CVE-2024-0520: MLflow command injection when loading dataset from HTTP source URL

Summary

CVE-2024-0520 affects mlflow/mlflow version 8.2.1. When loading a dataset from an HTTP source URL, the `mlflow.data.http_dataset_source.py` module builds the final file path from the Content-Disposition header or URL path without sanitization. Path traversal or absolute paths such as '../../tmp/poc.txt' let an attacker write arbitrary files, which could lead to command execution and access to data and model information.

Mitigation

The issue is fixed in version 2.9.0.

Affected packages in the Exposure Registry

Matched by package name and ecosystem. Each entry shows whether the package delegates to a language model and how many tracked packages depend on it.

• mlflowPyPILLM dependency since 2024-05-30 · 17 tracked dependents

Related items

• MediumCVE-2026-121936: easyVoice server-side request forgery through openaiBaseUrl in /api/v1/ttsSimilar attack · NVD/CVE Database
• MediumCVE-2026-118628: MLflow get-artifacts server-side request forgery information disclosureSimilar attack · NVD/CVE Database
• HighCVE-2026-118625: MLflow Tracking Server privilege escalation via get_endpointsSimilar attack · NVD/CVE Database
• InfoWhen trust becomes the payload: Inside the threats, etc quarterly reportSimilar attack · Elastic Security Labs
• MediumCritical Langflow flaws enable authentication bypass and RCESimilar attack · CSO Online