Microsoft Ships MAI-Cyber-1-Flash And Perception To Take On Anthropic In Security

Microsoft has launched its first cybersecurity-specialized AI model, MAI-Cyber-1-Flash, alongside Perception, an agentic security platform inside Defender.

Microsoft Ships MAI-Cyber-1-Flash And Perception To Take On Anthropic In Security

Microsoft on July 27, 2026 unveiled its first cybersecurity-specialized AI model, MAI-Cyber-1-Flash, alongside a new agentic security platform called Perception, taking a direct swing at Anthropic, OpenAI and Google in what is fast becoming the most crowded corner of enterprise AI.

MAI-Cyber-1-Flash and the MDASH harness

MAI-Cyber-1-Flash is a compact, code-tuned derivative of Microsoft's MAI-Thinking-1 family, trained in-house on the company's own exploit and remediation records. It is built to power MDASH, Microsoft's harness dedicated to finding and fixing software vulnerabilities in complex codebases. Microsoft says that pairing MAI-Cyber-1-Flash with OpenAI's GPT-5.4 inside MDASH scores 96% on the CyberGym benchmark — 12 points above Anthropic's Mythos 5 — while cutting cost in half versus the current MDASH configuration.

"We're shipping this into production immediately," Microsoft AI CEO Mustafa Suleyman said at a small San Francisco event announcing the model.

Perception's red, blue and green agents

Perception, delivered inside Microsoft Defender, coordinates three classes of agents across the security lifecycle: red agents that map attack paths and vulnerabilities, blue agents that investigate findings and determine real risk, and green agents that take corrective action and strengthen defenses. Microsoft says Perception collapses hours of manual work by security teams into minutes of automated triage and remediation, and that it can plug into the same MDASH pipeline that MAI-Cyber-1-Flash powers.

Microsoft security blog banner

Anthropic, Google and OpenAI in the crosshairs

Perception enters public preview on November 3, dropping into a field that already includes Anthropic's Mythos-based Project Glasswing and OpenAI's Daybreak. It also fits the broader Redmond push to sell defenders "AI at the scale and speed of attackers," as security VP Hayete Gallot put it, and lands alongside Microsoft's ongoing infrastructure buildout including the NVIDIA-led Open Secure AI Alliance. Microsoft's cyber push also arrives as rival platforms mint enterprise deals like AT&T's D-Wave agentic AI expansion.

Reporting based on coverage from TechCrunch and the Microsoft Blog.

Category: AI & Technology

Tags: AI Models Cybersecurity AI Foundation Models OpenAI AI Agents Anthropic

Related Articles