Skip to main content
Meet us at Black Hat USA 2026— Las Vegas, August 1–6Book a Meeting
Mallory
🇨🇳 CN9 malware familiesExploits CVEs in the wild

Velvet Ant

Also known asVelvet Ant

Velvet Ant is a China-linked / China-nexus cyber espionage threat group tracked by Sygnia. Sygnia attributed a long-running intrusion campaign dubbed Operation Highland to Velvet Ant, reporting that the actor remained inside one organization’s network for nearly a decade, with earliest observed activity dating to 2016 or 2017. The group moved from internet-facing systems through the IT network into a segregated critical infrastructure or air-gapped segment with no direct internet connectivity. The reporting describes Velvet Ant as focused on stealthy, long-term persistence and abuse of trusted infrastructure. In Operation Highland, the actor backdoored core Linux authentication components, including PAM modules and OpenSSH binaries, and also appended attacker-controlled keys to authorized_keys files. The modified PAM components enabled authentication bypass via a hardcoded secret password and credential harvesting from legitimate logins; Sygnia identified nine distinct pam_unix.so variants. The trojanized OpenSSH components captured credentials, logged commands typed during SSH sessions, and stored encrypted data on disk, with flags to suppress logging or disguise process names. Sygnia stated that this persistence survived password resets and session termination and made remediation difficult because replacing compromised authentication components incorrectly could lock administrators out of hosts. Velvet Ant also used modified GS-Netcat reverse shells on exposed Linux servers, renaming binaries to blend in and masquerading processes as kernel threads such as [khubd] or [kauditd]. The group used systemd unit files and SysVinit scripts for persistence, a Perl SOCKS5 proxy for tunneling and lateral movement, modified Nginx configurations, and a custom binary that established SSH connections into protected networks when triggered via HTTP requests. Reporting also notes use of reverse SSH tunnels / shells as encrypted command-and-control channels. Additional activity attributed to Velvet Ant includes abuse of F5 BIG-IP appliances for persistence, including a modified /etc/rc.local file, and use of custom tools VELVETSTING and VELVETTAP on compromised F5 devices. Sygnia also linked the actor to exploitation of Cisco NX-OS CVE-2024-20399 to plant the VELVETSHELL backdoor on Cisco Nexus switches and gain arbitrary command execution after authentication. On Windows systems, the content attributes to Velvet Ant use of DLL search order hijacking, a malicious DLL named iviewers.dll masquerading as the legitimate OLE/COM Object Viewer component, PlugX deployment, process injection into multiple svchost.exe processes, WMI-based remote execution via wmiexec.py, SMB and administrative shares for tool transfer, and firewall modification with netsh.exe to open random high-numbered listener ports. The group also attempted to disable local security tools and EDR software. The content further associates Velvet Ant with masquerading techniques, including process name masquerading on Linux, and with use of encrypted communications and reverse SSH for command and control. No additional aliases or sub-groups are provided in the source content beyond Velvet Ant itself.

Share:
Are they targeting you?

Know when an actor pivots toward your sector

Mallory correlates actor tradecraft and target patterns against your stack, your sector, and your geography. See overlap before they land.

OPERATIONAL PROFILE

Targeting

Who, where, and (when attributed) which flag flies behind the operation. Pulled from open-source reporting and Mallory's analyst review.

Where they're from

Attributed origin per open-source reporting.

  • CN
MITRE ATT&CK

Tradecraft

49 distinct techniques observed across reporting, grouped by tactic. Hover any cell for the evidence excerpt; click through for MITRE's full description.

