Skip to content

Image: research.checkpoint.com · rights & removal

Executive Summary

Activity was observed targeting Israeli organizations through Mobile Access portals on Check Point Security Gateways around late October 2026, approximately one week after a security fix release. Exploitation activity was associated with PHP2Shell, linked to CVE-2026-109618 and CVE-2026-109619. Initial access led to privilege escalation from a portal service account to root by exploiting CVE-2026-123039 in the DynamicCLI IPC interface. Attackers then deployed PHP webshells for command execution and persistence. Subsequent scanning on November 2nd targeted previously affected organizations and global entities without apparent sector focus. Exploitation signs included session artifacts under /opt/CPcvpn-R81.20/tmp/ and segmentation faults in Apache/PHP workers on October 30, 2026. Attackers established a passive backdoor by launching TSGA malware (/tmp/CPconfigdaemon) and redirecting output to /tmp/.bd.log. This backdoor employed an implant named slotCPconfigdaemon which deployed the TSProxy shared library (/lib/libcpfwcommon.so) for intercepting network functions. Lateral movement occurred to a VMware ESXi host via SSH, authenticating with the vpxuser service account, followed by the execution of the destroyv6 wiper on the host. Tools involved include PHP webshells, a Linux ELF wiper (destroy\v6), and the TSGA/TSProxy malware family associated with PhiliKit.

Facts Only

* Activity targeted Israeli organizations via vulnerable Mobile Access portals on Check Point Security Gateways in late October 2026.
* Exploitation activity was consistent with PHP2Shell, linked to CVE-2026-109618 and CVE-2026-109619.
* Privilege escalation occurred from a portal service account to root by exploiting CVE-2026-123039 in the DynamicCLI IPC interface.
* PHP webshells were deployed using privileged command execution to establish persistence.
* Session artifacts containing uploadprogresspwn markers were created under /opt/CPcvpn-R81.20/tmp/.
* Apache/PHP workers experienced segmentation faults during exploitation attempts on October 30, 2026.
* A passive backdoor, TSGA malware (/tmp/CPconfigdaemon), was deployed to redirect output to /tmp/.bd.log.
* The TSGA implant relaunched as slot\CPconfigdaemon in the boot persistence directory and injected TSProxy (/lib/libcpfwcommon.so) into appliance processes to intercept networking functions.
* The attackers scanned for internal assets using gogo and moved laterally to a VMware ESXi host over SSH with vpxuser credentials.
* A destroy\v6 wiper was executed on the ESXi host, targeting /tmp/des wiper, accompanied by the file DREAM\ANNEXATION.txt.
* The PHP webshell required a hardcoded key for command execution and supported commands in plain-text, Base64, and XOR-obfuscated formats.
* The ELF wiper destroy\v6 accepts target directories and can exclude files; it also writes DREAM\ANNEXATION.txt upon processing.
* Tools identified include tpl webshell, slot\CPconfigdaemon, CPconfigdaemon, gogo, destroy\v6, DREAM\ANNEXATION.txt, philistine, libdssafe.so.
* Infrastructure IP addresses used included 130.94.19[.]204 and 130.94.19[.]168.

Full Take

The narrative describes a multi-stage intrusion leveraging known vulnerabilities in Check Point Mobile Access portals to achieve deep system compromise. The initial access leverages specific CVEs and privilege escalation techniques to gain root access, demonstrating an understanding of internal service interfaces (DynamicCLI IPC). This immediately shifts the focus from simple exploitation to establishing persistent, stealthy control.
The transition to deploying PHP webshells demonstrates a clear methodology for maintaining web-based command and control, which is then evolved into a more sophisticated, low-level backdoor mechanism. The deployment of TSGA and TSProxy reveals an advanced technique: embedding functionality directly into system components (TSGA) while using shared libraries (TSProxy) to achieve stealthy network interception. This linkage to the PhiliKit family suggests the use of pre-existing, related toolsets, indicating a coordinated, established operational playbook rather than opportunistic attacks.
The final stage—lateral movement to an ESXi host and deployment of the destroy\v6 wiper—indicates an objective focused on destructive impact post-compromise, moving beyond simple data exfiltration or persistence toward disruption. The association with infrastructure previously linked to Chinese-nexus threat actors introduces a layer of complexity, suggesting potential false-flag operations or involvement by actors leveraging shared tooling lineage. This forces an inquiry into whether the observed activity reflects actor motivation (e.g., sabotage, espionage) or if it serves as a mechanism for operational testing utilizing known malware families and toolsets. What evidence exists to confirm the destructive action was the primary goal versus a smokescreen?

From the original · Check Point Research

SUBSCRIBE TO CYBER INTELLIGENCE REPORTS [contact-form-7 id=”26727″ title=”SUBSCRIBE TO CYBER INTELLIGENCE REPORTS”] We have now confirmed that, in late October 2026, approximately one week after a security fix became available, activity originating from 130.94.19[.]204 targeted Israeli organizations through vulnerable Mobile Access portals on Check Point Security Gateways.
Read the full story at research.checkpoint.com

Sentinel unavailable

The automated check of the source article's wording did not complete, so no result is shown.

This looks only at the wording of the original source article, not at this page's AI-written sections. A small local AI model made this estimate. It has not been checked against known human and machine texts, so treat it as provisional. It cannot show who wrote an article.

PHP2Shell Exploitation on Check Point Gateway | Huntaegis