Check Point has rushed out emergency patches for a zero-day flaw attackers have been exploiting since late October to seize full control of its security gateways — and in some cases wipe the servers sitting behind them.
The flaw, tracked as CVE-2026-123039, is a privilege-escalation bug in the DynamicCLI interface of Check Point's Gaia operating system. It carries a CVSS severity score of 7.8. Check Point said it has seen the flaw exploited against "a couple of dozen customers," and the US Cybersecurity and Infrastructure Security Agency (CISA) added it to its Known Exploited Vulnerabilities catalog on November 13, with a remediation deadline of November 16.
A world-writable socket running as root
The bug sits in dynamic_server, the process behind Gaia's DynamicCLI daemon, which runs with root privileges. It listens on a local socket, /tmp/dynamic_socket, created with world-writable 0777 permissions — and it never checks which process is talking to it. Any user with code already running on the appliance can send it a command and have it executed as root.
On its own, the flaw is not remotely exploitable: an attacker first needs a foothold on the gateway. But Check Point says attackers got that foothold by exploiting PHP2Shell, a pair of critical flaws in the PHP engine behind its Mobile Access portal, tracked as CVE-2026-109618 and CVE-2026-109619 and rated 9.2 out of 10. Once inside the portal, they used the zero-day to climb to root.
Backdoors, then a wiper
With root access, the attackers hid web shells inside the templates the portal uses to build its login page, and installed a backdoor known as TSGA and malware called TSProxy that survives reboots, tucks itself into backups and snapshots, and hides inside the appliance's own processes, according to Check Point's research.
In the most destructive cases, the attackers pivoted from the compromised gateway to a VMware ESXi host over SSH and deployed a wiper, destroy_v6, that zeroed and deleted files and left politically themed messages in each folder, signed by a group calling itself Dream Security Ltd. Check Point said the first destructive wave began on October 30 and mostly hit Israeli organisations; a second wave from November 2 spread across several countries and showed no destruction.
Who is behind the campaign remains unclear. Check Point said the implants belong to the PhiliKit malware family, which researchers at ESET have linked to the China-nexus group UNC5221 — but the company stressed the evidence does not conclusively prove attribution.
Patch now, or stop the service
The flaw affects Check Point Firewall, Security Gateway, Security Management Server, Multi-Domain management systems and Scalable Platforms running releases R81 through R82.20. Check Point Spark firewalls and Smart-1 Cloud are not affected, and the end-of-support R81 and R81.10 branches have no fix listed.
Fixes are available in Jumbo Hotfix Accumulator updates — R82.20 Take 14, R82.10 Take 52, R82 Take 132 and R81.20 Take 174. Check Point advised teams that cannot patch immediately to stop the dynamic_server process and prevent it from restarting, and to check /var/log/dynamicCli.log for commands that arrived without the expected Clish wrapper.
The incident is the latest in a run of Check Point zero-days: in September the company warned attackers were exploiting two other flaws, CVE-2026-93616 and CVE-2026-85102, in its Security Management Server and Security Gateway. CISA has now added five Check Point flaws to its exploited-vulnerabilities catalog over the past two years, two of which have also been abused by ransomware gangs.



