Detection: Cloud Security Groups Modifications by User

Description

The following analytic identifies unusual modifications to security groups in your cloud environment by users, focusing on actions such as modifications, deletions, or creations over 30-minute intervals. It leverages cloud infrastructure logs and calculates the standard deviation for each user, using the 3-sigma rule to detect anomalies. This activity is significant as it may indicate a compromised account or insider threat. If confirmed malicious, attackers could alter security group configurations, potentially exposing sensitive resources or disrupting services.

Search

1
2| 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 
3|  `drop_dm_object_name("All_Changes")` 
4| eventstats avg(unique_security_groups) as avg_changes , stdev(unique_security_groups) as std_changes by user 
5| eval upperBound=(avg_changes+std_changes*3) 
6| eval isOutlier=if(unique_security_groups > 2 and unique_security_groups >= upperBound, 1, 0) 
7| where isOutlier=1
8| `cloud_security_groups_modifications_by_user_filter`

Data Source

Name	Platform	Sourcetype	Source	Supported App
AWS CloudTrail	AWS	'aws:cloudtrail'	'aws_cloudtrail'	N/A

Macros Used

| cloud_security_groups_modifications_by_user_filter | search * |

Annotations

Default Configuration

This detection is configured by default in Splunk Enterprise Security to run with the following settings:

Implementation

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

• Suspicious Cloud User Activities

Risk Based Analytics (RBA)

References

• https://attack.mitre.org/techniques/T1578/005/

Detection Testing

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

Source: GitHub | Version: 2