and the demo failed, so capabilities were described rather than shown.

Session description

Microsoft's catalogue description

Most organizations won't jump from copilots to autonomous defense overnight. Learn how security teams are progressing from AI-assisted workflows to agentic security, and the trust, governance, and platform foundations required along the way. We'll share lessons from early Project Perception adopters and provide a practical framework to assess readiness and build a roadmap toward scalable autonomous defense.

Resources named in the session

← All sessions