https://cve.radiocsirt.org Contains only the most 10 recent entries. http://www.rssboard.org/rss-specification python-feedgen en Thu, 15 Oct 2026 06:13:01 +0000 https://cve.radiocsirt.org/vuln/certfr-2025-avi-1057 certfr-2025-avi-1057 https://cve.radiocsirt.org/vuln/certfr-2025-avi-1057 https://cve.radiocsirt.org/vuln/euvd-2026-215303 EUVD-2026-215303 https://cve.radiocsirt.org/vuln/euvd-2026-215303 https://cve.radiocsirt.org/vuln/fkie_cve-2025-24357 <p>vLLM is a library for LLM inference and serving. vllm/model_executor/weight_utils.py implements hf_model_weights_iterator to load the model checkpoint, which is downloaded from huggingface. It uses the torch.load function and the weights_only parameter defaults to False. When torch.load loads malicious pickle data, it will execute arbitrary code during unpickling. This vulnerability is fixed in v0.7.0.</p> <p>vLLM is a library for LLM inference and serving. vllm/model_executor/weight_utils.py implements hf_model_weights_iterator to load the model checkpoint, which is downloaded from huggingface. It uses the torch.load function and the weights_only parameter defaults to False. When torch.load loads malicious pickle data, it will execute arbitrary code during unpickling. This vulnerability is fixed in v0.7.0.</p> https://cve.radiocsirt.org/vuln/fkie_cve-2025-24357 https://cve.radiocsirt.org/vuln/ghsa-rh4j-5rhw-hr54 <p><strong>Affected:</strong> PyPI: vllm</p> <p>### Description 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.</p> <p>### Impact This vulnerability can be exploited to execute arbitrary codes and OS commands in the victim machine who fetch the pretrained repo remotely.</p> <p>Note that most models now use the safetensors format, which is not vulnerable to this issue.</p> <p>### References * https://pytorch.org/docs/stable/generated/torch.load.html * Fix: https://github.com/vllm-project/vllm/pull/12366</p> <p><strong>Affected:</strong> PyPI: vllm</p> <p>### Description 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.</p> <p>### Impact This vulnerability can be exploited to execute arbitrary codes and OS commands in the victim machine who fetch the pretrained repo remotely.</p> <p>Note that most models now use the safetensors format, which is not vulnerable to this issue.</p> <p>### References * https://pytorch.org/docs/stable/generated/torch.load.html * Fix: https://github.com/vllm-project/vllm/pull/12366</p> https://cve.radiocsirt.org/vuln/ghsa-rh4j-5rhw-hr54 https://cve.radiocsirt.org/vuln/pysec-2025-58 <p><strong>Affected:</strong> PyPI: vllm</p> <p>vLLM is a library for LLM inference and serving. vllm/model_executor/weight_utils.py implements hf_model_weights_iterator to load the model checkpoint, which is downloaded from huggingface. It uses the torch.load function and the weights_only parameter defaults to False. When torch.load loads malicious pickle data, it will execute arbitrary code during unpickling. This vulnerability is fixed in v0.7.0.</p> <p><strong>Affected:</strong> PyPI: vllm</p> <p>vLLM is a library for LLM inference and serving. vllm/model_executor/weight_utils.py implements hf_model_weights_iterator to load the model checkpoint, which is downloaded from huggingface. It uses the torch.load function and the weights_only parameter defaults to False. When torch.load loads malicious pickle data, it will execute arbitrary code during unpickling. This vulnerability is fixed in v0.7.0.</p> https://cve.radiocsirt.org/vuln/pysec-2025-58 https://cve.radiocsirt.org/vuln/wid-sec-w-2025-0868 <p>Ein entfernter, anonymer Angreifer kann eine Schwachstelle in PyTorch ausnutzen, um beliebigen Programmcode auszuführen.</p> <p>Ein entfernter, anonymer Angreifer kann eine Schwachstelle in PyTorch ausnutzen, um beliebigen Programmcode auszuführen.</p> https://cve.radiocsirt.org/vuln/wid-sec-w-2025-0868