CVE-2026-6859

HIGH

Instructlab: instructlab: arbitrary code execution due to hardcoded `trust_remote_code=true`

Title source: cna

Description

A flaw was found in InstructLab. The `linux_train.py` script hardcodes `trust_remote_code=True` when loading models from HuggingFace. This allows a remote attacker to achieve arbitrary Python code execution by convincing a user to run `ilab train/download/generate` with a specially crafted malicious model from the HuggingFace Hub. This vulnerability can lead to complete system compromise.

References (2)

Core 2

Core References

Vdb Entry, X_Refsource_Redhat vdb-entry x_refsource_redhat

https://access.redhat.com/security/cve/CVE-2026-6859

Issue Tracking, X_Refsource_Redhat issue-tracking x_refsource_redhat

RHBZ#2459998

https://bugzilla.redhat.com/show_bug.cgi?id=2459998

Scores

CVSS v3 8.8

EPSS 0.0035

EPSS Percentile 27.0%

Attack Vector NETWORK

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

CISA SSVC

Vulnrichment

Exploitation none

Automatable no

Technical Impact total

Details

CWE

CWE-829

Status published

Products (4)

pypi/instructlab 0PyPI

Red Hat/Red Hat Enterprise Linux AI (RHEL AI) 3

redhat/enterprise_linux_ai 3.0

redhat/instructlab

Published Apr 22, 2026

Tracked Since Apr 22, 2026