The Long Road To PHP2Shell: Revisiting The Story With AI
AIThis post was created with the assistance of artificial intelligence (AI).

🔍 Read the full analysis: The Long Road To PHP2Shell: Revisiting The Story With AI on ThorstenMeyerAI.com

Age 18–24?Offer from Amazon

Prime made for students and young adults

  • Fast, free delivery for dorm and study essentials
  • Prime Video and Amazon Music included
  • Member-only deals
Try Prime for Young Adults Free trial for eligible 18–24 year olds
As an affiliate, we earn on qualifying purchases.

TL;DR

PHP released fixes on October 21, 2026, for two critical remote code execution vulnerabilities known as PHP2Shell. A retrospective published by Thorsten Meyer traces the issue’s history to a 2004 warning and describes how researchers at Calif and Wiz used AI-assisted methods in separate investigations. The precise technical details and the full scope of affected systems are not established in the supplied account.

PHP released fixes on October 21 for two critical remote code execution vulnerabilities, CVE-2026-109618 and CVE-2026-109619, known as PHP2Shell. A retrospective published the next day by Thorsten Meyer traces the flaws to a long-running security concern and describes separate investigations by Calif and Wiz Research that used AI tools alongside human analysis.

Meyer’s account frames PHP2Shell as more than a story about recently released AI models. It follows a trail from a 2004 security advisory, through a 2019 PHP bug report about out-of-memory failures and temporary files, to research that led to the 2026 vulnerabilities. The underlying concern is that a request interrupted at an unsafe point may leave partially prepared data structures or other state for cleanup code to handle.

According to the retrospective, Calif identified several possible paths to remote code execution in August 2026 and reported two findings to PHP. Its timeline says the first report went to the PHP team on August 11, followed by a working proof of concept on August 14; PHP confirmed it the next day. Calif reported a second issue on August 21. Wiz Research says it found a PHP vulnerability on August 19, validated unauthenticated code execution against Drupal and WordPress in the following days, and reported it to PHP on August 24.

The account says the PHP security team made the fixes, released on October 21. It does not provide enough technical detail to establish precisely how the two vulnerabilities differ, which PHP versions are affected, or what administrators should do beyond applying the relevant fixes. The retrospective points readers seeking remediation to a separate patch-and-restart guide.

At a glance
reportWhen: PHP fixes shipped October 21, 2026; ret…
The developmentA retrospective published October 22 recounts how researchers used AI-assisted review to find two PHP remote code execution flaws that PHP patched the previous day.
Thorsten Meyer AI
2004 → 2026

Infographic 01 · Timeline

One class of bug, four chances to catch it.

The flaws fixed on 21 October 2026 continue a line that starts with a 2004 advisory about unsafe interruptions.

22years from the 2004 advisory to the 2026 fix
2CVEs, both rated 9.2 (CVSS v4.0)
3research groups credited: Calif, OpenAI, Wiz
4fixed PHP branches on 21 Oct 2026
  1. Jul 2004
    Stefan Esser: memory_limit can stop PHP “where an interruption is unsafe”Advisory 11/2004, CAN-2004-0594. Remote code execution was possible. (e-matters)
  2. 28 Nov 2019
    Bug #78876: long upload fields cause out-of-memory, temp files left behindFixed May 2020 as CVE-2019-11048, a denial-of-service issue at the time. (PHP bug tracker)
  3. 2024
    Quarkslab audit with OSTIF and the PHP FoundationAccording to Calif, it rediscovered the same OOM issue; exploitation was not proven, so it stayed open. (Calif, OSTIF)
  4. 30 Jul 2026
    Calif joins OpenAI’s Patch the Planet and picks PHPStefan Esser brings in the PHP Foundation through Volker Dusch. (Calif)
  5. 1–21 Aug 2026
    Calif finds several RCE paths and reports two bugs to PHPFirst report 11 Aug, working proof of concept 14 Aug, confirmed by PHP 15 Aug, second report 21 Aug. (Calif)
  6. 19–24 Aug 2026
    Wiz Research finds its first PHP flaw and reports on 24 AugUsing Wiz Atlas, Gemini 4 Argon and GPT-5.6 Sol; RCE confirmed on Drupal and WordPress. (Wiz)
  7. 21 Oct 2026
    PHP 8.5.12, 8.4.27, 8.3.36 and 8.2.35 releasedCVE-2026-109618 and CVE-2026-109619, both CVSS v4.0 9.2. Public disclosure the same day. (PHP)

“…possible for a remote attacker to trigger the memory_limit request termination in places where an interruption is unsafe.”Stefan Esser, e-matters Advisory 11/2004

Sources: e-matters Advisory 11/2004; PHP bug #78876; OSTIF; Calif: php2shell; Wiz Research (21 Oct 2026); PHP security advisory (21 Oct 2026).
Powered by Thorsten Meyer AI

Why PHP Runtime Security Matters

PHP runs a large share of web applications, and the reported ability to execute code remotely could have serious consequences if vulnerable deployments are exposed. The account’s references to validation against Drupal and WordPress illustrate why runtime flaws can matter beyond PHP specialists: applications built on the language may depend on its behavior. The source does not quantify affected installations or report confirmed exploitation in the wild.

