CVE-2026-6859

Published: Apr 22, 2026Modified: Jul 15, 2026

8.8

Vector

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

Exploitability: 2.8 / Impact: 5.9

Source: NVD

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.

Affected (2)

Products: Redhat: Enterprise Linux Ai, Instructlab

2 products

Enterprise Linux Ai

Configuration A

2 vulnerable

Vulnerable Software	Affected Versions
Redhat Enterprise Linux Ai	Version 3.0
Redhat Instructlab	All versions

Related CWEs

Inclusion of Functionality from Untrusted Control Sphere

The product imports, requires, or includes executable functionality (such as a library) from a source that is outside of the intended control sphere.

References (5)

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

Source: secalert@redhat.com

Vendor Advisory

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

Source: secalert@redhat.com

Issue TrackingVendor Advisory

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

Source: 0b0ca135-0b70-47e7-9f44-1890c2a1c46c

Vendor Advisory

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

Source: 0b0ca135-0b70-47e7-9f44-1890c2a1c46c

Issue TrackingVendor Advisory

https://security.access.redhat.com/data/csaf/v2/vex/2026/cve-2026-6859.json

Source: 0b0ca135-0b70-47e7-9f44-1890c2a1c46c

Timeline

No history available yet.