Security readout for executives and security teams
Plain-English summary
This flaw can let a malicious HuggingFace model run Python code when a user uses affected InstructLab training, download, or generation workflows. It requires user interaction, but successful exploitation could compromise the system running the workflow.
Executive priority
Treat this as high priority for AI engineering and platform teams using RHEL AI 3. The risk is not internet-wormable, but trusted-user model workflows can become a full host compromise path.
Technical view
InstructLab’s linux_train.py hardcodes trust_remote_code=True when loading HuggingFace models. A crafted model can therefore execute attacker-controlled Python if a user runs ilab train, download, or generate against it. Red Hat maps this to CWE-829 and CVSS 8.8.
Likely exposure
Exposure is most likely in Red Hat Enterprise Linux AI 3 environments using the listed CUDA bootc packages. The source marks ROCm packages as unaffected. No vulnerable version ranges are provided.
Exploitation context
The provided sources do not report known active exploitation, and KEV is false. Exploitation requires convincing a user to invoke affected ilab workflows with a malicious HuggingFace Hub model.
Researcher notes
Evidence is specific to InstructLab behavior in Red Hat Enterprise Linux AI 3 package contexts. Sources do not provide patch versions, exploit samples, or confirmed exploitation. Avoid broad claims beyond the listed packages.
Mitigation direction
- Check Red Hat CVE, Bugzilla, and VEX guidance for fixes or updated package status.
- Do not run ilab workflows against untrusted HuggingFace models.
- Restrict model sources to approved repositories and reviewed model revisions.
- Isolate training and generation workloads from sensitive hosts and credentials.
- Prioritize updates for affected RHEL AI 3 CUDA bootc images when available.
Validation and detection
- Inventory RHEL AI 3 systems and bootc image package names.
- Confirm whether CUDA bootc packages listed as affected are deployed.
- Verify ROCm bootc packages are treated according to the Red Hat unaffected status.
- Review ilab train, download, and generate usage with HuggingFace Hub models.
- Check Red Hat advisories for fixed builds or revised VEX status.
Public sources used
Generated from the cited source records. This long-tail analysis has not been individually reviewed by a named human.
Conservative CVE-to-ATT&CK context
These mappings and lookup hints may be relevant to the vulnerability behavior, CWE, affected product, or exposure path. Glexia-inferred context is not an official MITRE, ATT&CK, CWE, or CVE Program mapping.
ATT&CK lookup starting points
Use these exact CWE pages and searches to review the Glexia ATT&CK library from this CVE's weakness and description context.
CWE-829: Exact CWE lookup
Use the exact CWE identifier as the starting point before reviewing related ATT&CK behavior. Open the exact CWE lookup page first, then review the ATT&CK searches from that MITRE weakness context. This is a Glexia lookup hint, not an official ATT&CK mapping.
Open ATT&CK lookupExecution behavior lookup
The CVE wording references code or command execution, so execution technique review may help defensive triage. This is a Glexia inferred lookup path, not an official MITRE, ATT&CK, or CVE Program mapping.
Open ATT&CK lookupCVE-2026-6859 mapping review
Open the CVE-to-ATT&CK bridge for reviewed, inferred, or future official mappings tied to this CVE.
Open ATT&CK lookup- Severity
- High
- CVSS
- 8.8 (3.1)
- Known Exploited
- No
- Published
Vector: CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H
CNA and ADP enrichment extracted from CVE v5
These fields come from the CVE record and ADP containers, not from Glexia's Take. They preserve time-varying source decisions such as CISA SSVC, KEV status, CVSS metrics, and provider references.
SSVC decision data
CVSS vector scores
2 official scoresWe collect every scored CVSS vector available in the official CNA and ADP containers. When more than one version is present, the table keeps the source vectors side by side instead of collapsing them into the highest score.
CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H2.85.9redhatCVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H2.85.9redhat-SADPVulnerability scoring details
Base CVSS 3.1 score
8.8HighVector: CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H
Base CVSS 3.1 score
8.8HighVector: CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H
Vulnerability timeline
Timeline events are normalized from CVE metadata, CNA source timelines, ADP timelines, and KEV metadata when present.
- ADP timelineredhat-SADP
Reported to Red Hat.
- ADP timelineredhat-SADP
Made public.
- Source timelineredhat
Reported to Red Hat.
- Source timelineredhat
Made public.
- CVE reservedCVE Program
The CVE ID was reserved by the assigning CNA.
- CVE publishedCVE Program
The CVE record was published.
- CVE updatedCVE Program
The CVE record metadata indicates this as the latest update time.
ADP provider summaries
- 2026-04-15T00:00:00.000Z: Reported to Red Hat.
- 2026-04-15T00:00:00.000Z: Made public.
Source materials
- CVE List V5 sourceCVE List V5
- https://access.redhat.com/security/cve/CVE-2026-6859CVE reference, redhat · vdb-entry, x_refsource_REDHAT
- RHBZ#2459998CVE reference, redhat · issue-tracking, x_refsource_REDHAT
- https://security.access.redhat.com/data/csaf/v2/vex/2026/cve-2026-6859.jsonCVE reference, redhat-SADP · x_sadp-csaf-vex
Products and packages named in the record
CWE details
CWE links open Glexia weakness intelligence pages with official CWE context, developer remediation guidance, and related CVE mappings.
Inclusion of Functionality from Untrusted Control Sphere
Inclusion of Functionality from Untrusted Control Sphere represents a recurring weakness pattern that can create exploitable paths when design, validation, or implementation controls are missing.