The following analytic identifies the abuse of two undocumented registry keys that allow for a DLL to load into lsass.exe to potentially capture credentials. Upon successful modification of \CurrentControlSet\Services\NTDS\DirectoryServiceExtPt or \CurrentControlSet\Services\NTDS\LsaDbExtPt, a DLL either remote or local will be set as the value and load up into lsass.exe. Based on POC code a text file may be written to disk with credentials.
- Type: TTP
- Product: Splunk Enterprise, Splunk Enterprise Security, Splunk Cloud
- Datamodel: Endpoint
- Last Updated: 2022-08-22
- Author: Michael Haag, Splunk
- ID: 57fb8656-141e-4d8a-9f51-62cff4ecb82a
Kill Chain Phase
- CIS 10
1 2 3 4 5 6 | tstats `security_content_summariesonly` count min(_time) as firstTime max(_time) as lastTime FROM datamodel=Endpoint.Registry where Registry.registry_path IN ("*\\CurrentControlSet\\Services\\NTDS\\DirectoryServiceExtPt","*\\CurrentControlSet\\Services\\NTDS\\LsaDbExtPt") by Registry.registry_key_name Registry.user Registry.registry_path Registry.registry_value_data Registry.action Registry.dest Registry.process_guid | `drop_dm_object_name(Registry)` | `security_content_ctime(firstTime)` | `security_content_ctime(lastTime)` | `windows_autostart_execution_lsass_driver_registry_modification_filter`
The SPL above uses the following Macros:
windows_autostart_execution_lsass_driver_registry_modification_filter is a empty macro by default. It allows the user to filter out any results (false positives) without editing the SPL.
List of fields required to use this analytic.
How To Implement
To successfully implement this search you need to be ingesting information on process that include the name of the process responsible for the changes from your endpoints into the
Endpoint datamodel in the
Registry node. In addition, confirm the latest CIM App 4.20 or higher is installed and the latest TA for the endpoint product.
Known False Positives
False positives may be present on recent Windows Operating Systems. Filtering may be required based on process_name. In addition, look for non-standard, unsigned, module loads into LSASS. If query is too noisy, modify by adding Endpoint.processes process_name to query to identify the process making the modification.
Associated Analytic Story
|49.0||70||70||The registry values for DirectoryServiceExtPt or LsaDbExtPt were modified on $dest$.|
The Risk Score is calculated by the following formula: Risk Score = (Impact * Confidence/100). Initial Confidence and Impact is set by the analytic author.
source | version: 1