New Delhi: Microsoft has officially introduced its Project Perception; it is an AI-powered agentic security system that is specially designed to help organizations defend against AI-driven cyber threats. This platform also combines specialized AI agents, security context, AI models, and automated workflows to turn security signals into real-time protections while keeping humans in control of critical decisions.
Project Perception uses the coordinated workforce of specialized AI agents that analyze an organization’s security environment and work together to improve its overall security posture. However, this latest system includes three categories of agents:
- Red Team: These agents will simulate the attacker behavior to identify potential attack paths and vulnerabilities before they can be exploited.
- Blue Team: These agents investigate security incidents, analyze contextual information, and determine which risks require more attention.
- Green Team: These agents take more corrective actions to remediate identified issues and strengthen an organization’s security posture.
These agents operate as a continuous closed-loop system that discovers, evaluates, and responds to security risks rather than simply generating alerts. Project Perception is built on Microsoft’s latest Cyber Stack, which also brings together signals, security context, AI models, orchestration, and automated actions into a continuously learning security system.
Signals and sensors that collect security data across identities, endpoints, apps, data, cloud environments and AI systems. Security context which connects assets, identities, relationships, risks and activities to offer AI agents with shared, token-efficient understanding of an organization’s environment without repeatedly processing raw security data. A multi-modal AI architecture that combines frontier AI models with specialized cybersecurity models, by selecting the most suitable model for each task based on factors including quality, reliability, latency and cost.
An orchestration harness that coordinates AI models and specialized agents across security workflows. Actuators that translate AI-generated decisions into action across Microsoft Security products, by allowing organizations to reduce the risks instead of only identifying them.
Microsoft also unveiled MAI-Cyber-1-Flash; it is a specialized cybersecurity AI model that is initially being deployed within MDASH, Microsoft’s software vulnerability management system. MDASH, powered by the MAI-Cyber-1-Flash, achieved 96% on the CyberGym benchmark, which Microsoft stated is 12% points higher than Mythos. The configuration delivers nearly 50% lower operating costs than the current MDASH configuration and plans to expand MAI-Cyber-1-Flash to additional security workflows beyond software vulnerability management.
It will initially be available via Microsoft Defender, with support for additional Microsoft Security products planned in the future. Microsoft stated Project Perception will going to enter public preview on August 3.
