🔍 Read the full analysis: A PHP2Shell Chain Took Over 25 Palo Alto Servers—Here’s The Timeline on ThorstenMeyerAI.com
Prime made for students and young adults
- Fast, free delivery for dorm and study essentials
- Prime Video and Amazon Music included
- Member-only deals
TL;DR
Dream reported that an attacker used PHP2Shell and a previously disclosed Linux kernel flaw to gain root access on 24 Palo Alto Networks firewalls, then compromise an internal firewall-management server at a government-affiliated organization. Palo Alto Networks says Unit 42 independently verified the attacker’s tools and results. Dream found AI tools in the attacker’s environment but says it has no direct evidence they wrote the exploits or scripts.
An attacker compromised 25 servers, including 24 Palo Alto Networks PAN-OS firewalls, in a campaign using the newly fixed PHP2Shell flaws and a previously disclosed Linux kernel bug, according to Israeli security company Dream. Palo Alto Networks says its Unit 42 researchers independently verified the attacker’s tools and results. The activity matters because the attacker reached a government-affiliated network and targeted its Active Directory, which controls access to organizational accounts and systems.
Dream’s account says the attacker used PHP2Shell vulnerabilities CVE-2026-109618 and CVE-2026-109619 to run code on exposed firewall management pages. The attacker then chained them with Fragnesia, CVE-2026-46300, a Linux kernel privilege-escalation flaw disclosed in May. Dream says the combined path led to root access, the system’s highest level of privilege. It reported that 22 of the 24 rooted firewalls were left with a hidden web shell, while two belonged to the government-affiliated organization later breached internally.
The reported sequence began on October 23, two days after Palo Alto published an advisory about PHP2Shell. Dream says an automated run targeted 240 firewalls, selected from a larger set of internet-visible management pages, and gained root access on 24. On October 24, the attacker compromised a second firewall at the government-affiliated organization, established a connection back to the attacker and persistence that survived reboots, then reached the organization’s internal firewall-management server at 11:35 UTC. Dream identifies that machine as the 25th compromised server and says the attacker then targeted Active Directory.
Dream says the campaign’s targets were in the United States, Japan, Taiwan and South Korea. It also reported that the attacker’s setup included Claude Code and an offensive-security platform called aiscan configured with the Chinese models DeepSeek and Kimi. Their presence is evidence that AI tools were available in the environment, but not proof they generated the exploit chain. Dream said it had no direct evidence that AI wrote the exploit or scripts and that the operation did not appear fully autonomous.
Infographic 01 · PHP2Shell on Palo Alto firewalls
Two days from patch to root.
PHP fixed the bug on 21 October. Fixed firewall software for the attacked branches arrived almost a week later. The attack ran in between.
- 13–23 MayFragnesia goes public, kernel fix followsLinux privilege-escalation bug with a public exploit; CVE-2026-46300.
- 8 OctAttacker sets up an AI platformaiscan configured with DeepSeek and Kimi, per Dream’s timeline.
- 21 OctPHP fixes PHP2ShellPHP 8.5.12, 8.4.27, 8.3.36, 8.2.35. Palo Alto advisory the same day; PAN-OS fixes “upcoming”.
- 23 Oct, evening UTCBatch run: 240 targets, 24 rootedFound via device search engine, filtered to US, Japan, Taiwan, South Korea.
- 24 OctInto a government-affiliated networkTunnel, persistence, internal management server rooted (no. 25), Active Directory attacked.
- 26 OctDream publishes · CISA lists PHP2ShellFederal deadline 29 Oct.
- 27–29 OctFixed PAN-OS builds for 12.2 and 12.1CISA lists Fragnesia on 28 Oct; Unit 42 confirms Dream’s results on 29 Oct.
The fix for PHP existed. The fix for the firewall did not, yet. The open management page was the only door that mattered.Thorsten Meyer · my reading
Sources: Dream Research, 26 Oct 2026 (updated 29 Oct); php.net, 21 Oct 2026; Palo Alto Networks advisory (updated 29 Oct 2026); CISA KEV catalog, version 29 Oct 2026; oss-security, 13 May 2026; Unit 42, 29 Oct 2026. Day count: Thorsten Meyer.
Powered by Thorsten Meyer AI
Exposed Management Pages Enabled Intrusion
The reported breach shows how a flaw in a device’s management software can become a route into a wider organization when that management interface is reachable from the public internet. In this case, the attacker reportedly combined remote code execution with a kernel privilege-escalation flaw, then used compromised firewalls to reach an internal server responsible for managing other firewalls. The reported move toward Active Directory raises the potential stakes because that directory governs access across many organizational systems.
