Try in Splunk Security Cloud


The following analytic identifies users who are unsually modifying security group in your cloud enriovnment,focusing on actions such as modifications, deletions, or creations performed by users over 30-minute intervals. Analyzing patterns of modifications to security groups can help in identifying anomalous behavior that may indicate a compromised account or an insider threat.
The detection calculates the standard deviation for each host and leverages the 3-sigma statistical rule to identify an unusual number of users. To customize this analytic, users can try different combinations of the bucket span time and the calculation of the upperBound field. This logic can be used for real time security monitoring as well as threat hunting exercises.
This detection will only trigger on all user and service accounts that have created/modified/deleted a security group .
The analytics returned fields allow analysts to investigate the event further by providing fields like source ip and values of the security objects affected.

  • Type: Anomaly
  • Product: Splunk Enterprise, Splunk Enterprise Security, Splunk Cloud
  • Datamodel: Change
  • Last Updated: 2024-02-21
  • Author: Bhavin Patel, Splunk
  • ID: cfe7cca7-2746-4bdf-b712-b01ed819b9de




ID Technique Tactic
T1578.005 Modify Cloud Compute Configurations Defense Evasion
Kill Chain Phase
  • Exploitation
  • DE.AE
  • CIS 10
| tstats dc(All_Changes.object) as unique_security_groups values(All_Changes.src) as src values(All_Changes.user_type) as user_type values(All_Changes.object_category) as object_category values(All_Changes.object) as objects values(All_Changes.action) as action  values(All_Changes.user_agent) as user_agent values(All_Changes.command) as command from datamodel=Change WHERE All_Changes.object_category = "security_group" (All_Changes.action = modified OR All_Changes.action = deleted OR All_Changes.action = created)  by All_Changes.user  _time span=30m 
|  `drop_dm_object_name("All_Changes")` 
| eventstats avg(unique_security_groups) as avg_changes , stdev(unique_security_groups) as std_changes by user 
| eval upperBound=(avg_changes+std_changes*3) 
| eval isOutlier=if(unique_security_groups > 2 and unique_security_groups >= upperBound, 1, 0) 
| where isOutlier=1
| `cloud_security_groups_modifications_by_user_filter`


The SPL above uses the following Macros:

:information_source: cloud_security_groups_modifications_by_user_filter is a empty macro by default. It allows the user to filter out any results (false positives) without editing the SPL.

Required fields

List of fields required to use this analytic.

  • _time
  • All_Changes.object_id
  • All_Changes.action
  • All_Changes.status
  • All_Changes.object_category
  • All_Changes.user

How To Implement

This search requries the Cloud infrastructure logs such as AWS Cloudtrail, GCP Pubsub Message logs, Azure Audit logs to be ingested into an accelerated Change datamodel. It is also recommended that users can try different combinations of the bucket span time and outlier conditions to better suit with their environment.

Known False Positives

It is possible that legitimate user/admin may modify a number of security groups

Associated Analytic Story


Risk Score Impact Confidence Message
35.0 70 50 Unsual number cloud security group modifications detected by user - $user$

:information_source: The Risk Score is calculated by the following formula: Risk Score = (Impact * Confidence/100). Initial Confidence and Impact is set by the analytic author.


Test Dataset

Replay any dataset to Splunk Enterprise by using our tool or the UI. Alternatively you can replay a dataset into a Splunk Attack Range

source | version: 1