CVE-2026-96893

Undergoing Analysis Undergoing Analysis - In Progress

Authenticated AI Agent Instruction Injection in GitLab EE

Publication date: 2026-10-13

Last updated on: 2026-10-14

Assigner: GitLab Inc.

Description

GitLab has remediated an issue in GitLab EE affecting all versions from 19.1 before 19.2.8, 19.3 before 19.3.5, and 19.4 before 19.4.2 that, under certain conditions, could have allowed an authenticated user to inject specially crafted content into AI agent instructions, potentially causing unintended command execution attributed to other users, due to improper neutralization of special elements in AI conversation context construction.

CVSS Scores

EPSS Scores

Probability:
Percentile:

Meta Information

Published

2026-10-13

Last Modified

2026-10-14

Generated

2026-10-19

AI Q&A

2026-10-13

EPSS Evaluated

2026-10-18

NVD

CVE-2026-96893

EUVD

EUVD-2026-98213

Affected Vendors & Products

Vendor	Product	Version / Range
gitlab	gitlab	From 19.3.0 (inc) to 19.3.5 (exc)
gitlab	gitlab	From 19.4.0 (inc) to 19.4.2 (exc)
gitlab	gitlab	From 19.1.0 (inc) to 19.2.8 (exc)

Helpful Resources

Exploitability

CWE

KEV

CWE ID	Description
CWE-74	The product constructs all or part of a command, data structure, or record using externally-influenced input from an upstream component, but it does not neutralize or incorrectly neutralizes special elements that could modify how it is parsed or interpreted when it is sent to a downstream component.

Attack-Flow Graph

Executive Summary

This vulnerability in GitLab EE allows an authenticated user to inject crafted content into AI agent instructions. This could lead to unintended command execution being attributed to other users due to improper handling of special elements in AI conversation context.

Detection Guidance

This vulnerability requires checking GitLab versions for affected releases (19.1 before 19.2.8, 19.3 before 19.3.5, or 19.4 before 19.4.2). Use commands like 'gitlab-rake gitlab:env:info' or check version via UI/API. No direct detection commands are provided in the context.

Impact Analysis

An attacker could manipulate AI agents to execute unauthorized commands under another user's identity. This may result in data breaches, unauthorized access, or actions performed without consent, affecting system integrity and trust.

Compliance Impact

This vulnerability could lead to unauthorized data access or actions, violating GDPR's data protection principles and HIPAA's security rules. Non-compliance may result in legal penalties, reputational damage, and loss of trust.

Mitigation Strategies

Upgrade GitLab EE to version 19.2.8, 19.3.5, or 19.4.2 or later to address the vulnerability.

Hi! I’m here to help you understand CVE-2026-96893. Ask me anything about the vulnerability, its impact, or mitigation strategies.

0/70