SB2026072818 - Multiple vulnerabilities in MLflow

Description

This security bulletin contains information about 2 vulnerabilities.

1) Deserialization of Untrusted Data (CVE-ID: N/A)

CWE-ID: CWE-502 - Deserialization of Untrusted Data

CVSSv4: 8.6 [CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:A/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N]

The vulnerability allows a remote attacker to execute arbitrary code on the target system.

The vulnerability exists due to insecure input validation when processing serialized data in the pyfunc flavor. A remote attacker can pass specially crafted data to the application and execute arbitrary code on the target system.

Successful exploitation of this vulnerability may result in complete compromise of vulnerable system.

2) Missing Authorization (CVE-ID: CVE-2026-69148)

CWE-ID: CWE-862 - Missing Authorization

CVSSv4: 7.1 [CVSS:4.0/AV:N/AC:L/AT:N/PR:L/UI:N/VC:H/VI:N/VA:N/SC:N/SI:N/SA:N]

The vulnerability allows an authenticated remote attacker to read other users' private artifact files in MLflow deployments using built-in basic authentication and restricted experiment permissions.

The vulnerability exists due to missing READ authorization checks on source runs or logged models when creating model versions. An authenticated remote attacker with MANAGE permission on a registered model can reference a victim's artifact directory in a new version and download its files through the model version artifact endpoint.

Remediation

Install update from vendor's website.

References

• https://github.com/mlflow/mlflow/security/advisories/GHSA-gqvg-gmmx-x4hm
• https://github.com/mlflow/mlflow/security/advisories/GHSA-gqch-g4w5-7qcw