Microsoft has introduced its first cybersecurity-focused AI model, MAI-Cyber-1-Flash, alongside a new platform called Perception, marking a notable step in its expanding security strategy.
The company says the model is designed to uncover difficult vulnerabilities in complex codebases and support MDASH, Microsoft's system for identifying and fixing software flaws. According to Microsoft, the model is built for both strong performance and lower operating cost compared with rival systems.
Perception is designed to bring agentic AI into security operations by coordinating red, blue, and green teams. In Microsoft's framework, red teams simulate attack scenarios, blue teams detect and prioritize issues, and green teams carry out corrective actions. The goal is to speed up workflows that often require extensive manual effort.
Microsoft security leaders said the platform can help organizations respond to AI-driven threats with AI-powered defenses, matching the pace and scale of modern security challenges. The tools are scheduled to enter preview on November 3.
With this launch, Microsoft is positioning itself in a fast-growing field where AI is increasingly being used to strengthen digital protection and automate response. The move suggests a future where cybersecurity becomes faster, more adaptive, and more deeply integrated with intelligent systems.
