Microsoft Says New Cybersecurity AI Model Helps MDASH Score 95.95% at Half the Cost

Summary: Microsoft launched MAI-Cyber-1-Flash, its first cybersecurity-specific model, inside its MDASH vulnerability identification and remediation harness, and says MDASH running it alongside GPT-5.4 scored 95.95% on CyberGym at 50% less cost than its current best MDASH mix of GPT-5.4, GPT-5.4 mini, and GPT-5.3 Codex. The model card describes a sparse mixture-of-experts transformer with 137 billion total parameters, five billion active, and a 256,000-token context window, designed to handle up to 90% of MDASH tasks with GPT-5.4 reserved for the hardest 10%, available only to approved MDASH customers through an Azure AI Foundry private preview.
Bottom line: CyberGym's public leaderboard did not list the 95.95% result when checked on July 28, 2026 - it showed Wiz's Atlas first at 90.9% and Microsoft's May 12 MDASH submission at 88.4% - and Microsoft's materials do not say whether the result was submitted, nor disclose the token use, call volume, latency, task mix, or compute behind the 50% cost claim.

Read the original at The Hacker News ↗