Blog Post

Microsoft Defender Blog
5 MIN READ

Vulnerability shielding: Protect the time before the patch

Alon Rosental's avatar
Alon Rosental
Icon for Microsoft rankMicrosoft
Nov 17, 2026

AI has accelerated vulnerability discovery for defenders and attackers alike. An attacker can now turn a vulnerability into a working exploit within hours, but patching can’t keep pace. Our data shows that 38% of critical Patch Tuesday exposures remain unpatched for more than five days, and attackers operate in the gap between exposure and remediation.

To keep up, organizations need protection that moves as fast as critical vulnerabilities are discovered, without forcing rushed patches or disrupting the business. That’s why today we’re excited to introduce vulnerability shielding in Microsoft Defender, in preview:

Vulnerability shielding applies targeted, temporary protection to vulnerable systems in the gap between exposure and remediation, buying defenders time before they can apply a permanent patch. This AI-based capability reasons over the latest threat intelligence and each organization’s unique environment to choose the narrowest effective protection. Vulnerability shielding operates at two critical moments:

  • Proactively limit exposure: The Risk Reduction Agent in Microsoft Perception as part of the Reduce exposure risk playbook creates a tailored plan for shielding vulnerabilities when patching is not immediately available or possible. Shields are time bound, reversible, choose the path of minimal business disruption, and retire automatically once the patch lands. Nothing is implemented without your approval.
  • Protect during an active attack: When attack disruption in Microsoft Defender identifies an ongoing attack, it autonomously applies vulnerability shielding controls to exposed assets, cutting off new attack paths for adversaries and limiting the impact to your organization.

Vulnerability shielding does not replace patching. Patching is the fix; shielding is designed to protect the interim risk.

Autonomous protection enabled by a modern cyber stack

Microsoft's cyber stack connects context, agentic reasoning, and native enforcement in a continuous loop. That integration enables vulnerability shielding to move from exposure insight to targeted protection at the speed an attack demands.

  • Signals and context: Vulnerability shielding builds a deep, continuously updated exposure map of your environment across assets, identities, vulnerabilities, attack paths, and business context. This understanding helps determine where and how to disrupt potential attacks with the most effective protection, while preserving legitimate activity to minimize operational disruption.
  • Actuators: Vulnerability shielding applies hardening policies tailored to each vulnerability and device, placed at the point where an attack can be cut off most effectively. It uses native enforcement controls in Microsoft Defender including reducing network reachability, restricting access paths, and tightening permissions.
  • Agents: For preventive scenarios, green agents in Microsoft Perception can invoke vulnerability shielding when patching is not possible. They assess exposure and business impact, then recommend a coordinated plan that may combines patching, hardening, and temporary shielding.

Reduce exposure. 

Reduce your exposure: The Risk Reduction Agent in Microsoft Perception recommends tailored vulnerability management plans.
Figure 1: Reduce exposure risk playbook calling on the Risk Reduction Agent within Microsoft Perception, recommending a tailored shielding plan.

Microsoft Perception helps teams act before vulnerabilities become paths to compromise. The Reduce exposure risk playbook uses the Risk Reduction Agent to prioritize risk based on exploitability, asset criticality, attack paths, patch readiness, and business impact. It then explains what matters and recommends the safest next step for security and IT teams to review.

Vulnerability shielding is one of the capabilities the Risk Reduction Agent can invoke through the playbook. When a patch is available, the agent can coordinate patching through Microsoft Intune, with IT retaining control of deployment. When patching is not immediately possible e.g. because it could disrupt the business, the agent can recommend shielding. Shields apply targeted protection, buying teams time to work toward the permanent fix.

The agent’s workflow is as follows:

This transforms vulnerability management from a disconnected list of findings into a coordinated path from prioritization to protection. As part of public preview, the agent will provide vulnerability shielding for server-side RCE vulnerabilities, which make up 37% of vulnerabilities adversaries’ exploit.  

Once applied, the shield stays in place until the patch is fully deployed, then retires automatically. If you need to change course, you can roll it back with one click. Exclusions are managed through Microsoft Intune to keep critical devices and apps out of scope. You get room to remediate while staying in control of the business impact. Learn more about what else is new in Microsoft Perception.

Disrupt attacks:

Vulnerability shielding limits your blast radius
Figure 2. Vulnerability shielding applied to a SQL server with a known public vulnerability in attack disruption

Attack disruption in Microsoft Defender uses AI-powered analysis to detect and contain attacks at machine speed with over 99% confidence. It stops ransomware attacks in an average of 30 seconds. Every month Microsoft Defender disrupts about 68,000 attacks, protecting 580,000 devices from being encrypted. Utilizing the same confidence and technology, we are extending attack disruption to proactively shield vulnerabilities during ongoing attacks.

On by default, attack disruption now automatically contains the exploited path and shields new routes the attacker could use - boxing them in with precise protections that reduce network reachability, restrict access paths, and tighten permissions. It acts autonomously, combining the use of the latest Threat Intelligence about vulnerabilities that are being actively exploited, as well as the unique context of an organization’s environment, to always seek the narrowest effective action and limit impact on an organization’s productivity.

Crucially, the controls vulnerability shielding applies also analyze business context. Vulnerability shielding uses business context and baseline activity to understand who and what typically connects, it then applies precise network controls that block the exploit route while preserving established business traffic where possible. Security teams gain immediate protection, IT teams gain time to deploy the permanent fix safely, and finance users can keep working.

Managing vulnerability shielding

Shields are designed to be managed, not forgotten. Each one is temporary by design and retires once the patch is deployed, and every active shield is listed with the vulnerability it covers, the devices it applies to and why it was applied. You set the guardrails: exclude devices or apps that must never be touched, roll back any shield in one click, or turn shielding off. Before an attack, the Risk Reduction Agent simulates each shield plan against 90 days of real activity, so you see the expected impact before you approve. You can also extend shielding beyond published CVEs to vulnerabilities unique to your environment, such as in-house apps, legacy builds or AI-written code, and use your own detections to trigger the right shielding plan or attack disruption action.

AI has made time the new attack surface but with vulnerability shielding organizations no longer have to choose between waiting for the patch and breaking the business. Microsoft Defender closes the exploit window in minutes, keeps you protected while you remediate, and gets out of the way when the patch lands. Vulnerability shielding is giving you protected time to patch.

Learn more:

Updated Nov 17, 2026
Version 1.0
No CommentsBe the first to comment