Threat actors have once again set their sights on Qlik Sense enterprise servers, launching a coordinated campaign in October 2026 to deliver unauthorized PowerShell reverse shell commands across multiple environments. Security researchers have noted striking similarities between this latest activity and historical campaigns previously linked to sophisticated ransomware operations, raising fresh alarms for organizations utilizing these business intelligence platforms.
According to a detailed advisory released by Cynet’s CyOps and Research teams, the suspicious activity was identified last month targeting several customers located within the European Union. The attacks were executed within an unusually narrow window of just over three minutes, highlighting the automated, high-speed nature of the threat actor’s initial access and execution strategy. While the specific vulnerability leveraged to breach the Qlik Sense servers remains unidentified in this latest wave, the operational patterns closely mirror intrusions previously documented by Arctic Wolf in November 2023, which involved threat actors associated with the notorious CACTUS ransomware operation exploiting newly disclosed flaws to gain a foothold in target environments.

The historical vulnerabilities at the center of these past intrusions—patched by Qlik Sense in August and September 2023—impacted Qlik Sense Enterprise for Windows. Among them, CVE-2023-48365, widely referred to as DoubleQlik, served as an incomplete fix for a prior flaw tracked as CVE-2023-41265. During the November 2023 CACTUS ransomware campaign, malicious actors systematically exploited these security gaps to establish initial footholds, subsequently downloading secondary payloads to maintain persistence and establish remote control. Those historical attacks frequently involved the deployment of third-party administrative tools and remote access utilities, including ManageEngine Unified Endpoint Management and Security, AnyDesk, and Plink.
In the October 2026 incidents analyzed by Cynet, the threat actors employed a PowerShell command featuring a one-line reverse shell designed to communicate with an external IP address, specifically 45.115.26[.]23, over port 4455. Investigators observed that the malicious command was executed under the legitimate process name scheduler.exe, a behavioral fingerprint that closely resembles the 2023 CACTUS intrusions targeting Qlik environments. Although the automated attempts to spawn the reverse shell were successfully blocked during the initial phases of the attacks, subsequent analysis and controlled testing conducted by Cynet revealed the true intended mechanics of the interactive communication channel.

When simulated in a secure laboratory environment, the reverse shell permitted threat actors to conduct internal network reconnaissance and deliver advanced PowerShell scripts engineered to establish persistent access. This persistence mechanism relies heavily on a Windows Go-based reverse shell connecting to the same remote infrastructure on port
- Once active, the resulting backdoor actively scans the compromised system for running processes associated with Apache ActiveMQ. Furthermore, if the intrusion successfully breaches a Linux environment, the malware attempts to harvest sensitive system information and read critical authentication files, specifically
/etc/shadowand/etc/passwd.
Further compounding the threat, researchers discovered a secondary variant of the PowerShell script designed to automate the installation of OpenSSH on targeted systems. This script creates a scheduled task explicitly named "WinUpdCheck," which is configured to launch a PowerShell reverse shell written to C:ProgramDatasshwinc.ps1 at strict five-minute intervals. If the OpenSSH server daemon (sshd) is unavailable on the compromised machine, the script automatically downloads a legitimate OpenSSH archive, OpenSSH-Win64.zip, from the attacker’s staging server, extracts its contents, and executes the installation script. The utility is then configured for automatic startup, the service is initiated, and an inbound firewall rule is dynamically added to permit TCP traffic on port 22.

An extensive examination of the underlying threat actor infrastructure uncovered an exposed directory hosted at 45.115.26[.]23:8012. This open repository contained a diverse arsenal of malicious tools and utility scripts, including a custom Go backdoor tailored for Windows systems, advanced SSH persistence scripts, a proof-of-concept exploit targeting the Apache ActiveMQ vulnerability tracked as CVE-2023-46604, and a modified public Veeam toolchain engineered to exploit CVE-2024-40711. The presence of these multi-platform exploits and tools strongly indicates that the threat actor is actively laying the technical groundwork for broader, multi-vector campaigns against enterprise networks.
