-> Back to Writeups
Security Incident Report

Volt Typhoon: State-Actor Intrusion & Domain Compromise

TLP:CLEAR
Report ID INC2024-0324-VT Classification Simulated / Training Environment
Analyst F. Nanos Platform TryHackMe (SIEM investigation)
Severity Critical Status Closed (Investigation Complete)
Attributed Actor Volt Typhoon (state-sponsored, living-off-the-land tradecraft)
This report documents a simulated intrusion from a training environment (TryHackMe). Specific artifact values, account names, and addresses are generalized for public publication; the analytical structure, timeline reconstruction, and findings reflect the actual investigation performed. No production, client, or employer data is included.

1. Executive Summary

Over a six-day window, an Active Directory environment was subjected to a deliberate, multi-stage intrusion consistent with the tradecraft of Volt Typhoon, a state-sponsored threat group known for targeting critical infrastructure and for heavy reliance on living-off-the-land techniques. Initial access was obtained by abusing an unauthenticated password-reset weakness in an internet-facing identity self-service portal. The attacker then created a hidden administrative backdoor, extracted the entire domain credential database, harvested credentials from memory, staged sensitive financial records, deployed web shells for durable access, and finally cleared event logs and network configuration to frustrate investigation.

Although the attacker cleared the Windows event logs at the end of the operation, telemetry forwarded to the SIEM beforehand allowed a complete, corroborated timeline to be reconstructed. The single most consequential action was the extraction of the domain credential database, which must be treated as a full domain compromise requiring a complete credential reset.

2. Scope & Methodology

The investigation was conducted in the SIEM against forwarded telemetry from the identity self-service application, Windows Management Instrumentation activity, and PowerShell pipeline logging. The core analytical challenge was reconstructing the timeline after the attacker cleared all local Windows event logs in the final phase; because critical telemetry had already been forwarded, the SIEM retained the record that the endpoint no longer held.

A defining characteristic of this intrusion was the near-exclusive use of legitimate built-in Windows tooling (the domain database utility, WMI, netsh, and the event-log utility), which blends with normal administration and defeats signature-based detection. Reconstruction therefore depended on behavioral sequencing rather than malware signatures.

3. Attack Chain

3.1   Initial Access & Backdoor

An unauthenticated password reset in an internet-facing identity self-service portal was abused to seize an existing administrative account. Approximately two minutes later, a second attacker-controlled administrative account was created, a deliberate redundancy ensuring continued access even if the first compromise were remediated.

3.2   Credential Access

Following drive enumeration of two file servers via WMI, the attacker extracted the entire Active Directory credential database using a legitimate built-in domain utility, then copied it into a web-server directory. Separately, a credential-dumping tool was pulled from an external host and run against process memory, and the registry was searched for stored credentials belonging to remote-access software. These represent two independent credential-theft vectors: the on-disk domain database and live memory.

3.3   Collection & Masquerading

The stolen credential database was compressed into a password-protected, split archive and then renamed to an image-file extension, disguising it as a harmless picture to evade content inspection and blend with web traffic. Three years of financial records were staged in a temporary directory for collection.

3.4   Persistence via Web Shells

An ASP.NET web shell was created on one server by renaming an existing file to a web-executable extension, then copied to a second server under an audit-themed name. Web shells provide a browser-accessible command channel that survives account remediation.

3.5   Anti-Forensic Cleanup

In the final phase, all Windows event logs (Application, Security, Setup, System) were cleared, and a netsh port-forwarding rule was created to relay traffic to an external address, then deleted to remove the configuration evidence. The prior forwarding of telemetry to the SIEM is precisely what preserved the investigative record through this cleanup.

4. Indicators of Compromise

Type Indicator Context
Backdoor account[redacted, attacker-created admin]Persistence / redundancy
C2 domain[redacted, actor-themed domain]Tool download source
C2 relay[redacted, external IP:port]netsh port-forward target
Domain database dump[redacted, temp path]Full credential extraction
Disguised archive[redacted, .7z renamed to image ext]Staged for exfiltration
Web shells[redacted, .aspx / .jspx]Durable access, two hosts

5. MITRE ATT&CK Mapping

Technique ID Technique Tactic
T1190Exploit Public-Facing ApplicationInitial Access
T1136 / T1078.003Create Account / Valid AccountsPersistence
T1003.003OS Credential Dumping: NTDSCredential Access
T1003.001OS Credential Dumping: LSASS MemoryCredential Access
T1552.007Credentials from Password StoresCredential Access
T1560 / T1036.005Archive Collected Data / MasqueradingCollection / Defense Evasion
T1505.003Web ShellPersistence
T1070.001Clear Windows Event LogsDefense Evasion
T1090.001Internal Proxy (netsh port-forward)Command and Control

6. Root Cause & Recommendations

The intrusion was enabled by three compounding weaknesses: an unauthenticated password reset on an internet-facing identity portal (primary root cause), absent multi-factor authentication on privileged accounts, and no alerting on high-risk built-in tooling.

  1. Patch or isolate the identity self-service portal and remove direct internet exposure where possible.
  2. Enforce multi-factor authentication on all privileged and administrative accounts.
  3. Alert on domain-database utility execution and on any access to the credential database outside sanctioned backup windows.
  4. Alert on event-log clearing and on netsh port-forward rule creation, both high-fidelity intrusion indicators.
  5. Forward all critical logs to the SIEM in near real time so local log clearing cannot erase the investigative record, the control that made this reconstruction possible.
  6. On confirmed domain-database extraction, perform a domain-wide credential reset including a double reset of the krbtgt account.

7. Investigator's Notes

The instructive element of this investigation was the living-off-the-land tradecraft. With no custom malware to key on for most stages, detection depended on recognizing that legitimate administrative tools were being used in illegitimate sequences: a domain-database utility run outside backup windows, WMI used for multi-host enumeration, and the event-log utility used to wipe history. The standing lesson is that resilient SIEM forwarding plus behavioral detection on built-in tooling, rather than signature matching, is what defeats an actor who deliberately avoids bringing their own malware.

END OF REPORT : INC2024-0324-VT
Full technical detail, exact artifact values, and challenge-identifying information intentionally generalized for public publication. Happy to walk through the complete investigation in more depth in conversation: get in touch.