Browse / AI Framework and Platform Flaws / CVE-2025-3264
Transformers vulnerable to ReDoS attack through its get_imports() function
- Class
- AI Framework and Platform Flaws (filed by a model that read the advisory)
- Technique
- Denial of service
- Affected
- transformers (PyPI)
- CVSS
- CVSS:3.0/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:L
- EPSS
- 0.5% chance of exploitation in 30 days
How it works, in short
A catastrophic backtracking regex in the import-scanning helper makes crafted Python source consume excessive CPU when model code is loaded. (Our summary, written by a model. The source text is below.)
Details
A Regular Expression Denial of Service (ReDoS) vulnerability was discovered in the Hugging Face Transformers library, specifically in the get_imports() function within dynamic_module_utils.py. This vulnerability affects versions 4.49.0 and is fixed in version 4.51.0. The issue arises from a regular expression pattern \s*try\s*:.*?except.*?: used to filter out try/except blocks from Python code, which can be exploited to cause excessive CPU consumption through crafted input strings due to catastrophic backtracking. This vulnerability can lead to remote code loading disruption, resource exhaustion in model serving, supply chain attack vectors, and development pipeline disruption.
References
- https://nvd.nist.gov/vuln/detail/CVE-2025-3264
- https://github.com/huggingface/transformers/commit/0720e206c6ba28887e4d60ef60a6a089f6c1cc76
- https://github.com/huggingface/transformers/commit/126abe3461762e5fc180e7e614391d1b4ab051ca
- https://github.com/huggingface/transformers
- https://huntr.com/bounties/3c6f7822-9992-476d-8cf0-b0b1623427df
Source: GitHub Advisory Database, used under CC BY 4.0.