Check Point has confirmed that its gateway appliances have been targeted in a series of coordinated cyberattacks leveraging a pair of critical vulnerabilities collectively referred to as PHP2Shell. These vulnerabilities, tracked as CVE-2026-109618 and CVE-2026-109619, affect all versions of PHP from 5 onwards and allow for remote code execution. While official patches for these flaws were released on October 21, 2026, evidence indicates that threat actors began exploiting them shortly before the patches were available, with mass exploitation occurring in the days that followed.
The campaign unfolded in two distinct phases. The first wave, which began on October 30, 2026, primarily targeted Israeli organizations, specifically focusing on Mobile Access gateways. In these instances, attackers inserted PHP webshells into directories used by the portal to generate login pages. This provided the intruders with a persistent foothold, allowing them to issue commands and conduct further malicious activity within the affected environments.
To escalate their privileges, the attackers utilized a separate vulnerability, CVE-2026-123039, which affects the DynamicCLI process on Check Point gateways. By exploiting this flaw, the actors were able to gain root-level access to the appliances. Once elevated, they deployed the TSGA and TSProxy malware variants. These tools are identified as components of the PhiliKit framework, which has been previously associated with the threat group known as UNC5221 during operations targeting Ivanti products.
Following the initial compromise and privilege escalation, the attackers engaged in internal asset discovery and lateral movement. In the first wave of attacks, this activity culminated in the compromise of VMware ESXi hosts, where the actors deployed and executed a wiper tool. The attackers also left behind notes referencing a cybersecurity firm, though the intent behind these references remains a subject of investigation.
A second wave of attacks commenced on November 2, 2026, expanding the scope to include organizations worldwide, including some of the same entities targeted during the initial phase. Notably, Check Point reported that this second wave did not involve the destructive wiper activity observed in the first round of attacks. The company noted that neither wave showed a clear preference for specific industry sectors, suggesting that the initial access attempts were opportunistic rather than highly targeted.
The identity of the threat actors remains unconfirmed. While the operational timing aligns with typical Chinese working hours and the malware used has historical links to China-based operations, the destructive nature of the first wave—specifically the use of a wiper—is considered inconsistent with established patterns of known Chinese state-sponsored activity.
Check Point has stated that the available evidence does not allow for a high-confidence assessment of the actor’s true identity or motivations. The company is currently evaluating several possibilities, including the potential that the operation was conducted by a single group, that the attackers acted as initial access brokers for other entities, or that the destructive elements were employed as a false-flag operation or a diversionary tactic. Investigations into the full scope and origin of the campaign are ongoing.





