Platform: Windows Server 2019, 2022 and 2025, with the version and hosting restrictions described below.
Use audit mode to learn about a specific behaviour
Attack surface reduction rules target risky behaviours, not merely known malicious file signatures. An application can trigger a rule while performing a legitimate but security-sensitive operation. Audit mode records what a rule would affect without applying its block action, providing evidence for a controlled enforcement decision. It is not a statement that the recorded behaviour is harmless.
For managed Netcloud24 servers, the team reviews those events with the application owner before changing enforcement. Customers should not need to disable endpoint protection to keep an application running. This workflow pilots one supported rule, preserves other configured rules and establishes a route from audit evidence to a narrow, documented decision.
Confirm the protection prerequisites and management owner
Check the server release, Defender platform and rule support. Windows Server 2019 and later support relevant ASR capabilities, but individual rule prerequisites differ. Some rely on cloud protection or particular application behaviour. Verify the current rule reference rather than assuming a GUID supported elsewhere is effective on this server.
Confirm Microsoft Defender Antivirus is in the mode required for the rule. A passive or disabled engine is not equivalent to an active enforcement setup. Identify whether Group Policy, Intune, Configuration Manager or another approved management platform owns ASR settings. A local PowerShell change may be replaced at refresh or conflict with the source of truth.
Get-MpComputerStatus | Select-Object AMRunningMode,AntivirusEnabled,RealTimeProtectionEnabled,AMProductVersion
Get-MpPreference | Select-Object AttackSurfaceReductionRules_Ids,AttackSurfaceReductionRules_Actions
Get-DatePreserve the existing rule/action pairs. They are parallel lists, so read them together. Do not replace an entire list just to add one pilot, and do not interpret a blank selected property as a confirmed supported state without inspecting the actual object.
Choose one rule and representative workloads
The example uses the rule concerning credential stealing from LSASS, identified by 9e6c4e1f-7d60-472f-ba1a-a39ef669e4b2. Review its current Microsoft description, supported platforms and known interactions, including existing LSA protection. A credential-access event can come from security or management software and merits investigation rather than automatic dismissal.
Select a small server group with representative backup, monitoring, patching and application activity. Include scheduled and infrequent jobs in the observation window. A quiet afternoon on an idle pilot does not establish compatibility with month-end processing or a vendor update.
Define what evidence will be retained and who approves an exception or block rollout. Use application owner, security owner and device scope rather than accepting an unexplained blanket exclusion. Audit mode is the beginning of a behaviour review, not a permanent substitute for a security decision.
Add the audit setting without erasing existing rules
For a deliberately local pilot where that management method is authorised, add the selected rule in AuditMode. If it already exists, review and update its specific managed setting through the owning process instead of appending an ambiguous duplicate. Prefer the central platform for a centrally managed fleet.
$pilotRuleId = '9e6c4e1f-7d60-472f-ba1a-a39ef669e4b2'
Add-MpPreference -AttackSurfaceReductionRules_Ids $pilotRuleId `
-AttackSurfaceReductionRules_Actions AuditMode
$preferences = Get-MpPreference
for ($n=0; $n -lt $preferences.AttackSurfaceReductionRules_Ids.Count; $n++) {
[pscustomobject]@{
Rule=$preferences.AttackSurfaceReductionRules_Ids[$n]
Action=$preferences.AttackSurfaceReductionRules_Actions[$n]
}
}Set-MpPreference with a partial ASR list can replace more configuration than intended. Keep that distinction explicit in scripts and change review. Do not turn off tamper protection to force a local result; establish the approved management path if settings are protected.
Observe real events and retain useful context
Review Defender operational evidence during the agreed pilot. ASR audit events commonly use event 1122 and block events 1121; verify the actual records and rule identifiers on the target rather than treating the number alone as proof of this rule. Central telemetry depends on the configured security service and licensing.
Get-WinEvent -FilterHashtable @{
LogName='Microsoft-Windows-Windows Defender/Operational';
Id=1122;StartTime=(Get-Date).AddDays(-2)
} -MaxEvents 100 | Select-Object TimeCreated,Id,MessageCapture process path, signer, parent, target, user, rule and time as available. Protect exported data because it reveals software and activity details. Match events to the actual job or application operation. Do not run credential-dumping tools on production simply to generate a demonstration event. Use an approved safe security-validation method if a controlled trigger is required.
Investigate triggers rather than immediately excluding them
Confirm the software identity, version, signature and business purpose. A trusted vendor name does not automatically justify sensitive credential access, and an unexpected unsigned process should not be whitelisted to keep the graph quiet. Check whether an updated supported version removes the behaviour.
Determine whether the rule already overlaps another protection and whether any observed trigger would affect the business operation in block mode. Audit events indicate a candidate interaction, not guaranteed application failure. Use a controlled representative test to assess the effect before wider enforcement.
If an exception is genuinely needed, prefer the narrow supported rule-specific or applicable exclusion mechanism and document its scope and owner. Exclusion applicability varies by rule. A whole-drive or all-ASR exception can weaken unrelated controls and should not be inferred from one process event.
Move towards enforcement through an explicit decision
Review a representative observation period and agreed acceptance criteria. If the behaviour is unsupported or malicious, remediate it and coordinate enforcement. If the workload needs a justified exception, obtain approval and verify it does not accidentally bypass other rules. Roll out Block through a small ring and monitor the actual application and Defender events.
Warn mode is not universally available for every rule and platform, so do not assume it is an automatic intermediate stage. Keep the rollback method and owner ready. Test scheduled tasks, recovery tools and maintenance paths as well as ordinary user traffic, since those often run under different identities and at different times.
Restore only the pilot setting when necessary
If rollback is required, return this rule to its recorded prior state through the management source that owns it. Removing one local pilot must not remove the fleet’s existing ASR list. Re-read effective settings after policy refresh and confirm the application’s prior behaviour without suppressing evidence.
Keep the pilot findings, rule version assumptions, event examples, exceptions and enforcement decision with the server baseline. Reassess after major application or protection-platform changes. The managed result is a supported security control whose effect is understood, not an unreviewed collection of GUIDs placed in block mode.
Checking the real protection and update state of Microsoft Defender
Verify Defender’s intended active or passive role, inspect real protection and update sources, refresh intelligence, and investigate stale…
Opening one required Windows Firewall path with a narrow scope
Open one approved application flow with a named, source-scoped Windows Firewall rule, inspect effective filters and policy ownership, test…
Encrypting a server data volume with recoverable BitLocker protection
Encrypt a verified server data volume only after recovery custody and policy are ready, protect escrow secrets, observe conversion and…
Introducing Windows LAPS with controlled password retrieval
Deploy Windows LAPS through a narrow AD pilot, separate schema and account prerequisites from rotation policy, delegate retrieval and…
Building an audit policy that produces useful server evidence
Build a role-specific advanced audit policy from investigation needs, verify its effective subcategories and real events, measure…