12 of 15 tactics81 techniques×N= number of intelligence reports citing this technique
MITRE ATT&CK
TA0042
Resource Development
1 technique
T1587
Develop Capabilities
T1587.003
Digital Certificates
TA0001
Initial Access
3 techniques
T1078
Valid Accounts
T1078.003×2
Local Accounts
T1133×6
External Remote Services
T1190×4
Exploit Public-Facing Application
TA0002
Execution
5 techniques
T1047
Windows Management Instrumentation
T1053
Scheduled Task/Job
T1059×2
Command and Scripting Interpreter
T1059.001
PowerShell
T1059.004
Unix Shell
T1569
System Services
T1569.002
Service Execution
T1574
Hijack Execution Flow
T1574.001
DLL
T1574.006
Dynamic Linker Hijacking
T1574.007
Path Interception by PATH Environment Variable
TA0003
Persistence
9 techniques
T1037
Boot or Logon Initialization Scripts
T1037.004
RC Scripts
T1053
Scheduled Task/Job
T1078
Valid Accounts
T1078.003×2
Local Accounts
T1098
Account Manipulation
T1098.004
SSH Authorized Keys
T1112
Modify Registry
T1133×6
External Remote Services
T1505
Server Software Component
T1543
Create or Modify System Process
T1543.002×2
Systemd Service
T1556×5
Modify Authentication Process
T1556.003×3
Pluggable Authentication Modules
TA0004
Privilege Escalation
6 techniques
T1037
Boot or Logon Initialization Scripts
T1037.004
RC Scripts
T1053
Scheduled Task/Job
T1055×4
Process Injection
T1078
Valid Accounts
T1078.003×2
Local Accounts
T1098
Account Manipulation
T1098.004
SSH Authorized Keys
T1543
Create or Modify System Process
T1543.002×2
Systemd Service
TA0005
Stealth
7 techniques
T1014
Rootkit
T1027
Obfuscated Files or Information
T1036×5
Masquerading
T1055×4
Process Injection
T1070
Indicator Removal
T1070.006
Timestomp
T1078
Valid Accounts
T1078.003×2
Local Accounts
T1574
Hijack Execution Flow
T1574.001
DLL
T1574.006
Dynamic Linker Hijacking
T1574.007
Path Interception by PATH Environment Variable
TA0112
Defense Impairment
2 techniques
T1112
Modify Registry
T1556×5
Modify Authentication Process
T1556.003×3
Pluggable Authentication Modules
TA0006
Credential Access
7 techniques
T1003
OS Credential Dumping
T1056
Input Capture
T1056.001
Keylogging
T1187
Forced Authentication
T1555
Credentials from Password Stores
T1556×5
Modify Authentication Process
T1556.003×3
Pluggable Authentication Modules
T1557
Adversary-in-the-Middle
T1557.001
Name Resolution Poisoning and SMB Relay
T1649×2
Steal or Forge Authentication Certificates
TA0007
Discovery
2 techniques
T1049
System Network Connections Discovery
T1083
File and Directory Discovery
TA0008
Lateral Movement
2 techniques
T1021×2
Remote Services
T1021.002×2
SMB/Windows Admin Shares
T1021.004×3
SSH
T1210
Exploitation of Remote Services
TA0009
Collection
2 techniques
T1056
Input Capture
T1056.001
Keylogging
T1557
Adversary-in-the-Middle
T1557.001
Name Resolution Poisoning and SMB Relay
TA0011
Command and Control
4 techniques
T1090×4
Proxy
T1105
Ingress Tool Transfer
T1132
Data Encoding
T1573
Encrypted Channel
T1573.002
Asymmetric Cryptography
WEAPONIZED

Associated vulnerabilities

8 CVEs this actor has used in observed campaigns. 8 of them exploited in the wild.

CVE-2024-20399Command Injection in Cisco NX-OS CLIIn the wildEvidence3

The same actor had earlier abused F5 BIG-IP appliances. It also exploited CVE-2024-20399, a Cisco NX-OS zero-day, to plant the VELVETSHELL backdoor on Nexus switches.

CVE-2021-31207Post-auth Arbitrary File Write in Microsoft Exchange Server (ProxyShell)In the wildEvidence1

This analytic identifies potential exploitation attempts of ProxyShell (CVE-2021-34473, CVE-2021-34523, CVE-2021-31207) and ProxyNotShell (CVE-2022-41040, CVE-2022-41082) vulnerabilities in Microsoft Exchange Server.

CVE-2021-34473ProxyShell pre-auth SSRF in Microsoft Exchange AutodiscoverIn the wildEvidence1

This analytic identifies potential exploitation attempts of ProxyShell (CVE-2021-34473, CVE-2021-34523, CVE-2021-31207) and ProxyNotShell (CVE-2022-41040, CVE-2022-41082) vulnerabilities in Microsoft Exchange Server.

CVE-2021-34523Microsoft Exchange PowerShell Backend Elevation of Privilege (ProxyShell)In the wildEvidence1

This analytic identifies potential exploitation attempts of ProxyShell (CVE-2021-34473, CVE-2021-34523, CVE-2021-31207) and ProxyNotShell (CVE-2022-41040, CVE-2022-41082) vulnerabilities in Microsoft Exchange Server.

CVE-2022-26134Atlassian Confluence Server and Data Center OGNL Injection RCEIn the wildEvidence1

The following analytic detects attempts to exploit CVE-2022-26134, an unauthenticated remote code execution vulnerability in Confluence... This activity is significant as it allows attackers to execute arbitrary code on the Confluence server without authentication, potentially leading to full system compromise.

3 more CVEs tied to this actor tracked in Mallory.

ACTIVITY FEED

Recent activity

20 sources tracked across advisories, community write-ups, and news. New activity surfaces here as Mallory finds it.

What this page doesn’t show

The version that knows your environment.

This page is what’s public. Mallory adds the parts that aren’t: sector and geo overlap with your footprint, the IOCs they’re burning right now, detection coverage, and what to do next.
Target overlap

Match sector + geo + tech-stack targeting against your real footprint.

Tradecraft mapping49

Every observed MITRE ATT&CK technique, grouped by tactic.

Malware arsenal9

Families this actor is known to deploy, with IOCs and behavior.

Exploited CVEs8

CVEs this actor has used in known campaigns.

Detection signatures

YARA, Sigma, Snort, and vendor rules, auto-deployed to your SIEM.

Observables

Domains, IPs, and hashes tied to this actor, refreshed continuously.