Microsoft has launched Project Perception, a new AI-powered cybersecurity platform designed to identify vulnerabilities before attackers can exploit them. The platform uses AI agents to continuously detect risks, reason across complex environments, and help organizations take proactive action. Microsoft stated that traditional security tools cannot keep pace with AI-powered threats. Project Perception employs three types of specialized agents: Red team agents proactively search for vulnerabilities and potential attack paths; blue team agents analyze findings and assess which risks require attention; and green team agents apply corrective actions. The platform relies on real-time understanding of an organization’s users, devices, applications, relationships, and risks, and chooses different AI models based on accuracy, reliability, response speed, and cost. Microsoft emphasized that Project Perception follows its Responsible AI principles and includes enterprise-grade safeguards. The platform will be available in public preview on August 3. Additionally, Microsoft is enhancing MDASH with its MAI-Cyber-1-Flash model to improve vulnerability discovery.