CVE-2024-11393
Published: 22 November 2024
Summary
CVE-2024-11393 is a high-severity Deserialization of Untrusted Data (CWE-502) vulnerability in Huggingface Transformers. Its CVSS base score is 8.8 (High).
Operationally, exploitation aligns with the MITRE ATT&CK technique Exploitation for Client Execution (T1203); ranked in the top 0.9% of CVEs by exploit likelihood; it is not currently listed in the CISA KEV catalog.
This vulnerability is AI-related — categorised as NLP and Transformers; in the Supply Chain and Deployment risk domain; MITRE ATLAS techniques in scope: AI Supply Chain Compromise (AML.T0010).
Deeper analysis
CVE-2024-11393 is a deserialization of untrusted data vulnerability, tracked as CWE-502, that affects Hugging Face Transformers when parsing MaskFormer model files. The flaw stems from insufficient validation of user-supplied data during model file processing and carries a CVSS 3.1 score of 8.8.
Remote attackers can exploit the issue without authentication by supplying a malicious model file or page; successful exploitation results in arbitrary code execution in the context of the current user, though user interaction is required to trigger the payload.
The single referenced advisory from the Zero Day Initiative (ZDI-24-1514) confirms the vulnerability was originally reported as ZDI-CAN-25191 but does not detail specific patches or workarounds in the supplied information.
The vulnerability impacts an AI/ML framework component, and its EPSS score has remained steady at a peak of 0.7953 since disclosure.
EU & UK References
- 🇪🇺 ENISA EUVD: EUVD-2024-34102
Vulnerability details
Hugging Face Transformers MaskFormer Model Deserialization of Untrusted Data Remote Code Execution Vulnerability. This vulnerability allows remote attackers to execute arbitrary code on affected installations of Hugging Face Transformers. User interaction is required to exploit this vulnerability in that the…
more
target must visit a malicious page or open a malicious file. The specific flaw exists within the parsing of model files. The issue results from the lack of proper validation of user-supplied data, which can result in deserialization of untrusted data. An attacker can leverage this vulnerability to execute code in the context of the current user. Was ZDI-CAN-25191.
- CWE(s)
AI Security AnalysisAI
- AI Category
- NLP and Transformers
- Risk Domain
- Supply Chain and Deployment
- OWASP Top 10 for LLMs 2025
- None mapped
- Classification Reason
- The vulnerability affects Hugging Face Transformers library, specifically in handling the MaskFormer model during deserialization of model files. Hugging Face Transformers is a core library for transformer-based models, primarily associated with NLP and transformer architectures, even though MaskFormer is used for computer vision tasks.
Related Threats
MITRE ATT&CK Enterprise TechniquesAI
Why these techniques?
The deserialization vulnerability in Hugging Face Transformers MaskFormer model parsing allows remote code execution via exploitation of client software when loading untrusted model files or visiting a malicious page, directly mapping to T1203: Exploitation for Client Execution.
MITRE ATLAS TechniquesAI
MITRE ATLAS techniques
Affected Assets
Mitigating Controls
Likely Mitigating Controls AI
Per-CVE control mapping for this CVE has not run yet; the list below is derived from the weakness types (CWEs) cited in the NVD entry.
Penetration testing supplies malicious serialized objects, detecting unsafe deserialization and supporting corrective actions.
Evaluation of untrusted data handling (deserialization testing) reveals unsafe processing, which the required remediation process addresses.
Untrusted serialized data can be deserialized and observed inside the chamber, blocking gadget-chain exploitation outside the sandbox.
Validates or rejects untrusted serialized data before deserialization occurs.
Identifies and blocks malicious code introduced through deserialization of untrusted data at system boundaries.
Integrity verification of serialized information can detect tampering before deserialization occurs.
Provenance of associated data allows detection of untrusted sources before deserialization or processing occurs.