The Nosy Neighbor: How China’s LongNosedGoblin APT Uses Your IT Admin Tools Against You

A Masterclass in Living-Off-The-Land Attack Infrastructure Through Group Policy Abuse


1. Executive Summary

For years, defenders have been told to “trust built-in security tools” and to rely on centralized administration for consistency and control. LongNosedGoblin proves how dangerous that assumption has become.

This China-aligned advanced persistent threat (APT), uncovered by ESET Research in late 2025, demonstrates a quiet but deeply effective method of large-scale compromise: weaponizing Windows Group Policy itself. Instead of exploiting vulnerabilities on endpoints, the group focuses on compromising Active Directory and then using Group Policy Objects (GPOs) as a malware distribution, persistence, and re-infection mechanism.

Once domain administrator privileges are obtained, the attack no longer looks like an intrusion. It looks like normal IT operations.

This campaign marks a shift in tradecraft:
enterprise trust has become the attack surface.


2. Why This Campaign Is Different

Most APT campaigns still rely on:

  • Dropping malware through phishing
  • Exploiting exposed services
  • Evading endpoint protection

LongNosedGoblin uses those techniques only to get in the door. Once inside, they stop acting like attackers and start acting like administrators.

They:

  • Use Group Policy to deploy malware
  • Use legitimate .NET features to execute code
  • Use cloud storage platforms for command-and-control
  • Use Windows scheduled tasks for persistence

Every step blends into the background of a normal Windows enterprise.

This is why the campaign can persist for months or years without triggering alerts.


3. Initial Access and Privilege Escalation

LongNosedGoblin does not depend on a single entry vector. Observed access paths include:

  • Spear-phishing of government IT staff
  • Compromised VPN credentials
  • Abuse of exposed RDP or web portals

The real danger begins after initial access.

3.1 Privilege Escalation Paths

Once a foothold exists, attackers move quickly to obtain domain-level privileges, often abusing weaknesses that exist in many environments by default.

CVE-2025-21293 — Network Configuration Operators Abuse

  • Allows SYSTEM-level code execution
  • Exploits excessive registry permissions
  • Particularly dangerous because the group exists by default

CVE-2025-29810 — Active Directory ACL Abuse

  • Exploits misconfigured AD object permissions
  • Turns “modify permissions” into full admin access
  • Often invisible without detailed AD auditing

BadSuccessor (dMSA Abuse in Windows Server 2025)

  • No CVE, no patch at time of discovery
  • Allows privilege inheritance during account migration
  • Affects the majority of real-world environments

Once any of these succeed, Group Policy becomes fully controllable.


4. Weaponizing Group Policy

Group Policy is designed to:

  • Run automatically
  • Execute with elevated privileges
  • Apply repeatedly
  • Be trusted implicitly

LongNosedGoblin abuses all four.

Attackers create or modify GPOs to:

  • Deploy malware during startup or logon
  • Create scheduled tasks
  • Drop registry-based payloads
  • Reinstall malware after removal

From a logging perspective, the activity is indistinguishable from legitimate policy updates.


5. The Nosy Malware Ecosystem

LongNosedGoblin’s tooling is modular and purpose-built for espionage.

5.1 NosyHistorian — Browser Intelligence Collection

Purpose:
Reconnaissance and user profiling

What it collects:

  • Browser history from Chrome, Edge, Firefox
  • All user profiles on the system

Key behavior:

  • Deployed as History.ini
  • Runs via Group Policy
  • Uploads data to internal SMB shares
  • Avoids external traffic entirely

Why this matters:
Browser history reveals:

  • Government portals
  • Policy research
  • Inter-agency collaboration
  • High-value individuals

Strong IOC:

<username>_<hostname>_(chrome|edge|firefox)_History

5.2 NosyDoor — Primary Backdoor

NosyDoor is the backbone of LongNosedGoblin’s persistence.

Stage 1 — Dropper

  • Delivered via Registry.pol
  • Drops files into: C:\Windows\Microsoft.NET\Framework\
  • Creates scheduled tasks with cloud-themed names

Dropped artifacts:

  • SharedReg.dll
  • netfxsbs9.hkf
  • log.cached
  • UevAppMonitor.exe.config

Stage 2 — AppDomainManager Injection

This is one of the most important techniques in the campaign.

