Try in Splunk Security Cloud

Description

This search identifies users who have entered their passwords in username fields. This is done by looking for failed authentication attempts using usernames with a length longer than 7 characters and a high Shannon entropy, and looks for the next successful authentication attempt from the same source system to the same destination system as the failed attempt.

  • Type: Hunting
  • Product: Splunk Enterprise, Splunk Enterprise Security, Splunk Cloud
  • Datamodel: Authentication
  • Last Updated: 2022-05-11
  • Author: Mikael Bjerkeland, Splunk
  • ID: 5ced34b4-ab32-4bb0-8f22-3b8f186f0a38

Annotations

ATT&CK
ID Technique Tactic
T1078.003 Local Accounts Defense Evasion, Initial Access, Persistence, Privilege Escalation
T1552.001 Credentials In Files Credential Access
Kill Chain Phase
  • Reconnaissance
NIST
  • DE.CM
CIS20
  • CIS 3
  • CIS 5
  • CIS 16
CVE
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
| tstats `security_content_summariesonly` earliest(_time) AS starttime latest(_time) AS endtime latest(sourcetype) AS sourcetype values(Authentication.src) AS src values(Authentication.dest) AS dest count FROM datamodel=Authentication WHERE nodename=Authentication.Failed_Authentication BY "Authentication.user" 
| `drop_dm_object_name(Authentication)` 
| lookup ut_shannon_lookup word AS user 
| where ut_shannon>3 AND len(user)>=8 AND mvcount(src) == 1 
| sort count, - ut_shannon 
| eval incorrect_password=user 
| eval endtime=endtime+1000 
| map maxsearches=70 search="
| tstats `security_content_summariesonly` earliest(_time) AS starttime latest(_time) AS endtime latest(sourcetype) AS sourcetype values(Authentication.src) AS src values(Authentication.dest) AS dest count FROM datamodel=Authentication WHERE nodename=Authentication.Successful_Authentication Authentication.src=\"$src$\" Authentication.dest=\"$dest$\" sourcetype IN (\"$sourcetype$\") earliest=\"$starttime$\" latest=\"$endtime$\" BY \"Authentication.user\" 
| `drop_dm_object_name(\"Authentication\")` 
| `potential_password_in_username_false_positive_reduction` 
| eval incorrect_password=\"$incorrect_password$\" 
| eval ut_shannon=\"$ut_shannon$\" 
| sort count" 
| where user!=incorrect_password 
| outlier action=RM count 
| `potential_password_in_username_filter`

Macros

The SPL above uses the following Macros:

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

Required field

  • Authentication.user
  • Authentication.src
  • Authentication.dest
  • sourcetype

How To Implement

To successfully implement this search, you need to have relevant authentication logs mapped to the Authentication data model. You also need to have the Splunk TA URL Toolbox (https://splunkbase.splunk.com/app/2734/) installed. The detection must run with a time interval shorter than endtime+1000.

Known False Positives

Valid usernames with high entropy or source/destination system pairs with multiple authenticating users will make it difficult to identify the real user authenticating.

Associated Analytic story

RBA

Risk Score Impact Confidence Message
21.0 30 70 Potential password in username ($user$) with Shannon entropy ($ut_shannon$)

: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.

Reference

Test Dataset

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 | version: 1