CVE-2025-14921
Deserialization RCE in Hugging Face Transformers Transformer-XL
Publication date: 2025-12-23
Last updated on: 2025-12-23
Assigner: Zero Day Initiative
Description
Description
CVSS Scores
EPSS Scores
| Probability: | |
| Percentile: |
Meta Information
Affected Vendors & Products
| Vendor | Product | Version / Range |
|---|---|---|
| hugging_face | transformers | 3.0 |
Helpful Resources
Exploitability
| CWE ID | Description |
|---|---|
| CWE-502 | The product deserializes untrusted data without sufficiently ensuring that the resulting data will be valid. |
Attack-Flow Graph
AI Powered Q&A
Can you explain this vulnerability to me?
This vulnerability is a remote code execution flaw in the Hugging Face Transformers library, specifically in the Transformer-XL model. It occurs because the model files are parsed without proper validation, allowing deserialization of untrusted data. An attacker can exploit this by tricking a user into visiting a malicious webpage or opening a malicious file, which then lets the attacker execute arbitrary code with the user's privileges. [1]
How can this vulnerability impact me? :
If exploited, this vulnerability allows an attacker to execute arbitrary code on your system with the same privileges as the current user. This can lead to full compromise of confidentiality, integrity, and availability of your data and system, potentially resulting in data theft, data corruption, or denial of service. [1]
How can this vulnerability be detected on my network or system? Can you suggest some commands?
Detection of this vulnerability involves identifying usage of the Hugging Face Transformers library, specifically the Transformer-XL model, and monitoring for attempts to load or parse untrusted model files. Since exploitation requires user interaction with malicious files or webpages, monitoring file access logs and network traffic for suspicious model files or URLs related to Transformer-XL models may help. However, no specific detection commands or signatures are provided in the available resources. [1]
What immediate steps should I take to mitigate this vulnerability?
Immediate mitigation steps include avoiding opening untrusted model files or visiting untrusted webpages that may contain malicious Transformer-XL models. Ensure that only trusted sources provide model files and consider applying any available patches or updates to the Hugging Face Transformers library once released. Additionally, restrict user privileges and monitor for suspicious activity related to model deserialization. [1]