Skip to content
Submit incident
AII-1002-52861
ConfirmedA court, a regulator or the organization itself has stated the core facts, or the record cites a court or regulator source.

Hugging Face Transformers MaskFormer Model Deserialization RCE Vulnerability CVE-2024-11393

A remote code execution vulnerability was discovered in Hugging Face Transformers' MaskFormer model parser that allows attackers to execute arbitrary code through deserialization of untrusted model files. The vulnerability requires user interaction and was disclosed as a zero-day advisory after the vendor rejected the initial report.

11 September 2024 Β· Hugging Face Β· Security and cyberattacks
11 September 2024
AII-1002-528612 sources ↓
Quick facts
Category
Security and cyberattacks
Source
cve.org
Date
11 September 2024
Drafted by
Raidu Incident Detection

What happened

A vulnerability was discovered in the parsing of model files within Hugging Face Transformers that results from the lack of proper validation of user-supplied data, allowing deserialization of untrusted data.

The flaw enables remote attackers to execute arbitrary code on affected installations, with a CVSS score of 8.8. User interaction is required, as the target must visit a malicious page or open a malicious file.

Reported impact

An attacker can leverage this vulnerability to execute code in the context of the current user.

Index analysis

The vulnerability was reported to Hugging Face via a bug bounty platform on September 11, 2024, but the vendor rejected it on October 14, 2024.

The Zero Day Initiative notified the vendor of intention to publish the case as a zero-day advisory on November 6, 2024, and coordinated public release occurred on November 19, 2024.

Governance controls

Given the nature of the vulnerability, the only salient mitigation strategy is to restrict interaction with the application.

Classification

Category
Security and cyberattacks
Industry
IT infrastructure and hosting
Harm status
Potential harm
Intent
Unintentional
Timing
Post-deployment
Harmed
Consumers
AI systems
Transformers MaskFormer Model

Sources and evidence

This record was researched and written by the Index, based on the following original reporting.

cve.org
NVD entry
zerodayinitiative.com
https://www.zerodayinitiative.com/advisories/ZDI-24-1514/