  • Abuses legitimate .NET AppDomainManager functionality
  • Forces Windows to load a malicious DLL inside a trusted process
  • Uses UevAppMonitor.exe as a host

Why it’s effective:

  • Legitimate binary name
  • No suspicious parent/child process chains
  • Often bypasses AMSI and Defender heuristics
  • Appears as standard system behavior

Stage 3 — Backdoor Capabilities

NosyDoor supports:

  • Shell command execution
  • File exfiltration
  • File deletion
  • Registry modification
  • Process control

Command-and-Control:

  • Microsoft OneDrive
  • Google Drive
  • Yandex Disk (EU-focused operations)

All communication uses HTTPS and legitimate APIs.


5.3 NosyStealer — Browser Credential Theft

Targets:

  • Saved passwords
  • Cookies (session hijacking)
  • Certificates
  • Cached credentials

Execution chain:

  1. Configuration download (Google Docs)
  2. Data collection
  3. Encryption and archiving
  4. Upload to Google Drive

To defenders, this looks like normal cloud usage.


5.4 NosyDownloader — Fileless Payload Loader

  • Executes payloads entirely in memory
  • Leaves minimal disk artifacts
  • Enables rapid tool updates
  • Used selectively on high-value systems

5.5 NosyLogger — Keylogging and Clipboard Capture

  • Modified DuckSharp keylogger
  • Captures:
    • Keystrokes
    • Window titles
    • Clipboard contents
  • AES-encrypted logs
  • Periodic exfiltration

This gives attackers near-total visibility into user activity.


5.6 Reverse SOCKS5 Proxy

  • Enables network pivoting
  • Allows access to internal services
  • Supports lateral movement without direct exposure

6. MITRE ATT&CK Mapping

TacticTechniques
Initial AccessT1566.002 (Phishing), T1199 (Trusted Relationship)
ExecutionT1059, T1218
PersistenceT1547.014 (AppDomainManager), T1053.005 (Scheduled Tasks)
Privilege EscalationT1548.002, T1078
Defense EvasionT1027, T1562
Credential AccessT1555, T1056.001
DiscoveryT1615, T1087
Lateral MovementT1210, T1021
Command & ControlT1071.001, T1567
ExfiltrationT1041, T1020

7. Indicators of Compromise (IOCs)

7.1 File System (High Confidence)

C:\Windows\Microsoft.NET\Framework\SharedReg.dll
C:\Windows\Microsoft.NET\Framework\netfxsbs9.hkf
C:\Windows\Microsoft.NET\Framework\log.cached
C:\Windows\Microsoft.NET\Framework\UevAppMonitor.exe.config

7.2 Registry

HKLM\Software\Microsoft\.NETFramework\AppDomainManager
HKLM\SYSTEM\CurrentControlSet\Services\DnsCache
HKLM\SYSTEM\CurrentControlSet\Services\NetBT

7.3 Scheduled Tasks

OneDrive Reporting Task-S-1-5-21-*

7.4 Process Execution Patterns

ParentChildMeaning
gpscript.exeHistory.iniNosyHistorian
svchost.exeUevAppMonitor.exe (non-standard path)NosyDoor
UevAppMonitor.exepowershell.exeActive backdoor

7.5 Cloud C2 Indicators

Watch for:

  • OneDrive / Google Drive access by SYSTEM or service accounts
  • API access from .NET processes
  • Repeated downloads of the same cloud file IDs
  • After-hours access patterns

7.6 Encryption and Strings

DES Key:

UevAppMo

Common strings:

SharedReg.dll
UevAppMonitor
OneDrive Reporting Task
E:\Csharp\

8. Why This Works So Well

Because nothing “looks malicious.”

  • No exploits firing
  • No strange ports
  • No suspicious binaries
  • No obvious persistence keys

Everything operates inside trusted administrative infrastructure.


9. Defensive Lessons

  1. Treat Group Policy as Tier-0 infrastructure
  2. Monitor every GPO change
  3. Audit Active Directory permissions continuously
  4. Watch system processes using cloud services
  5. Assume domain admin compromise equals full breach

Final Thought

LongNosedGoblin represents the next logical step in APT tradecraft.

Why burn zero-days when defenders already trust the tools?
Why evade security when you can operate inside it?

The organizations that will survive this kind of threat are the ones that remember a simple rule:

In cybersecurity, trust is the most dangerous vulnerability of all.


Aegiron

Backed by 11+ years in cybersecurity and incident response, we decode the latest threats shaping today’s digital battlefield. This blog cuts through the noise with clear insights on vulnerabilities, emerging exploits, and the cyber news defenders can’t afford to miss.