Skip to main content
← All vulnerabilities
CVE-2026-57845EUVD-2026-109092GHSA-9HC7-P88P-MWM8

Hugging Face Datasets 5.1.0 Pickle Deserialization RCE via Pandas Builder

Published
Last modified
Status
published
Known exploitation
Not listed in CISA KEV
Fix
Fix or patch referenced
Public exploit
None on Exploit-DB

Summary

The Hugging Face Datasets library (through version 5.1.0 as stated, but without version numbers here) is affected when its deprecated pandas builder loads pickle files without validation. An attacker who supplies a malicious pickle dataset could get arbitrary Python code to run with the privileges of any user or process that loads it.

Summarized by anthropic/claude-haiku-5.5. Verify details against the original sources.

Read the original description

Hugging Face Datasets through 5.1.0 contains an insecure deserialization vulnerability that allows attackers to execute arbitrary Python code by crafting a malicious pickle file loaded through the deprecated pandas packaged builder. Attackers can supply a pickle-backed dataset that is processed by Pandas._generate_tables via pandas.read_pickle() without validation, so when a victim calls datasets.load_dataset("pandas", data_files=...), the embedded __reduce__ payload executes code with the victim process's privileges.

Affected versions

  • huggingface datasets

    through 5.1.0

References 3

GitHub · SaiTeja-Erukude/CVE-2026-57845-hf-datasets-rce

GitHub · huggingface/datasets

vulncheck.com