CVE-2026-96893

published 13 Oct 2026 · updated 14 Oct 2026

Improper Neutralization of Input Used for LLM Prompting in GitLab

7.3HighCVSS 7.3EPSS 1%

No exploitation in the wild has been reported. EPSS puts the chance of exploitation in the next 30 days at 1%. Fixed in GitLab 19.2.8, GitLab 19.3.5 and GitLab 19.4.2.

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

Vendors, products and packages with version ranges

VendorProductVersionsFixed
GitLabGitLab>=19.1 <19.2.8, >=19.3 <19.3.5, >=19.4 <19.4.219.2.8, 19.3.5, 19.4.2

Exploit evidence

1 public sources

Vendor advisories

Timeline

Every event on this vulnerability

Related

Other vulnerabilities in GitLab, strongest first

10.0CriticalPublic exploit

Embed

A live severity badge for an advisory, README or ticket

CVE-2026-96893 severity on Zeptric[![CVE-2026-96893](https://zeptric.io/badge/cve/CVE-2026-96893.svg)](https://zeptric.io/cve/CVE-2026-96893)