Vulnerabilities in pkg:pypi/transformers
PACKAGECVE-2026-80047 is a HIGH (CVSS 7.8) hugging face transformers downloads custom vulnerability (CWE-273: Improper Check for Dropped Privileges) affecting Transformers libraries in the PyPI ecosystem. The issue involves hugging face transformers downloads custom generation code before trust consent. Remediation is available in the latest vendor maintenance releases.
pkg:pypi/transformers
Vulnerability advisories affecting package pkg:pypi/transformers. Select an advisory below to view technical details or fixes.
| Advisory ID | CVE / NVD | Severity | Summary |
|---|---|---|---|
| GHSA-x9r9-c232-4q39 | CVE-2026-80047 | HIGH | Hugging Face Transformers downloads custom generation code before trust consent |
| GHSA-xrqw-3rrv-vx5w | CVE-2026-9856 | HIGH | Transformers save_pretrained path traversal allows arbitrary file writes through chat template names |
| GHSA-fgcw-684q-jj6r | CVE-2026-5241 | HIGH | huggingface/transformers: Arbitrary Code Execution During Model Initialization in the LightGlue Model Loading Path |
| GHSA-29pf-2h5f-8g72 | CVE-2026-4372 | HIGH | HuggingFace transformers vulnerable to remote code execution |
| GHSA-69w3-r845-3855 | CVE-2026-1839 | MODERATE | HuggingFace Transformers allows for arbitrary code execution in the `Trainer` class |
| GHSA-qq3j-4f4f-9583 | CVE-2025-2099 | MODERATE | Hugging Face Transformers Regular Expression Denial of Service |
| GHSA-4w7r-h757-3r74 | CVE-2025-6921 | MODERATE | Hugging Face Transformers vulnerable to Regular Expression Denial of Service (ReDoS) in the AdamWeightDecay optimizer |
| GHSA-rcv9-qm8p-9p6j | CVE-2025-6051 | MODERATE | Hugging Face Transformers library has Regular Expression Denial of Service |
| GHSA-59p9-h35m-wg4g | CVE-2025-6638 | MODERATE | Hugging Face Transformers is vulnerable to ReDoS through its MarianTokenizer |
| GHSA-37mw-44qp-f5jm | CVE-2025-3933 | MODERATE | Transformers is vulnerable to ReDoS attack through its DonutProcessor class |
| GHSA-9356-575x-2w9m | CVE-2025-5197 | MODERATE | Hugging Face Transformers Regular Expression Denial of Service (ReDoS) vulnerability |
| GHSA-489j-g2vx-39wf | CVE-2025-3262 | MODERATE | Transformers vulnerable to ReDoS attack through its SETTING_RE variable |
| GHSA-fpwr-67px-3qhx | CVE-2025-1194 | MODERATE | Transformers Regular Expression Denial of Service (ReDoS) vulnerability |
| GHSA-phhr-52qp-3mj4 | CVE-2025-3777 | LOW | Transformers's Improper Input Validation vulnerability can be exploited through username injection |
| GHSA-jjph-296x-mrcr | CVE-2025-3264 | MODERATE | Transformers vulnerable to ReDoS attack through its get_imports() function |
| GHSA-q2wp-rjmx-x6x9 | CVE-2025-3263 | MODERATE | Transformers's ReDoS vulnerability in get_configuration_file can lead to catastrophic backtracking |
| GHSA-6rvg-6v2m-4j46 | CVE-2024-12720 | MODERATE | Transformers Regular Expression Denial of Service (ReDoS) vulnerability |
| GHSA-wrfc-pvp9-mr9g | CVE-2024-11393 | HIGH | Deserialization of Untrusted Data in Hugging Face Transformers |
| GHSA-qxrp-vhvm-j765 | CVE-2024-11392 | HIGH | Deserialization of Untrusted Data in Hugging Face Transformers |
| GHSA-hxxf-235m-72v3 | CVE-2024-11394 | HIGH | Deserialization of Untrusted Data in Hugging Face Transformers |
| GHSA-v68g-wm8c-6x7j | CVE-2023-7018 | HIGH | transformers has a Deserialization of Untrusted Data vulnerability |
| GHSA-3863-2447-669p | CVE-2023-6730 | CRITICAL | transformers has a Deserialization of Untrusted Data vulnerability |
| GHSA-282v-666c-3fvg | CVE-2023-2800 | MODERATE | transformers has Insecure Temporary File |
| GHSA-37q5-v5qm-c9v8 | CVE-2024-3568 | LOW | Transformers Deserialization of Untrusted Data vulnerability |
CVE-2026-80047 at a Glance
Security Impact
CVSS: 7.8
The security impact metrics below are parsed directly from the CVSS v3.1 vector string (CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H).