The timing also matters for defenders. Palo Alto had published an advisory on October 21, but Dream says fixed builds for the affected 12.1 and 12.2 branches were released between October 27 and 29. That left a period when organizations needed to rely on mitigations, including keeping management interfaces off the internet, while awaiting applicable updates. The report does not establish the full impact on the government-affiliated organization or whether the attacker accessed sensitive information.
The AI element should be treated carefully. Finding AI products or model configurations on a suspected attacker’s servers does not establish that AI created the exploit, selected targets or carried out the intrusion autonomously. Dream’s own account distinguishes those observations from proof of AI use in the attack itself. The known chain also relied on publicly disclosed vulnerabilities and exposed administration services, making patching and access restrictions directly relevant regardless of AI involvement.
Products Worth Considering
LG 34GX90SA-W 34-inch Ultragear WQHD OLED Curved Gaming Monitor
Samsung 27 Odyssey OLED G5 (G50SF) QHD QD-OLED Gaming Monitor
Apple AirPods Pro 3 Wireless Earbuds with Active Noise Cancellation
As an affiliate, we earn on qualifying purchases.
From May Kernel Flaw to October Patch
Fragnesia was disclosed in May 2026, according to Dream’s timeline, with a public exploit and a kernel fix. The PHP project released updates on October 21, including PHP 8.5.12, 8.4.27, 8.3.36 and 8.2.35, and urged users to update. Palo Alto published its PHP2Shell advisory that day and described fixed PAN-OS builds as forthcoming. Dream says the firewalls use a bundled copy of PHP, meaning updates to PHP itself did not automatically update every affected firewall.
Dream says its researchers found roughly 19,000 PAN-OS firewalls with internet-facing management pages before the attacks. A custom scan returned version information for about 5,000, after which filtering by country and software version reduced the potential targets to 240. These figures describe Dream’s reported discovery and filtering process; they are not a count of confirmed vulnerable devices worldwide.
According to Dream, the attacker configured aiscan with DeepSeek and Kimi on October 8. Researchers found the attack artifacts on October 25, and Dream published its account on October 26, updating it on October 29. The source account also says the U.S. Cybersecurity and Infrastructure Security Agency added both PHP2Shell CVEs to its Known Exploited Vulnerabilities catalog on October 26 and set a three-day deadline for federal agencies. Palo Alto’s fixed builds for the affected branches followed from October 27 through October 29.
““No direct evidence” showed that AI tools wrote the exploit or scripts.”
— Dream
Impact and AI Role Remain Unclear
The available account does not establish whether the attacker successfully obtained sensitive data, changed access permissions, or maintained access beyond the persistence Dream describes. It also does not identify the government-affiliated organization or specify how far the Active Directory activity progressed. The full scope of the intrusion and any resulting damage remain unconfirmed in the supplied reporting.
AI use is also unresolved. Dream found Claude Code and configurations for DeepSeek and Kimi, but said it had no direct evidence those tools generated the exploit or attack scripts. The exact contribution, if any, of AI to reconnaissance, coding or operational decisions is not known. The number of exposed devices reported by Dream should not be read as the number of devices actually compromised or still vulnerable.
Patch and Check Management Access
Organizations running affected PAN-OS versions should consult Palo Alto Networks’ current security guidance, install the applicable fixed build and keep firewall management interfaces restricted to trusted networks rather than the public internet. Teams should also review relevant firewall and identity-system logs for signs of unauthorized access, and investigate whether management servers or Active Directory systems were reached. Applying a patch alone may not determine whether a device was already compromised, so incident review remains relevant for exposed systems.
Further reporting would need to clarify the government-affiliated organization’s impact, the scope of the Active Directory activity and any confirmed data access. Palo Alto’s advisory and subsequent fixed releases, alongside CISA’s catalog entry and federal deadline, provide the immediate response milestones. The extent to which AI tools contributed to the operation remains a separate open question.
Key Questions
What was PHP2Shell?
PHP2Shell refers to two vulnerabilities, CVE-2026-109618 and CVE-2026-109619, described in the report as enabling code execution through affected PAN-OS firewall management services. Dream says the attacker chained them with the Fragnesia Linux kernel flaw to reach root access.
How many servers did Dream say were compromised?
Dream reported 24 rooted firewalls and one internal firewall-management server, for 25 servers in total. It said two of the firewalls belonged to a government-affiliated organization, from which the attacker reached the internal server.
Did AI build or carry out the attack?
That has not been established. Dream found AI coding and model tools in the attacker’s environment but said it had no direct evidence they wrote the exploits or scripts. It also said the operation did not appear fully autonomous.
What should PAN-OS administrators do?
Administrators should follow Palo Alto Networks’ security advisory, install the fixed build that applies to their version and restrict management access from the public internet. Organizations should also assess exposed devices for signs of compromise, since updating does not by itself show whether an earlier intrusion occurred.
Source: ThorstenMeyerAI.com
