Security researchers at Unit 42 have uncovered a prolonged cyber-espionage campaign targeting government and critical infrastructure sectors in Southeast Asia. The campaign is attributed to a Chinese-speaking threat actor tracked as CL-STA-1062, which is also believed to be the same cluster as UAT-7237. Active since at least March 2022, the group has demonstrated a strategic focus on the Asia-Pacific region, with recent activities concentrating on state-owned enterprises in the energy and government sectors.
The attackers employ a hybrid toolkit, combining publicly available hacking tools like Mimikatz, SoftEther VPN, and JuicyPotato with a custom, previously undocumented backdoor named TinyRCT. This new malware is a lightweight .NET-based Remote Access Trojan (RAT) designed for stealth and persistence, enabling command execution, file exfiltration, and reconnaissance. The primary attack vector involves the exploitation of public-facing web applications to deploy web shells, which serve as a foothold for further malicious activities. This campaign highlights a persistent and evolving threat focused on long-term intelligence gathering against strategic targets.
The threat actor CL-STA-1062 has been conducting a series of targeted attacks against government entities and critical energy infrastructure (CEI) in Southeast Asia, with activity escalating throughout 2025. The group's operations show a clear intent of espionage, focusing on exfiltrating sensitive information, including database contents and web server source code. The campaign's scope extends beyond Southeast Asia, with earlier operations targeting strategic sectors in East Asia and web hosting infrastructure in Taiwan, indicating a broad and sustained regional focus.
Initial access is consistently achieved by exploiting vulnerabilities in public-facing web applications, leading to the deployment of ASPX web shells. These web shells act as the primary mechanism for the attackers to execute commands, perform reconnaissance, and deploy additional tools. The actor has been observed compromising multiple organizations, including at least ten entities between October and December 2025, and two state-owned critical energy providers. The attackers demonstrate operational sophistication by adapting their techniques to the target environment and using various tunneling tools for command and control (C2) and data exfiltration, often disguising them as legitimate system files to evade detection.
The attack lifecycle of CL-STA-1062 demonstrates a methodical approach, leveraging both custom and off-the-shelf tools. Their Tactics, Techniques, and Procedures (TTPs) map to several MITRE ATT&CK techniques.
T1190 - Exploit Public-Facing Application: The primary initial access vector is the exploitation of vulnerabilities in web applications to gain a foothold.T1505.003 - Server Software Component: Web Shell: Upon successful exploitation, the attackers deploy ASPX web shells. These shells are used to execute arbitrary commands via cmd.exe, as noted in observed command-line activity.T1059.003 - Command and Scripting Interpreter: Windows Command Shell: Web shells are used to interact with the underlying operating system and execute further commands for reconnaissance and tool deployment.T1068 - Exploitation for Privilege Escalation: The group uses known tools like JuicyPotato to escalate privileges on compromised Windows systems.T1003.001 - OS Credential Dumping: LSASS Memory: The use of Mimikatz indicates attempts to dump credentials from memory, particularly from the LSASS process, to facilitate lateral movement.T1090 - Proxy: The attackers heavily rely on tunneling tools to maintain C2 and exfiltrate data. Tools observed include SoftEther VPN, VNT, and yuze. These tools are often disguised with legitimate-sounding names like vmtools.exe to avoid suspicion.T1074.001 - Data Staged: Local Data Staging: Before exfiltration, stolen data is compressed into password-protected RAR archives.T1041 - Exfiltration Over C2 Channel: Data is exfiltrated through the established C2 tunnels. In some cases, simple tools like curl were used to send enumeration results directly to an actor-controlled IP address.TinyRCT is a new, lightweight backdoor written in C# for Windows systems. It was discovered on a server at 139.180.134.221 named PerfWatson2.exe. The backdoor provides core RAT functionalities, including:
The use of a custom backdoor suggests the threat actor has dedicated development resources and aims for long-term, stealthy operations that are less likely to be detected by signature-based security products.