How to Fix CVE-2026-80047
To remediate this vulnerability, upgrade all affected Transformers dependencies to a patched release. Both direct and transitive dependencies should be checked, as frameworks or security libraries may pull in vulnerable versions transitively.
pip install transformers>=Latest secure release
Run the following command in your project directory to verify that all resolved dependencies use the patched release:
pip show transformers
How to Check if You Are Affected
To determine whether your Python environment is affected by CVE-2026-80047, inspect your installed packages using pip:
pip show transformers
Analyze the terminal output to locate any version of Transformers in your dependency tree. If any resolved version matches the vulnerable ranges listed in the table below (for example, versions earlier than the patched release), your application is affected.
Note: The pip show transformers command checks currently installed packages in your active Python environment.
Technical Details
Vulnerability Mechanism & Technical Analysis
This vulnerability is classified under CWE-273: Improper Check for Dropped Privileges, where the product attempts to drop privileges but does not check or incorrectly checks to see if the drop succeeded. In Transformers, software state validation or input boundaries are circumvented during processing. Review the official source advisory statement below for specific affected components, classes, and reproduction details.
Source Advisory Statement
CVE-2026-80047 Affected Transformers Versions
1 component(s)| Product / Package | Ecosystem | Affected Versions | Fixed Version |
|---|---|---|---|
|
transformers
|
PyPI |
≥ 4.49.0
≤ 5.8.1
|
|
References
Frequently Asked Questions
What is CVE-2026-80047?
Which Transformers versions are affected?
What version fixes CVE-2026-80047?
Is transformers affected?
What is the CVSS score of CVE-2026-80047?
What CWE is associated with CVE-2026-80047?
How can I check PyPI dependencies for CVE-2026-80047?
Machine-readable OSV JSON
OSV v1.5 Schema
{
"schema_version" : "1.4.0",
"id" : "GHSA-x9r9-c232-4q39",
"modified" : "2026-09-30T23:40:28Z",
"published" : "2026-09-01T15:31:13Z",
"aliases" : [ "CVE-2026-80047" ],
"summary" : "Hugging Face Transformers downloads custom generation code before trust consent",
"details" : "A vulnerability in Hugging Face Transformers (versions 4.49.0, <= 5.8.1) allows remote Python files to be written to local disk without user consent when using GenerativePreTrainedModel.load_custom_generate(). The function fetches and caches a remote module file before performing the required trust_remote_code consent check, inverting the security model enforced by other code-loading paths (such as AutoConfig, AutoModel, and AutoTokenizer). As a result, attacker‑controlled Python code from custom_generate/generate.py is copied into the user’s ~/.cache/huggingface/modules directory even if the user declines the trust prompt. Although execution is correctly gated, the file write is not reversible and can persist across sessions. This can lead to persistent, unauthorized files on disk and stale cache collisions where cached attacker code may later be executed during trusted model loads. The issue stems from an unconditional file write in dynamic_module_utils.py prior to any trust verification.",
"severity" : [ {
"type" : "CVSS_V3",
"score" : "CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H"
} ],
"affected" : [ {
"package" : {
"ecosystem" : "PyPI",
"name" : "transformers"
},
"ranges" : [ {
"type" : "ECOSYSTEM",
"events" : [ {
"introduced" : "4.49.0"
}, {
"last_affected" : "5.8.1"
} ]
} ]
} ],
"references" : [ {
"type" : "ADVISORY",
"url" : "https://nvd.nist.gov/vuln/detail/CVE-2026-80047"
}, {
"type" : "WEB",
"url" : "https://github.com/huggingface/transformers/pull/48620"
}, {
"type" : "WEB",
"url" : "https://github.com/huggingface/transformers/commit/cbc1651a032b923da7f4b44b3d0e6f68e6ba6b55"
}, {
"type" : "PACKAGE",
"url" : "https://github.com/huggingface/transformers"
}, {
"type" : "WEB",
"url" : "https://kb.cert.org/vuls/id/456290"
}, {
"type" : "WEB",
"url" : "https://www.kb.cert.org/vuls/id/456290"
} ],
"database_specific" : {
"cwe_ids" : [ "CWE-273" ],
"severity" : "HIGH",
"github_reviewed" : true,
"github_reviewed_at" : "2026-09-30T23:40:28Z",
"nvd_published_at" : "2026-09-01T14:17:41Z"
}
}