CVE-2025-24357High· 7.5▾ Twilightvllm: Malicious model to RCE by torch.load in hf_model_weights_iterator
▾ Twilight zone — High severity, or a signal on a lesser flaw
impact 41.3 · likelihood 0.1 · exploitation 0
Need a working PoC? Pro members can cast a request and our team develops one — it lands right here.
Exploit-prediction probability, daily snapshots since Aug 7.
Disclosure to exploitation, from the record and what we observed since indexing it.
Disclosed via OSV
Last analysed / modified upstream
0.7%
The vllm/model_executor/weight_utils.py implements hf_model_weights_iterator to load the model checkpoint, which is downloaded from huggingface. It use torch.load function and weights_only parameter is default value False. There is a security warning on https://pytorch.org/docs/stable/generated/torch.load.html, when torch.load load a malicious pickle data it will execute arbitrary code during unpickling.
This vulnerability can be exploited to execute arbitrary codes and OS commands in the victim machine who fetch the pretrained repo remotely.
Note that most models now use the safetensors format, which is not vulnerable to this issue.
vllm < 0.7.0Upgrade to a patched release:
vllm 0.7.0Connected by shared product, vendor, weakness, or advisory.
GHSA-ggpf-24jw-3fcwCritical· 9.8CVE-2025-24357 Malicious model remote code execution fix bypass with PyTorch < 2.6.0
CVE-2026-69147Medium· 6.5vLLM is an inference and serving engine for large language models
CVE-2026-57173Medium· 6.5vLLM is an inference and serving engine for large language models
CVE-2026-73558Medium· 5.3vLLM is an inference and serving engine for large language models
CVE-2026-73560Medium· 6.5vLLM is an inference and serving engine for large language models
CVE-2026-71486Medium· 4.3vLLM is an inference and serving engine for large language models