Microsoft AI has launched a brand-new internally developed cybersecurity model

July 28th, Microsoft AI launched a new self-developed network security model, MAI-Cyber-1-Flash, which operates in the MDASH platform—a multi-agent framework that can detect and remediate vulnerabilities in large-scale codebases.

MDASH stated that the MAI-Cyber-1-Flash+GPT-5.4 model combination scored 12 points higher than Mythos in the CyberGym benchmark test. Microsoft AI claims that the pricing of this model configuration provided in MDASH is 50% lower than their previous best solution (GPT-5.4 + 5.4 mini + 5.3 codex). (Jinse)