Windows Unusual NTLM Authentication Destinations By User

SPLUNK detection rule by Steven Dick · Apache-2.0

The following analytic detects when an unusual number of NTLM authentications is attempted by the same user account against multiple destinations. This activity generally results when an attacker attempts to brute force, password spray, or otherwise authenticate to numerous domain joined Windows devices using an NTLM based process/attack. This same activity may also generate a large number of EventID 4776 events as well.

name: Windows Unusual NTLM Authentication Destinations By User
id: a4d86702-402b-4a4f-8d06-9d61e6c39cad
version: 10
creation_date: '2024-03-16'
modification_date: '2026-05-13'
author: Steven Dick
status: production
type: Anomaly
description: The following analytic detects when an unusual number of NTLM authentications is attempted by the same user account against multiple destinations. This activity generally results when an attacker attempts to brute force, password spray, or otherwise authenticate to numerous domain joined Windows devices using an NTLM based process/attack. This same activity may also generate a large number of EventID 4776 events as well.
data_source:
    - NTLM Operational 8004
    - NTLM Operational 8005
    - NTLM Operational 8006
search: |
    `ntlm_audit`
    EventCode = 8004
    SChannelName=* WorkstationName=*

    ```CIM alignment, remove leading \\ from some auth attempts ```
    | eval src = replace(WorkstationName,"\\\\","")

    ``` CIM alignment```
    | eval dest = SChannelName, user = UserName

    ``` Remove NTLM auths to self, improves accuracy for certain applications ```
    | where SChannelName!=src

    | stats count min(_time) as firstTime
                  max(_time) as lastTime
                  dc(eval(upper(dest))) as unique_count by user
    | eventstats avg(unique_count) as unique_avg
                 stdev(unique_count) as unique_std

    ``` adjust formula for sensitivity```
    | eval upperBound_unique=(1+unique_avg+unique_std*3)

    | eval isOutlier=CASE(unique_count > upperBound_unique, 1, true(), 0)
    | where isOutlier==1
    | `security_content_ctime(firstTime)`
    | `security_content_ctime(lastTime)`
    | `windows_unusual_ntlm_authentication_destinations_by_user_filter`
how_to_implement: The following analytic requires that NTLM Operational logs to be imported from the environment Domain Controllers. This requires configuration of specific auditing settings, see Microsoft references for further guidance. This analytic is specific to EventID 8004~8006.
known_false_positives: Vulnerability scanners, print servers, and applications that deal with non-domain joined authentications. Recommend adjusting the upperBound_unique eval for tailoring the correlation to your environment, running with a 24hr search window will smooth out some statistical noise.
references:
    - https://attack.mitre.org/techniques/T1110/003/
    - https://techcommunity.microsoft.com/t5/ask-the-directory-services-team/ntlm-blocking-and-you-application-analysis-and-auditing/ba-p/397191
    - https://techcommunity.microsoft.com/t5/microsoft-defender-for-identity/enriched-ntlm-authentication-data-using-windows-event-8004/m-p/871827
    - https://www.varonis.com/blog/investigate-ntlm-brute-force
    - https://learn.microsoft.com/en-us/openspecs/windows_protocols/ms-nrpc/4d1235e3-2c96-4e9f-a147-3cb338a0d09f
drilldown_searches:
    - name: View the detection results for - "$user$"
      search: '%original_detection_search% | search  user = "$user$"'
      earliest_offset: $info_min_time$
      latest_offset: $info_max_time$
    - name: View risk events for the last 7 days for - "$user$"
      search: '| from datamodel Risk.All_Risk | search normalized_risk_object IN ("$user$") | stats count min(_time) as firstTime max(_time) as lastTime values(search_name) as "Search Name" values(risk_message) as "Risk Message" values(analyticstories) as "Analytic Stories" values(annotations._all) as "Annotations" values(annotations.mitre_attack.mitre_tactic) as "ATT&CK Tactics" by normalized_risk_object | `security_content_ctime(firstTime)` | `security_content_ctime(lastTime)`'
      earliest_offset: 7d
      latest_offset: "0"
intermediate_findings:
    entities:
        - field: user
          type: user
          score: 20
          message: The user [$user$] attempted $count$ NTLM authentications against $unique_count$ destinations.
analytic_story:
    - Active Directory Password Spraying
asset_type: Endpoint
mitre_attack_id:
    - T1110.003
product:
    - Splunk Enterprise
    - Splunk Enterprise Security
    - Splunk Cloud
category: endpoint
security_domain: endpoint
tests:
    - name: True Positive Test
      attack_data:
        - data: https://media.githubusercontent.com/media/splunk/attack_data/master/datasets/attack_techniques/T1110.003/ntlm_bruteforce/ntlm_bruteforce.log
          source: XmlWinEventLog:Microsoft-Windows-NTLM/Operational
          sourcetype: XmlWinEventLog
      test_type: unit

[[ currentRule.title && currentRule.title.length > 60 ? currentRule.title.slice(0, 60) + '…' : currentRule.title ]]

e0fd5b05-6aed-4cb5-b8ef-3504ed7ffc58

Propose edits, view related rules, and engage with the community through comments.

Executable code, not a declarative rule

This submission is arbitrary [[ currentRule.format ? currentRule.format.toUpperCase() : '' ]] code, not a declarative detection rule — review it as you would any script before running it, and note that it cannot be safely aggregated or auto-executed alongside YARA/Sigma/Suricata/CRS rules.

Rejected pattern present

[[ reason ]]

What does this mean?

Format & documentation suggestions

Not a security issue — this rule is valid and was accepted as-is. These are optional fields/conventions its own format recommends, mainly affecting its quality score and how well it documents itself for other reviewers.

[[ reason ]]

What do these messages mean?

[[ currentRule.title && currentRule.title.length > 60 ? currentRule.title.slice(0, 60) + '…' : currentRule.title ]]

[[ currentRule.title ]]

Community Discussion