PHP2Shell: Two Critical PHP Vulnerabilities Demand Immediate Patching

Published 22 October 2026

Two critical remote code execution vulnerabilities have been disclosed in the PHP interpreter, affecting millions of websites worldwide and triggering an emergency out-of-cycle patch release from the PHP security team.

What happened

On October 20, 2026, details of two pre-authentication remote code execution flaws in PHP were leaked ahead of coordinated disclosure. The vulnerabilities, tracked as CVE-2026-109618 and CVE-2026-109619, were subsequently confirmed by the PHP security team, watchTowr, and multiple CERT and vendor advisories.

Both vulnerabilities are rated CVSS v4.0 9.2 (Critical). They have been dubbed PHP2Shell by researchers, reflecting the worst-case outcome: a single web request that gives an attacker a shell on the server. The PHP team released emergency patches on October 21, 2026 at 14:00 UTC, ahead of the originally scheduled disclosure date of October 22.

What the flaws are

CVE-2026-109618 is an out-of-memory corruption bug. When PHP exhausts its configured memory limit during request parsing, it can leave internal data structures only partially initialised. The subsequent cleanup routine then treats that uninitialised memory as valid, allowing an attacker to influence program execution.

CVE-2026-109619 is a use-after-free bug in global interpreter state. If a fatal error interrupts a request before cleanup completes, stale pointers can survive in worker-process memory. A later request handled by the same worker may then read or branch on that freed memory.

Both flaws sit in Zend, the C engine at the heart of PHP, so they are independent of the application code running on top. Updating WordPress, Drupal, or Magento alone does not fix them. The interpreter itself must be patched.

Who is affected

The PHP security team states that all PHP installations serving web requests are affected, including Apache mod_php, PHP-FPM, php-cgi, FrankenPHP, and LiteSpeed. The team considers mod_php deployments to be at highest risk, with a demonstrated exploit where an unauthenticated request triggers the attack before any application code runs.

Fixed versions are available for the four supported branches: PHP 8.5.12, 8.4.27, 8.3.36, and 8.2.35. PHP 8.1 and older are end-of-life and do not receive official fixes from the PHP project. The project has published minimal source backports for PHP 7.4, 8.0, and 8.1 to assist those building from source or without access to an LTS vendor.

CERT-EU and several security vendors have warned that PHP versions from 5.4 onward may be affected. Shodan data cited by Check Point Research shows roughly 884,000 internet-facing hosts advertising PHP 8.1 or older, plus another 740,000 reporting PHP without a version string. Those are banner counts, not confirmed vulnerable systems, but they indicate a large potential exposure.

Why this matters

PHP powers a substantial share of the web, including WordPress, WooCommerce, Drupal, Joomla, Magento, and many hosted and appliance-based products. Because the flaws are in the interpreter rather than in any single application, the attack surface is broad.

The exploitation mechanism is also unusually quiet. CrowdStrike verified that a single anonymous POST request can cause an Apache mod_php worker to launch a shell process and execute an attacker-specified command. The worker does not crash, the server returns a standard HTTP 200 response, and application logs record nothing unusual. That makes post-compromise detection harder than for a typical crash-and-restart exploit.

Cloudflare has deployed emergency WAF rules to all customers, including free plans, and reports observing early attacks. The company stresses that these protections are temporary and that administrators should install the official PHP updates as soon as possible.

What you should do

The PHP security team states plainly that there are no complete mitigations and that updating is the only safe choice. Organisations should:

Inventory every PHP instance, including PHP-FPM pools, container base images, vendor appliances, and hosting control panels. Stage the update path for each platform and schedule a change window around the October 21 release. Patch internet-facing hosts first, then restart PHP-FPM or the web server and verify the running version so the new interpreter actually loads.

Systems on end-of-life PHP versions should be flagged for migration. Where immediate patching is not feasible, CERT-EU advises isolating internet-facing applications from public networks until a fix can be applied. Temporary configuration changes such as disabling session upload progress or restricting multipart request handling reduce exposure to specific attack paths, but they do not make an installation safe.

Administrators should also review web server, WAF, and PHP error logs for unusual activity from August 1, 2026 onward, and hunt for new files in web roots, unexpected child processes spawned by PHP workers, and outbound connections from web servers to unknown destinations.

The bigger picture

This disclosure follows the now-deleted Hacker News post that first named the CVEs and their originally scheduled date. The leak forced the PHP team to move the release forward by one day, compressing the normal head start vendors rely on for testing and deployment.

History suggests the window for safe patching is short. The previous core PHP remote code execution vulnerability, CVE-2024-4577, drew botnet scanning within about 24 hours of disclosure. The same pattern is expected here.

Sources

PHP.net archive, PHP security advisory and release notes, October 21, 2026: https://www.php.net/archive/2026.php

BleepingComputer, “PHP warns admins of two critical flaws, emergency patches coming tomorrow,” October 20, 2026: https://www.bleepingcomputer.com/news/security/php-warns-admins-of-two-critical-flaws-emergency-patches-coming-tomorrow/

Sophos, Advisory SA-20261021-php2shell, October 21, 2026: https://www.sophos.com/en-us/security-advisories/sophos-sa-20261021-php2shell

CERT-EU, Advisory 2026-018, October 21, 2026: https://cert.europa.eu/publications/security-advisories/2026-018/

CrowdStrike, “CrowdStrike Protects Customers from Threats Delivered via PHP2Shell,” October 21, 2026: https://www.crowdstrike.com/en-us/blog/crowdstrike-protects-customers-from-php-rce-zero-days/

Check Point Research, “Patch Now: php2shell,” October 21, 2026: https://blog.checkpoint.com/security/php2shell-what-we-know-about-the-leaked-critical-php-vulnerabilities-cve-2026-109618-cve-2026-109619/

AWS, Security Bulletin 2026-143: https://aws.amazon.com/security/security-bulletins/2026-143-aws/

GitHub php-src security advisories: GHSA-w8hw-w9fx-j79w and GHSA-6wmf-4rv5-p3rp

How this analysis is researched, sourced and corrected: Editorial standards and corrections. Spotted an error? Contact Philip Hall.

Subscribe

Related articles

PHP2Shell: Two Critical Bugs, One Unavoidable Truth

# PHP2Shell: Two Critical Bugs, One Unavoidable TruthThe PHP...

AI Agents Breached Indian Companies Live Systems and Nobody Noticed

India's cyber agency CERT-In has issued a critical advisory after AI agents from multiple providers accessed live systems at Indian organizations, read personal data and confidential records, and none of the organisations detected the intrusion.

Meta and Apple Team Up to Put an 816B Open-Weight AI on Your Desk

Meta's Muse Spark Open is an 816B-parameter open-weight model that runs locally on two Mac Studios. It is the West's best open model, but China still leads in coding.

OpenAI’s Models Broke Into a Government Website, Spent Someone Else’s Test Crypto, and Said Nothing

OpenAI disclosed three new misalignment incidents where its AI models bypassed authorization on government sites, broke into research systems, and spent cryptocurrency from wallets they did not own. None of the models mentioned it in their final answers.
Philip Hall
Philip Hall
Philip Hall is a Sydney-based Cyber AI and Automation leader with more than 30 years of technology experience and a career in cyber security dating back to 2008. His work spans cyber architecture, cloud security, threat intelligence, assurance, incident support, AI-enabled defence and the security of autonomous agents.

This site uses Akismet to reduce spam. Learn how your comment data is processed.