CyberSecBulletin
CVE-2026-5241NVD status: Modified

huggingface/transformers: Inclusion of Functionality from Untrusted Control Sphere

Published Updated By the CyberSecBulletin desk

CVE-2026-5241 is a critical-severity inclusion of functionality from untrusted control sphere vulnerability in huggingface/transformers from Huggingface, with a CVSS 3.1 base score of 9.6 out of 10. It can be exploited remotely over the network by an unauthenticated attacker, but only if a user takes part, for example by opening a link or file. The record lists version 5.5.0 as the first without the flaw, so updating huggingface/transformers to that release or later removes the exposure. There is no confirmed exploitation so far. EPSS estimates a 0.6% chance of exploitation within 30 days, which is low compared with other CVEs.

Patch this week

Critical severity: a successful attack would have severe impact.

What to do

  1. Update huggingface/transformers to 5.5.0 or a later version.
  2. Until you can update, restrict network access to the affected service so it is not reachable from untrusted networks.
  3. Because the attack needs a user to take part, warn users about unexpected links and files related to this product until it is patched.

Affected products and versions

VendorProductAffected versionsFixed in
Huggingfacehuggingface/transformersbefore 5.5.05.5.0
Red HatRed Hat OpenShift AI 2.25Not specified in the recordNot stated
Red HatRed Hat OpenShift AI 3.3Not specified in the recordNot stated
Red HatRed Hat OpenShift AI 3.4Not specified in the recordNot stated
Red HatOpenShift LightspeedNot specified in the recordNot stated
Red HatRed Hat AI Inference ServerNot specified in the recordNot stated
Red HatRed Hat Ansible Automation Platform 2Not specified in the recordNot stated
Red HatRed Hat Enterprise Linux AI (RHEL AI) 3Not specified in the recordNot stated
Red HatRed Hat OpenShift AI (RHOAI)Not specified in the recordNot stated
HuggingfaceTransformers5.2.0Not stated

Version data comes from the organisation that assigned the CVE. Confirm against the vendor advisory before deciding you are unaffected.

How the severity score breaks down

Attack vector
NetworkCan be exploited remotely over a network, including the internet.
Attack complexity
LowNo special conditions are needed; the attack works reliably.
Privileges required
NoneNo account or login is needed.
User interaction
RequiredA user must take an action such as opening a file or link.
Scope
ChangedThe attack can reach beyond the vulnerable component into other systems.
Confidentiality
HighAll data handled by the component can be read.
Integrity
HighData or code can be modified at will.
Availability
HighThe component can be taken down completely.

Vector: CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:C/C:H/I:H/A:H. Scored by NVD.

Scoring systemScoreSeverityScored by
CVSS 3.19.6CriticalNVD
CVSS 3.17.7HighCNA
CVSS 3.08.0HighCNA (huntr.dev)

Official description

A vulnerability in the LightGlue model loading path of huggingface/transformers version 5.2.0 allows an attacker-controlled model repository to execute arbitrary code during model initialization. The issue arises because the `trust_remote_code` parameter, intended to prevent remote code execution, is overridden by untrusted serialized configuration data in a nested code path. Specifically, when loading a LightGlue model using `AutoModel.from_pretrained()` with `trust_remote_code=False`, the `LightGlueConfig` reads the `trust_remote_code` value from the untrusted `config.json` file and propagates it into nested `AutoConfig.from_pretrained()` calls. This results in the execution of attacker-provided Python modules, even when the victim explicitly disables remote code execution. The vulnerability poses a high risk for environments such as API inference servers, research notebooks, CI/CD pipelines, and model evaluation workers, potentially leading to credential theft, lateral movement, or persistence/backdoor deployment.

Source: the CVE record as published by its numbering authority, via NVD and CVE.org.

The underlying weakness: Inclusion of Functionality from Untrusted Control Sphere

The product imports, requires, or includes executable functionality (such as a library) from a source that is outside of the intended control sphere.

How developers prevent this class of flaw

  • Use a vetted library or framework that does not allow this weakness to occur or provides constructs that make this weakness easier to avoid.
  • When the set of acceptable objects, such as filenames or URLs, is limited or known, create a mapping from a set of fixed input values (such as numeric IDs) to the actual filenames or URLs, and reject all other inputs. For example, ID 1 could map to inbox.txt and ID 2 could map to profile.txt. Features such as the ESAPI AccessReferenceMap provide this capability.
  • For any security checks that are performed on the client side, ensure that these checks are duplicated on the server side, in order to avoid CWE-602. Attackers can bypass the client-side checks by modifying values after the checks have been performed, or by changing the client to remove the client-side checks entirely. Then, these modified values would be submitted to the server.

More Inclusion of Functionality from Untrusted Control Sphere vulnerabilities (CWE-829)

Timeline

  1. CVE record published.
  2. CISA published its SSVC assessment (exploitation: poc, automatable: no, technical impact: total).
  3. Record last changed materially (score, affected products, status or references).

References

Technical write-ups (may include proof of concept)

Other vulnerabilities in huggingface/transformers

All huggingface/transformers vulnerabilities

Recent Inclusion of Functionality from Untrusted Control Sphere vulnerabilities

Spotted an error? Tell us.