CVE-2026-96804

Answer in brief

CVE-2026-96804 records a High severity (CVSS 8.8) vulnerability in CVE-2026-96804. The current sources do not mark it as known exploited. The current feed maps MLflow/MLflow (generic). Check affected ranges and fixed versions before updating.

Analysis pending evidence review

HOL Guard separates source facts from reviewed analysis. See the methodology.

Published Sep 23, 2026Updated Sep 23, 2026Source checked Sep 23, 2026First seen by HOL Sep 23, 2026Material review Sep 23, 2026

Upstream Advisory

Key facts

Risk: High · CVSS 8.8
Exploitation: Not marked as known exploited
Affected software: 1 mapped package or product
Fix availability: Available

Why this deserves its current priority

CVSS is 8.8. The current sources do not mark it as known exploited. Treat this as a source-backed prioritization signal, not a statement about your environment.

Analysis status

Analysis pending evidence review

Factual feed record only; HOL analysis is not approved for indexing. Read the methodology.

Affected scope and exposure questions

The current feed maps MLflow/MLflow (generic). Check affected ranges and fixed versions before updating.

Mapped affected packages and fixed versions
Package	Affected range	Fixed version
MLflow/MLflowgeneric	>=2.1 <3.15.0	3.15.0

Recommended response

Check inventory. Check lockfiles and deployed manifests for MLflow/MLflow.
Review the reported fix. Update MLflow/MLflow to 3.15.0 if you use the affected versions. Test the change in a non-production environment first.

Evidence timeline and material changes

Published upstream
Sep 23, 2026
Evidence: source:cvelist:source_dates:source-dates:record
Source modified
Sep 23, 2026
Evidence: source:cvelist:source_dates:source-dates:record
First seen by HOL
Sep 23, 2026

Sources and claim methodology

Source referencekb.cert.org

Upstream source description

MLflow's statsmodel flavor, versions 2.1.0 to 3.14.0, omits the MLFLOW_ALLOW_PICKLE_DESERIALIZATION=False security control entirely in _load_model(), which allows a remote attacker to execute arbitrary code via a crafted MLmodel artifact.

Quoted source text, attributed separately from HOL analysis.