Security Arsenal is tracking a sophisticated espionage campaign targeting Central Asian government entities. Threat actors attributed to a Chinese-speaking cluster are actively deploying two distinct malware families: OctLurk and SilkLurk.
This campaign represents a shift in tactics for the region, leveraging highly customized loaders and stealthy backdoors designed for long-term persistence within sensitive networks. Given the high-value nature of government data and the potential for lateral movement to critical infrastructure, immediate defensive posture adjustments are required.
Technical Analysis
The Malware Framework
- OctLurk (The Loader): Analysis of recent samples indicates OctLurk serves as the initial access vector and loader. It is typically delivered via spear-phishing attachments (weaponized documents) or through web shells planted on internet-facing web servers. OctLurk is designed to bypass EDR through process hollowing and direct system calls. It decrypts and loads the second-stage payload in memory to avoid disk signatures.
- SilkLurk (The Backdoor): Once established, SilkLurk provides the actor with remote control capabilities. It utilizes a custom protocol over HTTPS to blend in with legitimate web traffic. Notable features include credential harvesting, screenshot capture, and the ability to proxy traffic for lateral movement.
Attack Chain
- Initial Access: Spear-phishing or exploitation of public-facing web applications.
- Execution: OctLurk loader executes, decrypting the payload in memory.
- Persistence: SilkLurk establishes persistence via Windows Registry run keys or WMI Event Subscriptions.
- C2 Communication: Beaconing to actor-controlled infrastructure using valid SSL/TLS certificates.
- Objective: Exfiltration of sensitive diplomatic documents and credential dumping.
Affected Platforms
- Operating Systems: Microsoft Windows 10 and Windows Server 2019/2022.
- Vector: Phishing (Rich Text Format/Office), Web Shells (IIS/Apache).
Detection & Response
SIGMA Rules
The following Sigma rules are designed to detect the specific behaviors associated with the OctLurk loader and SilkLurk persistence mechanisms.
---
title: Potential OctLurk Loader Process Injection
id: a1b2c3d4-5678-49ef-a1b2-c3d4e5f6a7b8
status: experimental
description: Detects suspicious process hollowing behavior often associated with the OctLurk loader, specifically targeting Windows system binaries.
references:
- https://thehackernews.com/2026/08/suspected-chinese-speaking-hackers.html
author: Security Arsenal
date: 2026/08/15
tags:
- attack.defense_evasion
- attack.t1055.012
logsource:
category: process_creation
product: windows
detection:
selection:
ParentImage|endswith:
- '\winword.exe'
- '\excel.exe'
Image|endswith:
- '\svchost.exe'
- '\dllhost.exe'
CommandLine|contains:
- '-inject'
- '-load'
condition: selection
falsepositives:
- Legitimate administration tools
level: high
---
title: SilkLurk Persistence via WMI Event Consumer
id: b2c3d4e5-6789-49ef-a1b2-c3d4e5f6a7b9
status: experimental
description: Detects the creation of WMI event consumers used by SilkLurk for persistence, characterized by specific script patterns.
references:
- https://thehackernews.com/2026/08/suspected-chinese-speaking-hackers.html
author: Security Arsenal
date: 2026/08/15
tags:
- attack.persistence
- attack.t1546.003
logsource:
category: wmi_event
product: windows
detection:
selection:
EventID: 10 # Event Filter created
Operation: Created
Query|contains:
- 'ActiveScriptEventConsumer'
- 'CommandLineEventConsumer'
ScriptText|contains:
- 'powershell -enc'
- 'Invoke-Expression'
condition: selection
falsepositives:
- Legitimate software installation
level: medium
KQL (Microsoft Sentinel / Defender)
This query hunts for the network artifacts and process execution patterns indicative of the SilkLurk beaconing activity.
// Hunt for SilkLurk C2 Traffic and OctLurk Execution
let TimeFrame = 1d;
DeviceProcessEvents
| where Timestamp >= ago(TimeFrame)
// Look for common OctLurk loader decoy processes
| where ProcessCommandLine has "-enc" and ProcessCommandLine has "powershell"
| extend ProcessName = tostring(FolderPath), Account = tostring(AccountName)
| join kind=inner (
DeviceNetworkEvents
| where Timestamp >= ago(TimeFrame)
// SilkLurk often uses high entropy User-Agent strings or specific intervals
| where RemotePort == 443
| where InitiatingProcessFileName in ("powershell.exe", "cmd.exe", "wscript.exe")
| summarize ConnectionCount = count(), MaxSentBytes = max(SentBytes) by DeviceId, InitiatingProcessAccountId, RemoteUrl
) on DeviceId, InitiatingProcessAccountId
| project Timestamp, DeviceName, Account, ProcessName, ProcessCommandLine, RemoteUrl, ConnectionCount, MaxSentBytes
| where ConnectionCount > 10 and MaxSentBytes < 5000 // Consistent small beacons
Velociraptor VQL
This Velociraptor artifact hunts for the file system artifacts and persistence mechanisms left behind by OctLurk.