The retrospective also offers a qualified account of AI’s role in security research. It attributes meaningful assistance to models and research platforms, but describes the process as guided by researchers’ experience with PHP and its older bug history. That distinction matters when interpreting claims about automated vulnerability discovery: the reported work does not establish that AI independently found the flaws or can identify all vulnerabilities.

Amazon

PHP security patch management tools

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

From a 2004 Warning to PHP2Shell

In July 2004, security researcher Stefan Esser published an advisory about PHP’s memory limit. As quoted in Meyer’s account, Esser warned that an attacker could trigger request termination at a point where interruption was unsafe, potentially enabling code execution. PHP fixed the specific paths identified then, but the retrospective says the wider class of risk—failure during sensitive operations—proved harder to eliminate.

A PHP tracker report filed on November 28, 2019, described large multipart form fields causing out-of-memory failures and leaving temporary files behind. PHP fixed that issue in May 2020; it received CVE-2019-11048 and was treated as a denial-of-service problem. Calif later viewed the leftover files as a clue to ask what other state might survive an out-of-memory event.

The 2024 PHP audit, conducted by Quarkslab with OSTIF and the PHP Foundation, listed 27 findings and four CVEs, according to Meyer. His account says Calif reported that auditors had again seen the out-of-memory issue, but that exploitation was not proven. OSTIF’s published summary does not mention that issue, so the retrospective’s account of the audit finding should be treated as an attributed claim, not as a detail confirmed by that summary.

“possible for a remote attacker to trigger the memory_limit request termination in places where an interruption is unsafe”

— Stefan Esser, in his July 2004 advisory, as quoted by Thorsten Meyer

Amazon

web application vulnerability scanner

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

Affected Versions and Exploitation Status

The supplied retrospective does not specify which PHP versions contain the flaws, the exact vulnerability mechanics, or the complete set of affected products and configurations. It also does not report whether attackers exploited either flaw before the fixes became available. The claims about Calif’s findings and Wiz’s validation are presented through the retrospective and cited research announcements; full technical advisories are not included in the source material.

The account says Calif and Wiz reached the same class of flaw independently, but the supplied timeline does not fully explain how their findings overlap or map to the two CVE identifiers. It is also unclear how widely PHP2Shell has been adopted as a name beyond the coverage described by the author.

Amazon

PHP debugging and troubleshooting tools

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

Patch Guidance and Disclosure Details

PHP users should consult the PHP security team’s release information to identify affected versions and apply the appropriate fixes. Meyer’s article directs administrators to a separate guide covering patching, restarting services, and checking deployments; the supplied material does not reproduce its steps or the official release notes.

Further clarity depends on the PHP team’s advisories and technical disclosures from Calif and Wiz. Those materials can establish the precise scope of the vulnerabilities, how the fixes address them, and whether there is evidence of exploitation. Until then, the confirmed development is that PHP shipped fixes on October 21, 2026; operational details remain incomplete in the retrospective.

Amazon

PHP security best practices book

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

Key Questions

What is PHP2Shell?

PHP2Shell is the name used in the retrospective for two critical PHP remote code execution vulnerabilities, CVE-2026-109618 and CVE-2026-109619. PHP released fixes on October 21, 2026.

Who found the vulnerabilities?

The account credits research to Calif, OpenAI and Wiz. It describes Calif and Wiz as investigating separately, with AI models or platforms assisting parts of their work.

Did AI discover PHP2Shell on its own?

The source does not support that conclusion. It describes researchers choosing targets and drawing on prior knowledge, with AI tools contributing to the investigations. Calif’s Ryan Austin cautioned that no model can find all bugs.

Are PHP2Shell flaws being exploited?

The supplied account does not say whether either flaw was exploited in the wild. It reports testing of unauthenticated code execution against Drupal and WordPress, which is not the same as evidence of real-world attacks.

What should PHP administrators do?

Check the PHP security team’s release information for affected versions and apply the relevant fix. The retrospective also points to a separate guide on patching, restarting, and checking a PHP deployment.

Source: ThorstenMeyerAI.com

HALLOWEEN

Halloween Picks

As an affiliate, we earn on qualifying purchases.

You May Also Like

The New Personal Agent Layer

OpenClaw and Hermes introduce a new layer of persistent personal action agents, enabling AI to act across digital environments with memory and tool use.

Enhance Your B2B Sales Funnel With Self-Qualifying Contact Tools

A new AI-powered widget enables B2B SaaS companies to automatically qualify leads, saving time and increasing sales pipeline quality.

The Test AI Models Keep Passing — and the One That Actually Matters

All five frontier AI models refused every manipulation — but only two closed the €55k deal they’d earned. Firmulate’s live wargame grades management, not chat.

14 Best AI Automation Software Tools for Smarter Workflows in 2026

Explore the 14 best AI automation software tools for 2026, including agent builders, coding assistants, and office copilots, to optimize workflows.