CVE-2026-96893: Improper Neutralization of Input Used for LLM Prompting in GitLab
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.
Affected Software
Remediation
Recommended actions to resolve this vulnerability, in priority order.
- Upgrade
Upgrade
GitLab EEto a version that resolves this vulnerability.Fixed in 19.2.8 - Upgrade
Upgrade
GitLab EEto a version that resolves this vulnerability.Fixed in 19.3.5 - Upgrade
Upgrade
GitLab EEto a version that resolves this vulnerability.Fixed in 19.4.2
Event History
Frequently Asked Questions
Which GitLab EE versions need to be remediated?
Affected releases are GitLab EE 19.1 before 19.2.8, 19.3 before 19.3.5, and 19.4 before 19.4.2. Upgrading to 19.2.8, 19.3.5, or 19.4.2 addresses the issue for the corresponding release line.
What access and interaction are required for exploitation?
An attacker must be an authenticated GitLab user and must provide specially crafted content that is incorporated into AI agent instructions. The vector is network-accessible, has low attack complexity, and requires user interaction.
What is the potential impact?
Crafted AI conversation content could cause unintended command execution attributed to other users. The reported impact includes high confidentiality and integrity impact, with no availability impact stated.