🔍 Read the full analysis: What “AI-Assisted” Looks Like In A Real PHP2Shell Attack on ThorstenMeyerAI.com
Get the latest gadgets delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
TL;DR
Dream reported that an attacker used PHP2Shell and a Linux kernel flaw to gain root access on 24 PAN-OS firewalls, then reached a server inside a US government-affiliated network. Dream found AI tools on the attacker’s servers but said it has no direct evidence they wrote the exploit or scripts; Palo Alto Networks says Unit 42 independently verified the tools and results.
An attacker gained root access to 24 PAN-OS firewalls and used two compromised devices to reach a server inside a US government-affiliated network, according to Israeli security company Dream. Its October 26 report, updated October 29, says the attacker combined the PHP2Shell vulnerabilities with a previously disclosed Linux kernel flaw; Palo Alto Networks says its Unit 42 researchers independently verified the attacker’s tools and results.
Dream says the attacker ran an automated batch against 240 firewalls on October 23 and obtained root access on 24. The targets were in the United States, Japan, Taiwan and South Korea. Dream reported that 22 of the rooted firewalls were left with hidden web shells, while two belonged to the government-affiliated organization. On October 24, the attacker used those devices to reach the organization’s internal firewall management server, bringing the reported total to 25 compromised servers. Dream says the attacker then targeted Active Directory, which manages accounts and access within an organization.
The reported intrusion chained PHP2Shell, CVE-2026-109618 and CVE-2026-109619, with Fragnesia, CVE-2026-46300, a Linux kernel privilege-escalation flaw disclosed in May. Dream says PHP2Shell allowed command execution under a low-privilege account on the firewall, while the kernel flaw enabled the attacker to obtain root. Dream estimates the chain, including establishing a persistent back door, could take under 15 minutes. These are descriptions and timing estimates in Dream’s account, not independently detailed findings in the supplied vendor statement.
The “AI-assisted” label comes from tools Dream found on attacker-controlled servers. The company reported that Claude Code was installed and that the Chinese AI models DeepSeek and Kimi were configured in an offensive-security platform called aiscan. Dream said it had “no direct evidence” that AI tools wrote the exploit or attack scripts and that the operation did not appear fully autonomous. The presence of tools does not establish how, or whether, they were used during the intrusion.
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
Internet-Facing Firewalls Were Exposed
The incident matters because management interfaces reachable from the public internet gave the attacker a route to systems that control network access. Dream’s account describes a rapid move from vulnerable firewalls to an internal management server and attempts to reach Active Directory, where access to many organizational resources can be governed. That makes the reported compromise more consequential than an isolated firewall intrusion.
The case also shows why the AI angle should not displace established security lessons. Dream’s evidence supports the presence of AI-related software on attacker infrastructure, but not that AI created the exploit or carried out the operation autonomously. The reported attack relied on known vulnerabilities and exposed management services. The vendor’s advisory and the timing of fixed software builds are central to assessing organizations’ exposure.
Popular right now
TCL QM64L 85-Inch QD-Mini LED 4K Smart Fire TV
Vitamix Propel Series 750 Professional-Grade Blender, 4 Programs, Black
Sony BRAVIA 5 85-Inch Mini LED 4K HDR Smart TV (K-85XR50)
As an affiliate, we earn on qualifying purchases.
Patch Dates and Attack Timeline
Dream places the attack after the relevant PHP project fixes were released on October 21. Palo Alto Networks published an advisory that day, but Dream says fixes for the affected PAN-OS 12.1 and 12.2 branches arrived between October 27 and 29. The source account says PAN-OS includes its own PHP copy, identified by Dream as version 8.0.30. Until fixed builds were available, Palo Alto’s stated protective measure was to keep the management interface off the internet.
Dream says its researchers identified about 19,000 PAN-OS firewalls with internet-facing management pages using FOFA, a search engine for internet-connected devices. A custom scanner obtained version information from about 5,000, and filtering by country and software version narrowed the potential targets to 240. These figures describe Dream’s reported discovery and selection process; they do not establish that all 19,000 devices were vulnerable or compromised.
The first reported batch ran on October 23, with the government-affiliated organization’s first firewall compromised that evening. Dream says a second firewall was breached the next morning, followed by the internal management server. The company published its findings October 26 and updated them October 29. On October 26, CISA added both PHP2Shell vulnerabilities to its Known Exploited Vulnerabilities catalog and set a three-day deadline for federal agencies, according to the supplied source material.
““no direct evidence””
— Dream
AI’s Role Remains Unverified
How the AI tools were used is not established. Dream found Claude Code installed and DeepSeek and Kimi configured in aiscan, but said it had no direct evidence that they generated exploit code or scripts. The available account does not establish whether the tools were used for research, coding, target selection, or another task, nor how much human involvement the operation required.
The supplied reporting also does not identify the attacker or provide a complete account of the government-affiliated organization’s response. It says Active Directory was targeted, but does not specify whether the attacker obtained access to it, what data or systems may have been affected, or whether the intrusion led to broader disruption. Dream’s counts and technical timeline are attributed to the company; details beyond the reported independent verification by Unit 42 remain limited in the source material.
Patch Deployment and Incident Review
Organizations using affected PAN-OS branches need to check Palo Alto Networks’ advisories and deploy the fixed builds applicable to their systems. The supplied source says fixes for 12.1 and 12.2 shipped between October 27 and 29; administrators should confirm their installed version and whether management access is exposed publicly. Palo Alto’s stated interim protection was to keep the management page off the internet.
For organizations that may have been exposed, the next steps include investigating for signs of compromise, reviewing firewall and identity-system activity, and assessing whether credentials or access paths need to be reset. The source material does not report a final impact assessment or explain whether the targeted organization contained the intrusion. Further statements from the affected organization, Palo Alto Networks, Dream or government agencies could clarify the extent of the breach and the attacker’s use of AI tools.
Key Questions
What happened in the PHP2Shell attack?
Dream says an attacker combined two PHP2Shell vulnerabilities with the Linux kernel flaw Fragnesia to gain root access on 24 PAN-OS firewalls and reach an internal management server in a US government-affiliated network.
Did AI write the exploit?
That is not confirmed. Dream found AI tools on attacker-controlled servers but said it had no direct evidence that they wrote the exploit or scripts.
How many systems were reported compromised?
Dream reported 24 rooted firewalls and one internal firewall management server, for a reported total of 25 servers. The account does not establish that all systems in the organization were compromised.
What should PAN-OS administrators do?
Check Palo Alto Networks’ advisory for the relevant fixed release, apply the applicable update, and keep the management interface off the public internet. Organizations that may have been exposed should also investigate for signs of compromise.
Source: ThorstenMeyerAI.com
