A sophisticated and persistent wave of cyberattacks has been hammering servers worldwide for over a week, as threat actors aggressively exploit critical vulnerabilities collectively known as "PHP2Shell." While the volume of exploitation attempts is massive, security researchers have confirmed that the number of successful, full-scale compromises remains relatively low. However, the nature of the attacks—which include the sophisticated chaining of multiple vulnerabilities—has put security teams and organizations on high alert.
Among the most significant developments in this campaign is the targeted exploitation of Palo Alto Networks firewalls. A China-linked hacking group has successfully leveraged PHP2Shell flaws in tandem with the "Fragnesia" kernel privilege escalation bug to gain root access to these devices. This level of sophistication, which includes the use of artificial intelligence to assist in the chaining of distinct exploits, marks a concerning evolution in how threat actors are approaching perimeter defense infrastructure.
The Anatomy of the Palo Alto Networks Attacks
The attacks against Palo Alto Networks equipment are particularly notable due to the precision required to execute them. The hackers utilized two distinct exploit chains: the first targeted the PHP2Shell vulnerabilities identified as CVE-2026-109618 and CVE-2026-109619, which reside within the PAN-OS software used on Palo Alto Networks devices. By chaining these with the Fragnesia kernel exploit, the attackers were able to bypass standard security boundaries, effectively deploying webshells with root-level privileges on the target firewalls.
Cybersecurity firm Dream first identified the campaign earlier this week, providing evidence that at least 25 firewalls—out of a pool of hundreds analyzed—had been compromised. Palo Alto Networks has since conducted an independent investigation and confirmed the findings. In an official statement, the company noted that its Unit 42 research division verified the attacker’s tooling and the reported outcomes.
In response to the threat, Palo Alto Networks has issued patches for its PAN-OS 12.2, 12.1, 11.2, 11.1, and 10.2 software versions, covering both PA-Series and VM-Series firewalls, as well as the Panorama management platform. The company emphasized that no special configuration is required for a system to be vulnerable, noting that active exploitation has been observed specifically targeting management interfaces that are exposed to the public internet.
Palo Alto Networks has been proactive in its remediation efforts, clarifying that by default, the management interface for its devices is not exposed to the internet. The company urged customers to immediately restrict access to these interfaces and apply the available patches. According to the manufacturer, only a tiny fraction of its global installed base—approximately 0.02%—had internet-facing management interfaces configured and was subsequently impacted. The company has taken the step of contacting each of these affected customers individually and continues to monitor the situation closely to provide necessary updates.
A Massive Landscape of Exposed PHP Instances
The threat posed by PHP2Shell is not limited to specialized networking hardware. Because PHP is the backbone of a significant portion of the modern web, the potential attack surface is enormous. As of October 29, telemetry data identified over 22.6 million exposed PHP instances that are potentially vulnerable to the PHP2Shell bugs. This is a staggering figure, particularly when considering how deeply integrated PHP is into the infrastructure of major content management systems (CMS) such as WordPress, Drupal, and Joomla.
While the mass exploitation attempts have targeted these common CMS platforms, they have remained largely unsuccessful in achieving widespread compromise, according to insights from the web security firm Defiant. As of October 28, Defiant reported that there were no confirmed compromises of WordPress sites linked to the PHP2Shell campaign, despite the firm recording more than 77,000 related requests originating from over 2,000 distinct IP addresses.
The failure of these automated, mass-scale attacks to achieve broad success is attributed to the complexity of the environments being targeted. Defiant noted that their testing revealed that the exploits often require significant adjustments to function correctly on a given target. Factors such as Address Space Layout Randomization (ASLR), specific PHP build versions, the presence of various running extensions, memory limits, and the individual server’s method of request handling all play a role in whether an exploit succeeds. The attackers are continuously modifying their toolkits to overcome these hurdles, and the consensus among security researchers is that updating PHP environments remains an essential defense measure, as a successful breach could grant an attacker full control over site files, databases, and sensitive credentials.
Assessing the Threat Landscape and Attacker Behavior
Despite the high volume of traffic associated with the PHP2Shell campaign, success remains confined to a very small number of threat actors. Cato Networks, which has been tracking the campaign, noted that roughly 96% of the attacks it blocked targeted CVE-2026-109618. Of those attempts, approximately 41% were specifically designed to facilitate remote code execution (RCE), while the remainder consisted of network scanning or invalid proof-of-concept (PoC) requests.
The geographic distribution of these attacks is widespread, with attempts originating from a handful of IP addresses located in the United States, the Netherlands, and Germany. Most of these attempts appear to be repetitive, PoC-style requests focused on command execution or deserialization. Regarding the targets, Cato Networks found that healthcare organizations were hit the hardest, followed by entities in the technology, security, professional services, automotive, education, and manufacturing sectors. Interestingly, the attacks do not appear to be highly targeted in a strategic sense; rather, they seem opportunistic. The same source IP addresses were observed sending requests to a diverse range of organizations across different sectors within the same day, suggesting an automated "spray and pray" approach.
WatchTowr, another security firm monitoring the situation, reported that it has observed successful PHP2Shell exploitation from a single IP address located in Germany. Despite the large volume of traffic and the variety of payloads being tested, the post-exploitation activity from this actor has been remarkably minimal. In each case identified by WatchTowr, the attacker executed a simple "curl" command—a standard utility for transferring data—to confirm that the exploitation was successful, after which the activity ceased without further post-exploitation maneuvers.
Industry analysts suggest that this behavior indicates that threat actors are still in the testing phase, refining their techniques and developing more reliable payloads. Jake Knott, head of threat intelligence at WatchTowr, noted that as attackers develop more appliance-specific exploits and improve their methods, the industry is likely to see an increase in public reporting regarding highly targeted, successful compromises of edge appliances. The current situation serves as a stark reminder of the risks associated with exposed management interfaces and the critical importance of timely patching. As the security community continues to analyze the evolving PHP2Shell threat, the primary recommendation for administrators remains the same: ensure all internet-facing systems are fully patched, limit management access to authorized networks only, and maintain a vigilant watch for anomalous traffic patterns.






