Automated Threat Modeling

Eclipse GlassFish CVE-2026-2586: Brief Summary of a Critical Expression Language Injection Leading to Remote Code Execution

A brief summary of CVE-2026-2586, a critical Expression Language injection vulnerability in the Eclipse GlassFish Administration Console that lets authenticated administrators execute arbitrary OS commands through crafted requests.

CVE Analysis

6 min read

ZeroPath CVE Analysis
ZeroPath CVE Analysis

2026-05-19

Eclipse GlassFish CVE-2026-2586: Brief Summary of a Critical Expression Language Injection Leading to Remote Code Execution
Experimental AI-Generated Content

This CVE analysis is an experimental publication that is completely AI-generated. The content may contain errors or inaccuracies and is subject to change as more information becomes available. We are continuously refining our process.

If you have feedback, questions, or notice any errors, please reach out to us.

[email protected]

Introduction

An Expression Language injection flaw in the Eclipse GlassFish Administration Console lets an authenticated administrator execute arbitrary operating system commands on the underlying host. It carries a CVSS score of 9.1. The reporting researchers have already built functional exploit tooling and shared it with the Eclipse Foundation, so the gap between disclosure and patch availability is a meaningful risk for organizations running affected versions.

Eclipse GlassFish is an open source application server that fully implements the Jakarta EE platform. It is designed for high availability through clustering, scalability, and load balancing. Data indicates that more than 1200 companies use it, which gives it a significant presence in the enterprise Java ecosystem. Its Administration Console is the primary web interface for managing server configurations, deployments, and resources.

Technical Information

Root Cause: Expression Language Injection

The vulnerability comes from improper neutralization of Expression Language (EL) statements in the GlassFish Administration Console. It is classified under both CWE-94 (Code Injection) and CWE-917 (Improper Neutralization of Special Elements used in an Expression Language Statement).

Security researchers Gabriel A. Hinostroza Ayala and Camilo Galdos found that the alertSummary and alertDetail parameters on multiple administration console endpoints do not sanitize user input before the EL engine processes it. The primary confirmed vulnerable endpoint is:

/web/configuration/virtualServerEdit.jsf

Attack Flow

Exploitation follows these steps:

  1. An attacker authenticates to the GlassFish Administration Console with valid administrator credentials.
  2. The attacker navigates to (or directly sends a crafted request to) a vulnerable endpoint such as /web/configuration/virtualServerEdit.jsf.
  3. The attacker injects a malicious Expression Language payload into the alertSummary or alertDetail parameter.
  4. The server-side EL engine evaluates the injected expression without proper sanitization.
  5. The payload invokes java.lang.Runtime, which allows direct execution of arbitrary operating system commands.
  6. The commands run with the privileges of the GlassFish service account on the host system.

Scope and Impact

The CVSS 3.1 vector string for this vulnerability is:

CVSS:3.1/AV:N/AC:L/PR:H/UI:N/S:C/C:H/I:H/A:H

The attack requires high privileges (an authenticated administrator), but it works over the network with low complexity and needs no user interaction. The scope change (S:C) indicates that compromising the GlassFish application leads directly to compromise of the underlying host environment. The result is high impact to confidentiality, integrity, and availability.

The reporting researchers gave the Eclipse Foundation comprehensive reproduction materials: a functional proof of concept, an automated exploit script named exploitSSTI.py, a demonstration video, and a detailed list of vulnerable endpoints. These artifacts confirm that the vulnerability is highly reproducible, and they lower the barrier to exploitation.

Affected Systems and Versions

ComponentVersionStatus
Eclipse GlassFish7.1.0Confirmed vulnerable
Eclipse GlassFish8.0.0Confirmed vulnerable
Eclipse GlassFish8.0.1Under review by maintainers
glassfish jsftemplating< 4.2.0Vulnerable (core EL processing component)
glassfish jsftemplating4.2.0Contains the fix for EL expression restrictions
Eclipse GlassFish8.0.2Integrates the patched jsftemplating component

GlassFish 7.1.0 and 8.0.0 are confirmed affected. Maintainers are still checking whether version 8.0.1 belongs on the affected list. The fix has been merged into glassfish jsftemplating 4.2.0, which blocks access to public static fields, calls to static methods, and all reflective access within EL expressions. Full remediation is scheduled for GlassFish 8.0.2, which integrates the patched component.

Until GlassFish 8.0.2 is deployed, organizations should consider disabling the Administration Console entirely if daily operations do not require it. If the console must stay active, restrict network access to trusted management IP addresses and audit all authenticated administrator accounts.

Vendor Security History

The GlassFish Administration Console has had multiple earlier security vulnerabilities:

CVEVulnerability TypeAffected Component
CVE-2024-10029Reflected cross-site scriptingAdministration Console in version 7.0.15
CVE-2024-8646URL redirection to untrusted sitesVersions prior to 7.0.10
CVE-2017-3249Unauthenticated network compromiseOracle GlassFish Server via LDAP

This run of flaws in the administration interface is a good reason to treat the management plane as a high-value target and to put it behind strict network segmentation and access controls. The Eclipse Foundation has responded actively to CVE-2026-2586. Maintainers have merged fixes into the templating component and scheduled the platform release.

References

Detect & fix
what others miss

Works with
  • GitHub
  • GitLab
  • Bitbucket
  • Azure DevOps Services
  • Gerrit
  • CVS
  • Jira
  • Linear
  • Slack
  • Security Compass
Explore integrations
Security magnifying glass visualization