Skip to content
Fencer Vulns
high 8.8

CVE-2026-96775 json

MLflow dspy bypasses pickle deserialization control

published 2026-09-23 47 d ago modified 2026-09-23 46 d ago also EUVD-2026-85410

cvss
8.8
version 3.1, CNA
cisa kev
Not listed
not in CISA's catalog
epss
0.6%
EPSS 0.6% chance of exploitation in 30 days, percentile 51%
public exploits
0
none known

Summary

MLflow's dspy flavor, versions >= 2.0, applies the MLFLOW_ALLOW_PICKLE_DESERIALIZATION=False security control only when the model_path ends in .pkl, which allows a remote attacker to execute arbitrary code via a crafted MLmodel artifact.

Vendor advisories

What the vendors themselves say about their products: affected, fixed or not affected, and what to do.

red hat product security vex CVE-2026-96775 updated cvss 8.8

mlflow: MLflow dspy: Arbitrary code execution via crafted MLmodel artifact

  • Red Hat OpenShift AI (RHOAI) rhoai/odh-mlflow-rhel9
    affected: no fix available
  • Red Hat OpenShift AI 3.4 rhoai/odh-th06-cpu-torch210-py312-rhel9 1791874335, rhoai/odh-th06-rocm64-torch291-py312-rhel9 1791872119, rhoai/odh-th06-cuda130-torch210-py312-rhel9 1791961113, rhoai/odh-training-cuda128-torch29-py312-rhel9 1791786619, rhoai/odh-pipeline-runtime-pytorch-cuda-py312-rhel9 1791829182, rhoai/odh-pipeline-runtime-pytorch-rocm-py312-rhel9 1791868244, +12 more
    fixed: vendor fix RHSA-2026:84734
  • Red Hat OpenShift AI 3.5 rhoai/odh-pipeline-runtime-pytorch-cuda-py312-rhel9 1791816462, rhoai/odh-workbench-jupyter-pytorch-cuda-py312-rhel9 1791817064, rhoai/odh-workbench-jupyter-pytorch-rocm-py312-rhel9 1791816223, rhoai/odh-workbench-jupyter-trustyai-cpu-py312-rhel9 1791816316, rhoai/odh-pipeline-runtime-datascience-cpu-py312-rhel9 1791952873, rhoai/odh-pipeline-runtime-tensorflow-cuda-py312-rhel9 1791951047, +6 more
    fixed: vendor fix RHSA-2026:86891
  • Red Hat OpenShift AI (RHOAI) rhoai/odh-th-torch-cpu-py312-rhel9, rhoai/odh-th-torch-cuda-py312-rhel9, rhoai/odh-th-torch-rocm-py312-rhel9
    not affected: vulnerable code not in execute path
  • Red Hat OpenShift AI 3.4 rhoai/odh-cli-rhel9 1791509689, rhoai/rhai-cli-rhel9 1791509689, rhoai/odh-automl-rhel9 1791535789, rhoai/odh-mlflow-rhel9 1791865967, rhoai/odh-autorag-rhel9 1791557539, rhoai/odh-trainer-rhel9 1791869783, +82 more Red Hat OpenShift AI 3.5 rhoai/odh-cli-rhel9 1791945905, rhoai/rhai-cli-rhel9 1791945905, rhoai/odh-automl-rhel9 1791916705, rhoai/odh-mlflow-rhel9 1791947857, rhoai/odh-autorag-rhel9 1791903266, rhoai/odh-trainer-rhel9 1791952138, +101 more
    not affected: vulnerable code not present

Red Hat CSAF VEX © Red Hat, Inc., from https://security.access.redhat.com/data/csaf/v2/vex/; changed: reduced to product statuses and remediations. CC BY 4.0 · csaf

CISA SSVC

Exploitation
none
Automatable
no
Technical impact
total
Assessed
46 d ago

Exploit prediction

0.65% chance of exploitation in the next 30 days, higher than 51% of all scored CVEs (FIRST EPSS, 2026-11-08).

CVSS scores

assigned by version score vector
CISA ADP 3.1 8.8 CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H
CNA 3.1 8.8 CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H

Affected products

  • MLflow MLflow

Classification

Weakness
  • CWE-502 Deserialization of Untrusted Data

References

Sources

Fencer Vulns keeps one entry per vulnerability: the CVE id wins, then GHSA, and records that alias each other are merged, except that two CVEs always keep separate entries. The same data is available as JSON.