PatchSiren

PatchSiren cyber security CVE debrief

CVE-2026-57845 huggingface CVE debrief

Hugging Face Datasets through 5.1.0 contains an insecure deserialization vulnerability allowing attackers to execute arbitrary Python code by crafting a malicious pickle file loaded through the deprecated pandas packaged builder. This vulnerability is triggered when a victim calls datasets.load_dataset('pandas', data_files=...), processing the pickle-backed dataset without validation, and executing the embedded __reduce__ payload with the victim process's privileges. The vulnerability allows for arbitrary code execution with the privileges of the victim process. Defenders should assess exposure and prioritize verification and remediation efforts.

Vendor
huggingface
Product
datasets
CVSS
HIGH 8.7
CISA KEV
Not listed in stored evidence
Original CVE published
2026-10-23
Original CVE updated
2026-10-29
Advisory published
2026-10-23
Advisory updated
2026-10-29

Who should care

Defenders responsible for maintaining and securing environments using Hugging Face Datasets should assess exposure and prioritize verification and remediation efforts. This includes operators, platform administrators, vulnerability management teams, and security teams that use Hugging Face Datasets.

Why it matters

CVE-2026-57845 is a high-severity vulnerability in Hugging Face Datasets that allows for arbitrary code execution. Defenders should prioritize verification, remediation, and monitoring efforts to prevent exploitation.

  • Potential for arbitrary code execution with victim process privileges
  • Need for verification of Hugging Face Datasets version and exposure in environments
  • Priority for updating to a secure version or implementing compensating controls
  • Importance of monitoring for suspicious activity related to Hugging Face Datasets

Technical summary

The Hugging Face Datasets library through version 5.1.0 contains an insecure deserialization vulnerability. This vulnerability allows attackers to execute arbitrary Python code by crafting a malicious pickle file loaded through the deprecated pandas packaged builder. The vulnerability is triggered when a victim calls datasets.load_dataset('pandas', data_files=...), which processes the pickle-backed dataset without validation, executing the embedded __reduce__ payload with the victim process's privileges.

Defensive priority

Defenders should prioritize verifying and updating Hugging Face Datasets to a secure version, assessing exposure in their environments, and implementing compensating controls to prevent exploitation.

Recommended defensive actions

  • Verify Hugging Face Datasets version and update to a secure version if necessary
  • Assess exposure in environments using Hugging Face Datasets
  • Implement compensating controls to prevent exploitation
  • Monitor for suspicious activity related to Hugging Face Datasets
  • Review relevant monitoring, detection, and logs for exposed assets that need extra review
  • Track exceptions, retest remediated assets, and close the item only after evidence is documented
  • Confirm whether affected product deployments exist in managed environments and assign an owner for follow-up

Evidence notes

The CVE record and source item provide details on the vulnerability, including its description, affected versions, and references to additional information. The vulnerability is confirmed to exist in Hugging Face Datasets through version 5.1.0. There are no known details on currently affected or unaffected versions beyond 5.1.0. Defenders should verify Hugging Face Datasets version and exposure in their environments.

Sources and references

Verified primary and authoritative sources

  • CVE-2026-57845 CVE Program record

    Publisher, destination, and source semantics verified

    URL: https://www.cve.org/CVERecord?id=CVE-2026-57845

    CVE Program - Official CVE Program record with source-provided CVE metadata.

  • CVE-2026-57845 NVD vulnerability detail

    Publisher, destination, and source semantics verified

    URL: https://nvd.nist.gov/vuln/detail/CVE-2026-57845

    NIST National Vulnerability Database - Official NIST NVD detail page and source-specific vulnerability assessment.

Supplemental references

  • Hugging Face Datasets through 5.1.0 Pickle Deserialization RCE via Pandas Builder

    Unverified legacy reference

    URL: https://raw.githubusercontent.com/CVEProject/cvelistV5/main/cves/2026/57xxx/CVE-2026-57845.json

    cve_program_cvelist_v5

  • Source reference

    Unverified legacy reference

    URL: https://github.com/SaiTeja-Erukude/CVE-2026-57845-hf-datasets-rce

    Supplemental source - technical-description, exploit

  • Source reference

    Unverified legacy reference

    URL: https://github.com/huggingface/datasets/pull/8879

    Supplemental source - patch

  • Source reference

    Unverified legacy reference

    URL: https://www.vulncheck.com/advisories/hugging-face-datasets-pickle-deserialization-rce-via-pandas-builder

    Supplemental source - third-party-advisory

Methodology and review provenance

AI-assisted synthesis based on stored public vulnerability evidence. System validation, approval state, and publication status do not by themselves establish human review of this revision. PatchSiren helps prioritize defensive review and does not prove exposure or remediation on any system.