This is the Trace Id: ea08fc7505d42225469a742ae31128d3
Skip to main content Microsoft Defender Microsoft Entra Microsoft Intune Microsoft Purview Microsoft Security Copilot Microsoft Sentinel SIEM Microsoft Sentinel platform View all products AI-powered cybersecurity Cloud security Data security & governance Identity & network access Privacy & risk management Security for AI Small and medium business Unified SecOps Zero Trust Pricing Services Partners Why Microsoft Security Cybersecurity awareness Customer stories Security 101 Product trials How we protect Microsoft Industry recognition Microsoft Security Insider Microsoft Digital Defense Report Security Response Center Microsoft Security Blog Microsoft Security Events Microsoft Tech Community Documentation Technical Content Library Training & certifications Compliance Program for Microsoft Cloud Microsoft Trust Center Security Engineering Portal Service Trust Portal Microsoft Secure Future Initiative Business Solutions Hub Contact Sales Start free trial Microsoft Security Azure Dynamics 365 Microsoft 365 Microsoft Teams Windows 365 Microsoft AI Azure Space Mixed reality Microsoft HoloLens Microsoft Viva Quantum computing Sustainability Education Automotive Financial services Government Healthcare Manufacturing Retail Find a partner Become a partner Partner Network Microsoft Marketplace Software companies Blog Microsoft Advertising Developer Center Documentation Events Licensing Microsoft Learn Microsoft Research View Sitemap
AGENTIC SYSTEM

Project Perception

A workforce of specialized AI agents, now in preview – red, blue, and green – that reason across your security data, tools, and workflows to expose gaps, investigate threats, and remediate them continuously. You set the strategy; the agents carry the load.
A person pointing at a screen with a blue background.
Overview

Multi-agent defense

Coordinated agents. Shared intelligence. Protection that never stops.
  • 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. 
    A person sitting at a desk with a laptop.
  • 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.
    A person using a laptop.
  • 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.
    A close-up of hands using a phone.
WHAT’S INCLUDED

Inside Project Perception

The building blocks of agentic defense — intelligence, agents, and models to power your security stack while keeping humans in charge.
Actuators
Decisions become action
The mechanism an agent uses to act on its environment, turning a decision into a real-world effect, not just a recommendation.
Agents
Red, blue, & green agents
Specialized agents that expose weaknesses, detect and respond, and remediate and harden across the entire attack lifecycle.
Harness
Orchestration for agents
The framework, tools, and controls that orchestrate every agent and model to operate reliably, be tested, or be guided effectively. 
Models
Purpose-built cyber models
A multi-model approach – including the new MAI-Cyber-1-Flash model – to reason over threats with deep, security-specific expertise. 
Context
Real-time security context
Continuously enriched intelligence that connects TI, signals, and organizational knowledge to give agents operational context.
Signals & sensors
End-to-end signals
Complete visibility across a breadth of signals from sensors spanning endpoints, identities, clouds, and apps so no threat slips past unseen.
THE AGENTS

Meet your workforce: red, blue, & green agents

Three specialized agents that cover you across the full attack lifecycle, and hand off to each other automatically.
A white background with black lines.
FAQ

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; 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, measured in Security Compute Units (SCUs). Different agents consume SCUs at different rates depending on the intensity of the task they perform.

  • A Security Compute Unit (SCU) is the unit used to measure Project Perception usage. As agents run scenarios in the portal, they consume SCUs. More intensive tasks consume more SCUs, so your cost reflects the actual work performed.

A person using a laptop
Get started

Protect everything

Make your future more secure. Explore your security options today.

Follow Microsoft Security

English (United States) Consumer Health Privacy Sitemap Contact Microsoft Privacy Manage cookies Terms of use Trademarks Safety & eco Recycling About our ads