Analytics Story: Gh0st RAT

Description

Gh0st RAT is a long-running Windows remote access trojan family known for full interactive control, surveillance, and data theft.

Variants implement a custom binary wire protocol over TCP (often high ports), peer-to-peer relaying, and modular features such as keylogging, screen and camera capture, audio recording, file management, and remote shell.

Operators frequently achieve persistence via Run keys, services, or scheduled tasks, and may load capability through side-loaded DLLs or abused LOLBins.

Because Gh0st tooling is widely shared and re-branded, detections should emphasize behavioral chains including ingress staging, non-standard process ancestry, unusual outbound sessions, and registry or service changes associated with remote access—rather than brittle file hashes alone.

Why it matters

Gh0st samples typically establish a foothold through spear-phishing, drive-by downloads, or supply-chain delivery, then unpack a loader or injector that decrypts the core implant in memory.

The implant beacons to attacker-controlled infrastructure using its proprietary framing; some builds add encryption, compression, or domain generation to resist network inspection.

On the endpoint, the malware often registers autostart mechanisms under standard persistence locations, may masquerade as legitimate software or use stolen certificates, and sometimes stages payloads under user-writable or public directories before execution.

Operational use spans credential harvesting, lateral movement as a foothold for follow-on tools, and long-term espionage.

Effective coverage combines host telemetry (process creation, module loads, WMI or service creation, and authentication events for remote access features) with firewall and proxy logs highlighting repeated connections to uncommon ports, symmetric upload/download ratios on non-web protocols, and TLS anomalies where HTTPS wrappers are used. Correlating registry edits that enable remote access or weaken authentication with subsequent interactive sessions helps distinguish Gh0st-style remote control from benign administrative activity.

Detections

Name ▲▼ Technique ▲▼ Type ▲▼
CMD Carry Out String Command Parameter Windows Command Shell Hunting
Ping Sleep Batch Command Time Based Checks Anomaly
Registry Keys Used For Persistence Registry Run Keys / Startup Folder TTP
Rundll32 Process Creating Exe Dll Files Rundll32 TTP
Windows Access Token Manipulation SeDebugPrivilege Create Process with Token Anomaly
Windows Cmdline Tool Execution From Non-Shell Process JavaScript Anomaly
Windows Hosts File Access Query Registry Anomaly
Windows Net System Service Discovery System Service Discovery Hunting
Windows Process Execution in Temp Dir Create or Modify System Process, Match Legitimate Resource Name or Location Anomaly
Windows Routing and Remote Access Service Registry Key Change Modify Registry Anomaly
Windows Rundll32 with Non-Standard File Extension Rundll32 Anomaly
Windows Service Created with Suspicious Service Name Service Execution Anomaly
Windows Service Created with Suspicious Service Path Service Execution TTP
Windows Service Creation Using Registry Entry Services Registry Permissions Weakness Anomaly
Windows Service Stop Attempt Service Stop Hunting

Data Sources

Name ▲▼ Platform ▲▼ Sourcetype ▲▼ Source ▲▼
CrowdStrike ProcessRollup2 Other crowdstrike:events:sensor crowdstrike
Sysmon EventID 1 Windows icon Windows XmlWinEventLog XmlWinEventLog:Microsoft-Windows-Sysmon/Operational
Sysmon EventID 11 Windows icon Windows XmlWinEventLog XmlWinEventLog:Microsoft-Windows-Sysmon/Operational
Sysmon EventID 13 Windows icon Windows XmlWinEventLog XmlWinEventLog:Microsoft-Windows-Sysmon/Operational
Windows Event Log Security 4663 Windows icon Windows XmlWinEventLog XmlWinEventLog:Security
Windows Event Log Security 4688 Windows icon Windows XmlWinEventLog XmlWinEventLog:Security
Windows Event Log Security 4703 Windows icon Windows XmlWinEventLog XmlWinEventLog:Security
Windows Event Log System 7045 Windows icon Windows XmlWinEventLog XmlWinEventLog:System

References


Source: GitHub | Version: 1