A critical remote code execution vulnerability exists in all versions of the HuggingFace transformers library prior to version 5.3.0. The vulnerability allows an attacker to craft a malicious `config.json` file containing the `_attn_implementation_internal` field set to an attacker-controlled HuggingFace Hub repository ID. When a victim loads this model using the standard `AutoModelForCausalLM.from_pretrained()` API, the library downloads and executes arbitrary Python code from the attacker's repository with the victim's full OS privileges. This issue arises due to unfiltered deserialization of configuration attributes, insufficient sanitization of internal fields, and unsandboxed execution of downloaded kernels. The vulnerability bypasses the `trust_remote_code` security mechanism, is invisible to the victim, and exploits the standard documented usage pattern, making it particularly severe. Users are advised to upgrade to version 5.3.0 or later to mitigate this issue.

Project Subscriptions

Vendors Products
Huggingface Subscribe
Transformers Subscribe
Advisories

No advisories yet.

Fixes

Solution

No solution given by the vendor.


Workaround

No workaround given by the vendor.

History

Tue, 26 May 2026 16:30:00 +0000

Type Values Removed Values Added
Metrics ssvc

{'options': {'Automatable': 'no', 'Exploitation': 'none', 'Technical Impact': 'total'}, 'version': '2.0.3'}


Mon, 25 May 2026 11:45:00 +0000

Type Values Removed Values Added
First Time appeared Huggingface
Huggingface transformers
Vendors & Products Huggingface
Huggingface transformers

Sun, 24 May 2026 14:00:00 +0000

Type Values Removed Values Added
Description A critical remote code execution vulnerability exists in all versions of the HuggingFace transformers library prior to version 5.3.0. The vulnerability allows an attacker to craft a malicious `config.json` file containing the `_attn_implementation_internal` field set to an attacker-controlled HuggingFace Hub repository ID. When a victim loads this model using the standard `AutoModelForCausalLM.from_pretrained()` API, the library downloads and executes arbitrary Python code from the attacker's repository with the victim's full OS privileges. This issue arises due to unfiltered deserialization of configuration attributes, insufficient sanitization of internal fields, and unsandboxed execution of downloaded kernels. The vulnerability bypasses the `trust_remote_code` security mechanism, is invisible to the victim, and exploits the standard documented usage pattern, making it particularly severe. Users are advised to upgrade to version 5.3.0 or later to mitigate this issue.
Title Arbitrary Remote Code Execution via `_attn_implementation_internal` Config Injection in huggingface/transformers
Weaknesses CWE-1066
References
Metrics cvssV3_0

{'score': 7.8, 'vector': 'CVSS:3.0/AV:L/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H'}


Projects

Sign in to view the affected projects.

cve-icon MITRE

Status: PUBLISHED

Assigner: @huntr_ai

Published:

Updated: 2026-05-26T15:21:02.376Z

Reserved: 2026-03-18T10:04:09.683Z

Link: CVE-2026-4372

cve-icon Vulnrichment

Updated: 2026-05-26T15:20:57.850Z

cve-icon NVD

Status : Awaiting Analysis

Published: 2026-05-24T14:16:16.917

Modified: 2026-05-26T20:06:20.310

Link: CVE-2026-4372

cve-icon Redhat

No data.

cve-icon OpenCVE Enrichment

Updated: 2026-05-25T11:30:23Z

Weaknesses