-- Hunt for OctLurk and SilkLurk Artifacts
SELECT
OSPath.Basename as FileName,
OSPath.Path as FilePath,
Size,
Mtime as ModifiedTime,
Atime as AccessedTime,
Ctime as CreatedTime,
Mode
FROM glob(globs="C:\Windows\Temp\**\*.tmp")
WHERE FileName =~ 'octlurk'
OR FileName =~ 'silk'
OR FileName =~ 'loader'
UNION ALL
-- Hunt for Suspicious Registry Persistence
SELECT
key.path as KeyPath,
key.value as ValueName,
key.data as ValueData,
key.mtime as ModifiedTime
FROM read_reg_key(globs="HKEY_LOCAL_MACHINE\\SOFTWARE\\Microsoft\\Windows\\CurrentVersion\\Run\\*")
WHERE ValueData =~ 'powershell'
AND ValueData =~ '-enc'
Remediation Script (PowerShell)
This script assists in identifying potential OctLurk processes and SilkLurk persistence mechanisms on an endpoint.
# OctLurk and SilkLurk Identification Script
# Requires Administrator Privileges
Write-Host "[+] Scanning for OctLurk and SilkLurk indicators..." -ForegroundColor Cyan
# 1. Check for Suspicious Processes (OctLurk Loader Patterns)
$suspiciousProcesses = Get-Process | Where-Object {
$_.ProcessName -match "svchost" -or
$_.ProcessName -match "dllhost"
} | Where-Object {
$_.MainWindowTitle -eq "" -and
$_.StartTime -lt (Get-Date).AddHours(-24)
}
if ($suspiciousProcesses) {
Write-Host "[!] Alert: Found potentially spoofed system processes:" -ForegroundColor Red
$suspiciousProcesses | Select-Object ProcessName, Id, Path, StartTime | Format-Table
} else {
Write-Host "[-] No suspicious loader processes detected." -ForegroundColor Green
}
# 2. Check Registry Run Keys for SilkLurk Persistence
$runKeys = @(
"HKLM:\SOFTWARE\Microsoft\Windows\CurrentVersion\Run",
"HKCU:\SOFTWARE\Microsoft\Windows\CurrentVersion\Run"
)
Write-Host "[+] Scanning Registry Run Keys for encoded commands..." -ForegroundColor Cyan
foreach ($key in $runKeys) {
if (Test-Path $key) {
Get-Item $key | Select-Object -ExpandProperty Property | ForEach-Object {
$propValue = (Get-ItemProperty -Path "$key" -Name $_).$_
if ($propValue -match "powershell" -and $propValue -match "-enc|-e|-encodedcommand") {
Write-Host "[!] Suspicious persistence found in $key" -ForegroundColor Red
Write-Host " Name: $_" -ForegroundColor Yellow
Write-Host " Value: $propValue" -ForegroundColor Yellow
}
}
}
}
Write-Host "[+] Scan complete." -ForegroundColor Cyan
Remediation
To neutralize the OctLurk and SilkLurk threat, security teams must execute a multi-layered remediation strategy:
- Isolate Compromised Systems: Immediately isolate hosts identified by the detection rules from the network to prevent lateral movement and data exfiltration.
- Terminate Malicious Processes: Kill processes identified as OctLurk loaders. Be aware that they may be masquerading as legitimate system processes (e.g.,
svchost.exerunning from an incorrect path). - Remove Persistence Mechanisms: Delete identified Registry keys, WMI event consumers, and scheduled tasks associated with SilkLurk.
- Credential Reset: Assume credential theft has occurred. Force a password reset for all accounts used on compromised systems, specifically privileging domain admin and service accounts.
- Network Blocking: Update firewall and web proxy rules to block identified C2 IP addresses and domains associated with the SilkLurk malware.
- Hunt for Web Shells: If the initial access vector was web-facing, conduct a thorough scan of web server logs (IIS/Apache) for suspicious HTTP requests indicative of web shell uploads.
Related Resources
Security Arsenal Healthcare Cybersecurity AlertMonitor Platform Book a SOC Assessment healthcare Intel Hub
Is your security operations ready?
Get a free SOC assessment or see how AlertMonitor cuts through alert noise with automated triage.