🔍 Read the full analysis: What The PHP2Shell Attack Reveals About AI-Assisted Hacking on ThorstenMeyerAI.com
Get business pricing on monitors, keyboards and dev gear
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
TL;DR
Dream reported that an attacker used two vulnerabilities to gain root access on 24 Palo Alto Networks firewalls and then entered a government-affiliated network, compromising an internal firewall-management server. Dream found AI tools in the attacker’s environment but said it had no direct evidence they created the exploit or scripts; the incident also highlights exposed management interfaces and the delay before fixed PAN-OS builds were available.
Security firm Dream reported that an attacker gained root access to 24 Palo Alto Networks PAN-OS firewalls in an automated campaign shortly after fixes for two PHP vulnerabilities were released. The attacker then used compromised equipment to reach a U.S. government-affiliated network and take control of an internal firewall-management server; Dream found AI tools in the attacker’s environment but said their role in developing or running the attack is unproven.
Dream’s October 26 account, updated October 29, says the campaign targeted 240 firewalls across the United States, Japan, Taiwan and South Korea. It reports root access on 24. Palo Alto Networks said its Unit 42 researchers independently verified the attacker’s tools and results, according to Dream. The firm’s account describes a total of 25 compromised servers: the 24 firewalls and one internal management server.
The attacker chained PHP2Shell vulnerabilities CVE-2026-109618 and CVE-2026-109619 with Fragnesia, a Linux kernel privilege-escalation flaw assigned CVE-2026-46300 and publicly disclosed in May, Dream said. The PHP bugs enabled code execution on the firewall’s management web server; Fragnesia was then used to obtain administrator-level access. Dream says 22 rooted firewalls were left with hidden web shells, while two firewalls at the government-affiliated organization provided a route into its network.
Dream’s timeline places the automated attack on October 23, two days after PHP and Palo Alto Networks issued advisories. Palo Alto’s fixed builds for the affected PAN-OS 12.1 and 12.2 branches became available between October 27 and 29, according to the source account. Before those builds were released, the stated mitigation was to keep the management interface off the public internet.
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 Firewall Management Interfaces
The incident shows how a vulnerability in a management service can become a route from an internet-facing device into an organization’s internal systems. Dream says the attacker moved from compromised perimeter firewalls to a server used to manage the organization’s firewalls, then targeted Active Directory, which controls access and logins. The account does not say that the attacker succeeded in taking over Active Directory.
The timing also matters for defenders. An advisory can identify a risk before a vendor has shipped a fixed build for every affected product branch. In the interval, exposure reduction—particularly restricting management access to trusted networks—can limit opportunities for attackers. Dream’s report describes a campaign that followed public disclosure of an older kernel flaw with rapid targeting of devices whose management pages were reachable online.
The AI angle should be treated cautiously. Finding AI software on an attacker’s server may indicate that such tools were available to the operator; it does not establish that AI generated the exploit, selected targets, or carried out the intrusion. The confirmed security lesson is not that an autonomous system breached the network, but that known flaws and exposed administration services were combined in a successful attack.
Popular right now
Breville Bambino Espresso Machine, Brushed Stainless Steel (BES450BSS)
Apple AirPods Pro 3 Wireless Earbuds with Active Noise Cancellation
Dyson Cyclone V10 Origin Cordless Vacuum
As an affiliate, we earn on qualifying purchases.
From PHP Fixes to PAN-OS Builds
On October 21, PHP released updates for the affected PHP branches, while Palo Alto Networks published an advisory about the firewall issue. The source account says the affected PAN-OS software contained its own copy of PHP, version 8.0.30. Fixed PAN-OS builds for the 12.1 and 12.2 branches arrived from October 27 through October 29, after the reported October 23 campaign.
Dream says the attacker first identified roughly 19,000 firewalls with internet-facing management interfaces, then gathered version information from about 5,000 and narrowed the pool to 240 targets. These are figures from Dream’s reconstruction, not a measure of all vulnerable firewalls. The company also reports that the attacker had set up an offensive-security platform called aiscan on October 8, with DeepSeek and Kimi configured.
The earlier Fragnesia vulnerability was disclosed in May with a public exploit and a kernel fix, according to the source material. Its presence in the chain meant the attacker could combine a newly relevant weakness in the firewall’s PHP service with a separate, already-known route to higher privileges.
““No direct evidence” shows that AI tools wrote the exploit or scripts.”
— Dream
AI’s Role Remains Unverified
Dream found Claude Code installed and DeepSeek and Kimi configured in aiscan, but it said it had no direct evidence those tools produced the exploit or attack scripts. The report does not establish whether AI helped with reconnaissance, coding, troubleshooting or another part of the operation. Its presence is evidence of access to the tools, not proof of how they were used.
The source account also does not identify the attacker or provide a confirmed motive. It says the attacker targeted Active Directory after reaching the internal management server, but does not confirm that the directory was compromised or specify what data, if any, was accessed or taken. The extent of any impact beyond the reported servers remains unclear from the material provided.
Dream’s figures and reconstruction describe the campaign as it understood it; the full set of affected organizations and whether additional compromised devices were found are not established here. The account also does not provide a complete remediation status for every targeted firewall.
Patch and Restrict Management Access
Organizations using affected PAN-OS branches should apply the fixed builds Palo Alto Networks released between October 27 and 29, following the vendor’s advisory. Where updates have not yet been applied, the source identifies keeping the management interface off the public internet as a protective measure. Administrators should also review firewall and network-management systems for signs of compromise and follow vendor and government guidance.
U.S. federal agencies were given three days after CISA added both PHP2Shell CVEs to its Known Exploited Vulnerabilities catalog on October 26, according to the source account. Further reporting or incident-response updates may clarify the government-affiliated organization’s exposure, whether Active Directory was affected, and whether the attacker used AI tools during any part of the operation.
Key Questions
What happened in the PHP2Shell attack?
Dream reported that an attacker chained two PHP vulnerabilities with the known Fragnesia Linux kernel flaw to gain root access on 24 PAN-OS firewalls. The attacker then used compromised firewalls to reach an internal management server in a government-affiliated network.
Did AI create or carry out the attack?
That has not been established. Dream found AI tools on the attacker’s systems but said it had no direct evidence they wrote the exploit or scripts, and that the operation did not appear fully autonomous.
What is confirmed about the government-affiliated network?
Dream says two of the compromised firewalls belonged to the organization and that the attacker reached and took over an internal firewall-management server. The report says the attacker then went after Active Directory, but does not confirm a successful takeover of it.
What should PAN-OS administrators do?
Apply the fixed PAN-OS builds for affected branches and follow Palo Alto Networks’ advisory. Dream’s account identifies keeping the management interface inaccessible from the public internet as protection before a fix is applied.
Source: ThorstenMeyerAI.com
