Security

Microsoft MDASH Security AI Finds Windows Vulnerabilities

3 min read

Summary

Microsoft says its MDASH multi-agent AI security system has moved from research into production workflows across Windows, Azure, and identity engineering teams. The platform now feeds validated findings into GitHub Advanced Security, Azure DevOps, and Microsoft Defender, helping teams discover and remediate high-impact vulnerabilities earlier in the development lifecycle.

Need help with Security?Talk to an Expert

Introduction

Microsoft is pushing AI deeper into its internal security engineering with codename MDASH, a multi-model agentic scanning system designed to find and help remediate software vulnerabilities at enterprise scale. For IT and security leaders, the announcement matters because it shows how Microsoft is using AI not just for detection, but to shift vulnerability discovery earlier into the software development lifecycle.

What’s new with MDASH

Since its initial launch, Microsoft says MDASH is now being used in active engineering workflows across:

  • Windows
  • Azure core infrastructure
  • Identity systems
  • Hyper-V and the Windows kernel
  • Active Directory Domain Services

Rather than acting as a standalone scanner, MDASH integrates into existing security and DevSecOps tools:

  • GitHub Advanced Security for inline code scanning alerts and pull request visibility
  • Azure DevOps for build gating and remediation work items
  • Microsoft Defender for prioritization alongside threat intelligence and runtime signals

Microsoft also shared that the latest MDASH version reached 96.5% on the CyberGym benchmark for “any crash,” reflecting improvements in the early prepare and scan phases of the pipeline.

Vulnerabilities Microsoft says MDASH helped uncover

Microsoft highlighted multiple Patch Tuesday discoveries across critical Windows components, including:

  • Hyper-V remote code execution flaws
  • Windows kernel remote code execution vulnerabilities
  • Active Directory Domain Services vulnerabilities
  • HTTP.sys and Remote Desktop Client flaws
  • DNS Client elevation of privilege issues
  • DHCP Client information disclosure bugs

Several listed CVEs carry high severity scores, including CVE-2026-45657 and CVE-2026-47291, both rated 9.8 CVSS.

Why this matters for IT administrators

For security teams and administrators, the biggest takeaway is not just the benchmark score. It is that Microsoft is embedding AI-driven vulnerability discovery into the same pipelines developers already use, making findings more actionable and less likely to sit in backlogs.

This approach could improve:

  • Faster identification of exploitable flaws before release
  • Better prioritization of code-level security issues
  • Tighter integration between development, security, and operations teams
  • More proactive protection for platforms many enterprises depend on

Next steps

IT pros should:

  1. Review this month’s Patch Tuesday updates, especially for Windows, Hyper-V, AD DS, and HTTP.sys.
  2. Track Microsoft’s MDASH preview and related security tooling announcements.
  3. Consider how AI-assisted code scanning and DevSecOps workflows could fit into internal application security programs.

Microsoft’s broader message is clear: security teams need to operate at AI speed, and vulnerability management is becoming more integrated, automated, and pipeline-driven.

Need help with Security?

Our experts can help you implement and optimize your Microsoft solutions.

Talk to an Expert

Stay updated on Microsoft technologies

Microsoft SecurityMDASHWindows vulnerabilitiesDevSecOpsGitHub Advanced Security

Related Posts

Security

Microsoft Sentinel Named a 2026 SIEM Leader

Microsoft has been named a Leader in the 2026 Gartner Magic Quadrant for SIEM, highlighting the continued momentum behind Microsoft Sentinel. The announcement also underscores Microsoft's push toward integrated, AI-powered security operations through deep Sentinel integration with the new Integrated Security Operations Center in Microsoft Defender.

Security

Microsoft Perception GA: Agentic Security at Ignite

Microsoft has made Microsoft Perception generally available, introducing an agentic security platform that coordinates red, blue, and green team agents through automated playbooks. Announced at Ignite 2026, the update also expands Defender and Sentinel capabilities to help security teams detect, investigate, and remediate AI-speed threats while keeping administrators in control.

Security

Device Code Phishing Risks for AI Coding Agents

Microsoft warns that device code phishing can be used against AI coding agents such as GitHub Copilot CLI, not just user mailboxes or cloud apps. In a controlled lab, a single approved device code sign-in could enable remote session steering, memory poisoning, synced session exposure, and private repository access depending on account permissions and enabled features.

Security

Microsoft 365 Outage Phishing Campaign Targets Users

Microsoft Threat Intelligence reported a high-volume phishing campaign that impersonates Microsoft 365 service-health notifications to trick users into clicking malicious links. The campaign matters because it uses outage-themed urgency and recipient-specific tracking to target multiple industries, increasing the risk of credential theft and follow-on account compromise.

Security

SCKit ScreenConnect Attack: MSPs Must Hunt Now

Microsoft Threat Intelligence confirmed a campaign abusing ConnectWise ScreenConnect to deploy unauthorized remote-access clients and the SCKit worm on MSP-managed endpoints. The activity can also target technician sessions, making this a high-priority incident for defenders because removing a rogue client alone does not eliminate persistence or invalidate exposed sessions.

Security

AI Agent Supply Chain Attacks Target Coding Tools

Microsoft Security researchers detailed two post-training supply chain attack paths that can backdoor coding and terminal agents: dataset poisoning and model poisoning. The findings matter because compromised agents can misuse tools, execute commands, and access sensitive environments even when prompts appear benign, reinforcing the need for stronger controls around training data, model provenance, and tool permissions.