Microsoft Debuts Its First Cybersecurity AI Model and Agentic Security Platform

Microsoft launched its first cybersecurity-specialized AI model, MAI-Cyber-1-Flash, alongside a new AI cybersecurity platform called Perception. The company described MAI-Cyber-1-Flash as built to find challenging vulnerabilities in complex codebases. The model is designed to work with MDASH, Microsoft’s harness for software vulnerability identification and remediation.

Perception is designed to deploy teams of agents that assist with and automate security workflows, including bug identification and remediation. The platform can integrate with MDASH. Microsoft said Perception uses agentic red teams, blue teams, and green teams: red teams simulate potential attacks, blue teams detect and triage bugs, and green teams take corrective actions.

Microsoft claimed MAI-Cyber-1-Flash is more powerful and cost-effective than competitor models based on performance on an established AI cybersecurity benchmark. Mustafa Suleyman, CEO of Microsoft AI, said the company is “shipping this into production immediately.” Microsoft said its new security tools will be available in preview on November 3.

The launch places Microsoft deeper into a crowded AI cybersecurity market that includes offerings from Anthropic and OpenAI. Microsoft framed Perception as a way for enterprise defenders to respond to AI-enabled attacks with AI-driven defenses. For security leaders, the practical takeaway is to watch how well Microsoft’s claims translate into real-world vulnerability discovery, prioritization, detection, posture fixes, and code fixes.