CISA Confirms PHP2Shell Is Being Exploited in the Wild
The agency added both PHP flaws to its Known Exploited Vulnerabilities catalogue on 26 October 2026, a week after saying it had seen no successful exploitation, and gave federal civilian agencies until 29 October to act.
CISA added the two PHP flaws known collectively as PHP2Shell to its Known Exploited Vulnerabilities catalogue on 26 October 2026, listing CVE-2026-109618 as a PHP memory corruption vulnerability and CVE-2026-109619 as a PHP stale internal request-state vulnerability. The agency updated its standing alert, Critical Vulnerabilities in PHP Require Immediate Patching, in the same notice. Both flaws score 9.2 on CVSS and allow unauthenticated remote code execution.
The listing is a reversal. The Register reported that a week earlier CISA had warned about the danger PHP2Shell posed while saying it had not yet observed successful exploitation in the wild. Adding a vulnerability to the catalogue means the agency now considers it actively exploited.
A shell on a honeypot, then the catalogue
What moved in between was evidence from outside government. The Hacker News reported that Adel Karimi, detection and response lead at Perplexity, confirmed successful remote code execution against honeypot systems on 24 October 2026. Computer Weekly reported that the detection involved an Apache mod_php installation running PHP 8.4.26 in a sequestered honeypot environment, and that it emerged publicly on Sunday 25 October.
Karimi was careful about what the result proves.
The evidence demonstrates this chain on the intentionally favorable single-worker mod_php profile; it is not a measurement of exploit reliability on other worker counts, PHP builds, or applications.
That caveat matters because the flaws reach almost every way of running PHP against web traffic. CISA's advisory names Apache with mod_php, PHP-FPM, php-cgi, FrankenPHP and LiteSpeed, and singles out mod_php as carrying particularly high risk. The only demonstrated chain so far runs against the narrowest of those configurations.
Parallax Nexus reported on 24 October that botnet traffic, a working ASLR bypass and a publicly browsable attacker toolkit had all appeared around PHP2Shell without any vendor reporting a successful compromise. That is the gap the honeypot result and the catalogue entry have now closed, at least on the question of whether the flaws work.
What the 29 October date actually requires
Federal civilian executive branch agencies have until 29 October 2026 to act. The obligation runs through Binding Operational Directive 26-04, Prioritizing Security Updates Based on Risk, which tells agencies to prioritise rapid remediation of catalogued vulnerabilities on publicly exposed assets that grant total control of the asset after exploitation, and to defer lower-risk work. The directive also sets expectations for when agencies must check whether attackers compromised a system before the patch was applied. CISA encourages non-federal organisations to follow the same approach without binding them to it.
The patches have existed since 21 October 2026, in PHP 8.5.12, 8.4.27, 8.3.36 and 8.2.35. Censys estimates more than 32.51 million PHP services still run an affected version, and warns that hosts on end-of-life branches between PHP 5.4 and 7.3 will never receive a fix and have to move to a supported branch. W3Techs puts PHP behind roughly 70 per cent of websites. On 26 October the PHP Foundation also shipped Windows builds 8.5.12-1, 8.4.27-1, 8.3.36-1 and 8.2.35-1 to address CVE-2026-111497, an out-of-bounds write in the bundled Brotli decoder.
What CISA has not said is who was hit. The Register noted that the agency disclosed neither the compromised organisations nor any detail of the attacks behind the decision, and that the listing confirms exploitation without confirming how many production systems were successfully breached. Dataminr, publishing on 27 October, went further and said it remained difficult to identify active successful exploitation from a reputable source.
The honest reading is that the catalogue entry is a procurement instrument, not a damage report. It converts a contested research finding into a legal deadline for several hundred federal networks, which is useful even if the underlying exploitation turns out to be narrow and opportunistic. The number worth watching is not 32.51 million exposed services but how many of them sit behind mod_php on a single worker, the one configuration anybody has demonstrated a reliable chain against.
What happens next?
- Federal civilian agencies must remediate or mitigate both CVEs by 29 October 2026 and, under BOD 26-04, assess whether compromise preceded patching.
- Vendors that embed PHP in management interfaces are still publishing exposure assessments, which will determine how much of the installed base can be patched by its owners at all.
- Whether CISA names affected organisations or publishes indicators will determine how much defenders can do beyond patching.
- Hosts on end-of-life PHP branches from 5.4 to 7.3 have no patch path and will need a branch migration rather than an update.
Related topics
Sources & references
- 01CISA Adds Two Known Exploited Vulnerabilities to Catalog — Cybersecurity and Infrastructure Security AgencyprimaryAlert of 26 October 2026 naming both CVEs and setting out the BOD 26-04 obligation.
- 02CISA Confirms Active Exploitation of Two PHP2Shell Flaws Enabling Remote Code Execution — The Hacker NewsnewsReported 27 October 2026; source of the Karimi honeypot finding and quote, the 29 October deadline, patch versions and the Censys and W3Techs figures.
- 03CISA confirms PHP2Shell under attack as researchers clock successful exploits — The RegisternewsReported 27 October 2026; source for the shift from CISA's earlier position and for what the agency has not disclosed.
- 04PHP2Shell: US updates Kev catalogue as exploits spread — Computer WeeklynewsDetail of the mod_php 8.4.26 honeypot detection and the 25 October timeline.
- 05Examining the Vulnpocalypse and Defensive Responses — DataminrreportIntel brief of 27 October 2026 dissenting on how firmly successful exploitation can be established.
More from Security
Stolen GitHub Tokens Put Malware in a LiteLLM Release
A compromised release pipeline put a poisoned LiteLLM build on PyPI and Docker for under three hours on 21 October 2026. It harvested cloud secrets, AI provider API keys and SSH keys, and Wiz says the attackers were spending stolen AWS credentials about four and a half hours later.
Attackers Are Escalating to Root Through a Japanese Web Firewall
A command injection flaw in the SiteGuard Server Edition web application firewall, CVE-2026-106931, is being exploited to run commands as root or SYSTEM. JPCERT/CC alerted users on 14 October 2026 and the vendor told support customers on 29 September.
China-Linked Actor Had Root on BeyondTrust Since August
A critical flaw in BeyondTrust Remote Support appliances, CVE-2026-102764, was exploited as a zero-day in late August 2026 by a suspected PRC-nexus espionage actor that Google tracks as ARMORY CASTLE. BeyondTrust patched its SaaS tenants on 9 October and disclosed the vulnerability on 15 October. Self-hosted customers have to apply the patch themselves.
China-Linked Scanner Holds 1,300 Exploit Scripts, FBI Says
A joint advisory led by the FBI describes MicroScan, a Python web application holding more than 1,300 penetration-testing scripts that China-linked actors have used since at least 2017. CISA added five of the eight named flaws to its exploited-vulnerability catalogue with a federal deadline that expired on 11 October.



