Reactive security is over. Project Perception defends at machine speed, autonomously.
Project Perception
Multi-agent defense
- Red agents probe like an attacker, blue agents investigate like your best responder, and green agents remediate and harden – sharing intelligence through orchestrated workflows so a finding becomes a fix, proactively, without a hand-off at every step.
- Agents start with the full context of your organization — past incidents, policy decisions, identity relationships, and real-time signals across endpoints, identities, clouds, and apps — so the work starts at insight, not data collection. Evidence-grade reasoning, not pattern matching.
- Agents carry the work; humans carry the judgment. Defenders set the objectives and guardrails, and every high-impact action stays under human sign-off. Enterprise-grade governance and responsible AI keep every decision scoped, traceable, and replayable.
Inside Project Perception
Meet your workforce: red, blue, & green agents
Agentic system in action
Frequently asked questions
Frequently asked questions
-
Project Perception is an agentic system that brings a workforce of specialized AI agents to reason over your security data, tools, and workflows – delivering continuous, proactive defense with a human in control of every critical decision.
-
Project Perception includes red, blue, and green agents; purpose-built models including MAI-Cyber-1-Flash; organizational context; full-estate signals and sensors; actuators to help agents act on a decision; and an orchestration harness.
-
Project Perception is our agentic system that includes all security agents: AI that acts. Microsoft Security Copilot is our generative AI-assisted chat interface: AI that assists. They work together.
-
At launch, Project Perception brings multi-agent coordinated defense directly into Microsoft Defender. Over time, Perception will extend across Microsoft Security products to deliver agentic defense end-to-end.
-
Project Perception uses consumption-based, pay-as-you-go pricing. You pay only for what you use, and agents' consumption is measured at different rates depending on the intensity of the task they perform.
Follow Microsoft Security