The activities of CL-STA-1062 pose a significant espionage threat to governments and critical infrastructure operators in the Asia-Pacific region. The primary impact is the loss of confidentiality, as the attackers have successfully exfiltrated sensitive data, including database information and proprietary web server source code. This stolen information could be used for further intelligence gathering, to plan future attacks, or to gain a strategic advantage.
For critical energy infrastructure entities, a compromise could lead to severe consequences beyond data theft. While the current campaign appears focused on espionage, the access and persistence gained by the attackers could potentially be leveraged for disruptive or destructive purposes in the future. The compromise of state-owned enterprises could have economic and national security implications for the affected countries. The group's ability to remain active for several years and continuously evolve its toolkit indicates a persistent, well-resourced adversary capable of inflicting significant damage.
The following Indicators of Compromise were explicitly mentioned in the source report.
139.180.134.221PerfWatson2.exe (TinyRCT).PerfWatson2.exeSecurity teams may want to hunt for the following patterns, which could indicate related activity:
*.aspxcurl.exerar.exe-p) and archiving.vmtools.exeyuze.exe, vnt.exe*unrar* e -p* *Defenders should focus on a multi-layered detection strategy to identify CL-STA-1062 activity.
Web Server Monitoring: Implement File Integrity Monitoring (FIM) on web servers to detect the creation of unauthorized files, especially .aspx web shells. Analyze web server logs for suspicious requests that may indicate exploitation or web shell interaction. This aligns with D3FEND technique D3-FA: File Analysis.
Endpoint Detection and Response (EDR): Deploy EDR solutions to monitor for suspicious process execution. Create detection rules for:
Mimikatz, JuicyPotato, and tunneling tools (SoftEther, VNT, yuze).w3wp.exe) spawning command shells (cmd.exe) or curl.exe.vmtools.exe that is not signed by VMware). This leverages D3FEND's D3-PA: Process Analysis.Network Security Monitoring: Utilize D3FEND's D3-NTA: Network Traffic Analysis to detect C2 communications. Monitor for outbound connections to suspicious IPs like 139.180.134.221. Baseline normal traffic from critical servers and alert on anomalies, such as unexpected VPN tunnels or large data transfers.
Credential Theft Detection: Monitor for signs of credential dumping, such as direct access to the LSASS process by non-system processes. Enable and monitor Windows Event Logs related to credential access.
Upon detection, incident response teams should immediately isolate compromised hosts, revoke compromised credentials, and conduct a thorough investigation to determine the full scope of the breach before eradication.
Organizations should implement the following tactical and strategic controls to defend against this threat actor.
Secure Web Applications: The primary defense is to prevent initial access. Regularly scan and patch public-facing applications (M1051 - Update Software). Implement a Web Application Firewall (WAF) to block common exploitation techniques. This corresponds to D3FEND's D3-AH: Application Hardening.
Execution Prevention: Use application allowlisting (M1038 - Execution Prevention) to prevent the execution of unauthorized tools like Mimikatz, tunneling software, and custom malware. This can be implemented via D3FEND's D3-EAL: Executable Allowlisting.
Network Segmentation and Filtering: Restrict outbound traffic from web servers to only what is explicitly required (M1037 - Filter Network Traffic). Implement network segmentation to prevent easy lateral movement from a compromised web server into the internal network.
Privilege and Credential Management: Enforce the principle of least privilege. Harden systems to prevent known privilege escalation techniques. Implement M1032 - Multi-factor Authentication on all critical systems and accounts to mitigate the impact of stolen credentials. Utilize M1026 - Privileged Account Management solutions to secure and monitor administrative accounts.
Regularly patch public-facing web applications and underlying servers to eliminate vulnerabilities exploited for initial access.
Mapped D3FEND Techniques:
Implement egress filtering rules to block outbound connections from web servers to untrusted IP addresses and ports, disrupting C2 and exfiltration channels.
Mapped D3FEND Techniques:
Use application allowlisting to prevent the execution of unauthorized tools like Mimikatz, JuicyPotato, and tunneling software on critical servers.
Mapped D3FEND Techniques:
Enforce the principle of least privilege and closely monitor the use of administrative accounts to limit the impact of credential theft.
Require MFA for all remote access and for access to sensitive systems and data to prevent unauthorized access even if credentials are compromised.
