AI System Card - RHEL Lightspeed

AI System Name RHEL command line assistant
Version 0.1
AI system developed by Red Hat, Inc.
Contact for security concerns security@redhat.com
Date of system card creation February 10, 2026
Publish date March 4, 2026

License: Creative Commons Attribution 4.0 International (CC BY 4.0)

Purpose

The RHEL command line assistant provides an optional AI enabled assistant which is available from the RHEL command line. It uses a natural language interface to answer RHEL-related questions, help with troubleshooting, and can be used as a learning aid. The command line assistant can provide a productivity boost for RHEL users ranging from novices to experts.

Technical Information

AI model Gemini 2.5 Flash
Hosting platform Google Vertex AI
Development stack python, postgres
Inference mechanism Google Vertex AI
Languages English

Data Provenance and Pedigree

The RHEL command line assistant hosted service utilizes the Gemini 2.5 Flash model, hosted on the Google Vertex AI platform. Other components of the infrastructure are hosted on Red Hat infrastructure.

RHEL Lightspeed uses a RAG (retrieval-augmented generation) database that contains Red Hat documentation and errata, knowledge base articles and solutions, as well as data about CVEs. The RHEL Lightspeed system retrieves relevant information from the RAG database to enhance the LLM’s knowledge of Red Hat products and processes.

Security and Safety of the AI System

Intent and use:

The RHEL command line assistant is intended to be used to help answer RHEL-related questions, help with troubleshooting on RHEL systems, and can be used as a learning aid to help improve RHEL related skills. Users should always review AI-generated content prior to use.

Scope:

Users provide a prompt to the command line assistant, and can also attach files, or pipe output from other commands to the command line assistant. This content is processed by the command line assistant backend (Red Hat infrastructure and Google Vertex API), and a result is returned to the user. Users can also optionally enable a terminal capture mode that enables users to reference output from commands previously run.

Do not include any personal information or other sensitive information in your input. Interactions may be used to improve Red Hat's products or services. Always review AI-generated content prior to use.

Security considerations:

Security testing in accordance with Red Hat Secure Software Development Lifecycle was conducted on the codebase before it was released.

Governance

Any security or safety issues related to the AI system should be reported to security@redhat.com.

References