Microsoft launched its first artificial intelligence model specifically for cybersecurity, MAI-Cyber-1-Flash, and a new AI-powered platform named Perception on Monday in San Francisco. The move positions the tech giant to compete directly with major AI companies in the cybersecurity sector.

Microsoft's New AI Cybersecurity Tools

MAI-Cyber-1-Flash is designed to identify difficult vulnerabilities within complex codebases. This model integrates with MDASH, Microsoft’s system dedicated to finding and fixing software weaknesses.

The Perception platform employs teams of AI agents. These agents automate security tasks, including identifying and fixing bugs. Perception can also connect with the MDASH system.

Performance and Competitive Claims

Microsoft claims MAI-Cyber-1-Flash outperforms competing models on an established AI cybersecurity benchmark, offering greater power at a lower cost. Mustafa Suleyman, CEO of Microsoft AI, stated the system is being shipped into production immediately.

Suleyman highlighted the system's benchmark performance: "We have MAI-1 Cyber Flash binded [sic] with GPT 5.4 inside of the MDASH harness — which beats out Gemini, GPT 5.5 Cyber, GPT 5.6 Sol, and Mythos 5 on Cyber Gym, which is the primary benchmark that we all use. The golden benchmark."

Defending Against AI-Powered Attacks

Microsoft's vice president for security, Hayete Gallot, described Perception as a vital defense mechanism. She explained it allows enterprises to "defend against AI with AI at the scale and speed that the attackers have." This comes as cybercriminals increasingly leverage AI in their attacks.

Perception utilizes agentic teams: red teams simulate potential attacks and threat actors, blue teams detect and triage existing bugs, and green teams implement corrective actions. Dave Weston, lead engineer for Perception, noted the platform significantly speeds up security work. He stated, "We’ve gone from this taking hours and hours of manual work... and in minutes, we have a fix for all of this."

Market Entry and Availability

Microsoft's new security tools will enter a competitive AI cybersecurity market. The company announced the tools will be available in preview starting November 3. Earlier this year, competitors like Anthropic and OpenAI also launched their own AI security platforms.