Mapped D3FEND Techniques:
Deploy and maintain endpoint protection solutions to detect and block known malicious tools and the custom TinyRCT backdoor.
Mapped D3FEND Techniques:
Use a Web Application Firewall (WAF) to inspect and filter inbound web traffic, blocking malicious requests that attempt to exploit vulnerabilities or upload web shells.
Mapped D3FEND Techniques:
Given that CL-STA-1062's primary entry point is exploiting public-facing web applications, maintaining a rigorous patch management program is the most critical defense. Organizations must establish an automated process to scan for, identify, and remediate vulnerabilities in all web-facing software, including web servers, content management systems, and custom applications. Prioritize patches for vulnerabilities known to be exploited in the wild (KEVs). This process should not be limited to the application layer; the underlying operating systems and third-party components must also be kept up-to-date. A risk-based approach should be adopted, applying critical patches to internet-facing systems within 24-48 hours of release. This directly disrupts the attacker's ability to gain initial access, forcing them to expend more resources on developing zero-day exploits, which are significantly rarer and more valuable.
To counter CL-STA-1062's use of a diverse toolkit including Mimikatz, JuicyPotato, and various tunneling tools, organizations should implement application allowlisting on critical servers, especially web servers. By defining a strict policy that only permits known, legitimate executables to run, defenders can block the execution of the attacker's entire post-exploitation arsenal. This should be deployed in 'enforce' mode on production servers. The policy should be granular, preventing web server user accounts (e.g., IUSR, Network Service) from executing anything beyond their required binaries. This strategy is highly effective against both the known open-source tools and the custom TinyRCT backdoor, as the hash or signature of these unauthorized files will not be on the allowlist. This forces the attacker into more difficult fileless or living-off-the-land techniques, which are easier to detect through behavioral analysis.
CL-STA-1062 relies on tunneling tools like SoftEther VPN and VNT for command and control and data exfiltration. To defeat this, organizations must implement strict egress filtering on perimeter firewalls and network security groups. Web servers should be placed in a segmented DMZ and denied all outbound internet access by default. Explicit firewall rules should be created to allow only necessary outbound traffic to specific IP addresses and ports (e.g., for software updates or API calls). This 'default-deny' posture breaks the attacker's C2 channel, preventing them from receiving commands or exfiltrating data, even if they successfully compromise a server. Monitoring for and alerting on any denied outbound connection attempts from the DMZ can serve as a high-fidelity indicator of a successful breach, enabling rapid incident response.
CL-STA-1062 activity is assessed to have begun, with initial operations observed across East Asia.
Throughout 2025, Unit 42 observed a cluster of activity from CL-STA-1062 targeting government and critical infrastructure in Southeast Asia.
Attackers compromise a Southeast Asian government entity, deploying web shells and exfiltrating MSSQL database information.
Between October and December 2025, at least ten different organizations in Southeast Asia were likely compromised by the threat actor.
The threat actor compromises two state-owned critical energy infrastructure (CEI) entities in a Southeast Asian country.

Cybersecurity professional with over 10 years of specialized experience in security operations, threat intelligence, incident response, and security automation. Expertise spans SOAR/XSOAR orchestration, threat intelligence platforms, SIEM/UEBA analytics, and building cyber fusion centers. Background includes technical enablement, solution architecture for enterprise and government clients, and implementing security automation workflows across IR, TIP, and SOC use cases.
Help others stay informed about cybersecurity threats
Every tactic, technique, and sub-technique used in this threat has been identified and mapped to the MITRE ATT&CK framework for consistent, actionable threat language.
Observables and indicators of compromise (IOCs) have been extracted and cataloged. Risk has been assessed and correlated with known threat actors and historical campaigns.
Detection rules, incident response steps, and D3FEND-aligned mitigation strategies are included so your team can act on this intelligence immediately.
Structured threat data is packaged as a STIX 2.1 bundle and can be visualized as an interactive graph — relationships between actors, malware, techniques, and indicators.
Sigma detection rules are derived from the threat techniques in this article and can be converted for deployment across any major SIEM or EDR platform.