Transformers Deserialization of Untrusted Data vulnerability
The huggingface/transformers library is vulnerable to arbitrary code execution through deserialization of untrusted data within the `load_repo_checkpoint()` function of the `TFPreTrainedModel()` class. Attackers can execute arbitrary code and commands by crafting a malicious serialized payload, exploiting the use of `pickle.load()` on data from potentially untrusted sources. This vulnerability allows for remote code execution (RCE) by deceiving victims into loading a seemingly harmless checkpoint during a normal training process, thereby enabling attackers to execute arbitrary code on the targeted machine.
describes (1)
Related CVEs.
| ID | severity | exploitable | ||
|---|---|---|---|---|
| CVE-2024-3568US | critical9.6 | Exploit | ||
The huggingface/transformers library is vulnerable to arbitrary code execution through deserialization of untrusted data within the `load_repo_checkpoint()` function of the `TFPreT… | ||||
references
External citations — patches, advisories, analysis.
- https://github.com/advisories/GHSA-37q5-v5qm-c9v8ADVISORY
- https://github.com/huggingface/transformersPACKAGE
- https://github.com/huggingface/transformers/commit/693667b8ac8138b83f8adb6522ddaf42fa07c125WEB
- https://huntr.com/bounties/b3c36992-5264-4d7f-9906-a996efafba8fWEB
- https://nvd.nist.gov/vuln/detail/CVE-2024-3568ADVISORY
- https://pypi.org/project/transformersPACKAGE
Advisories like this name the exposed products attackers go after on your perimeter. See what is exposed on your perimeter with a free passive scan.
Free scan