Microsoft Unveils MDASH, a Multi‑Model AI System for Vulnerability Discovery
Enable the MDASH pilot for your environment and configure AI agents to match your vulnerability taxonomy.
Enable the MDASH pilot for your environment and configure AI agents to match your vulnerability taxonomy.
Summary
Microsoft has introduced MDASH, a multi‑model AI‑driven system designed to accelerate vulnerability discovery and remediation at scale. MDASH, short for Multi‑Model Agentic Scanning Harness, is model‑agnostic and deploys bespoke AI agents tailored to specific vulnerability categories. The system is currently in a limited private preview, with select customers testing its capabilities in real‑world environments. MDASH integrates with existing security tooling to provide automated scanning, risk scoring, and remediation guidance. By leveraging multiple AI models, MDASH aims to reduce false positives and increase coverage across complex codebases. Microsoft plans to expand the preview to a broader audience once the platform stabilizes. The initiative reflects the growing trend of AI‑assisted security operations. MDASH is expected to become a key component of Microsoft's security portfolio for enterprise customers.
Key changes
- MDASH is a multi‑model AI system for vulnerability discovery.
- It uses bespoke AI agents tailored to specific vulnerability categories.
- The system is model‑agnostic and integrates with existing security tooling.
- MDASH is in a limited private preview with select customers.
- Microsoft plans to broaden the preview after stabilization.