HighVulnerabilityLLM-specific

CVE-2025-24357: vLLM deserialization of untrusted model weights via torch.load

Source: NVD(opens in a new tab)
Identifier: CVE-2025-24357
Published: January 27, 2025
Record updated: October 9, 2026

Affected: vllm < 0.7.0
Fixed in: 0.7.0
released January 27, 2025, on the day of disclosure
Known exploitation: Not listed in the CISA Known Exploited Vulnerabilities catalog at the last check.
EPSS: 0.9%

Summary

CVE-2025-24357 affects vLLM, a library for LLM inference and serving. In vllm/model_executor/weight_utils.py, the hf_model_weights_iterator function loads model checkpoints downloaded from Hugging Face using torch.load with weights_only defaulting to False. When torch.load loads malicious pickle data, it executes arbitrary code during unpickling. The weakness is classified as CWE-502, Deserialization of Untrusted Data.

Mitigation

This vulnerability is fixed in v0.7.0.

Details

Scores and exploitation

CVSS score: 7.5(high)
Patch available: Yes
EPSS: 0.9%Estimated probability that this CVE is exploited in the next 30 days, from the FIRST Exploit Prediction Scoring System. Read on November 4, 2026.

Weaknesses and techniques

Weakness (CWE): CWE-502
Attack pattern (CAPEC): CAPEC-586
AI threat technique (MITRE ATLAS): AML.T0010

Affected

Packages: vllm@< 0.7.0 (fixed: 0.7.0)

Classification

Attack type: Supply chain
Attack sophistication: Moderate
AI component targeted: Inference
Impact: Confidentiality,
Integrity,
Availability
Labels: security

How this record was produced

Summaries, mitigations and labels on this page are generated automatically from the source. The linked source is authoritative. The source gives this advisory a description and no title, so the title above is a headline generated from the source text.

Original source: https://nvd.nist.gov/vuln/detail/CVE-2025-24357

First tracked: February 15, 2026 at 08:43 PM ET, from NVD/CVE Database

Classifier: claude-haiku-5-5, prompt v4

Classifier confidence: 95% (self-assessed)

Affected packages in the Exposure Registry

Matched by package name and ecosystem. Each entry shows whether the package delegates to a language model and how many tracked packages depend on it.

vllmPyPILLM dependency since 2023-08-25 · 15 tracked dependents

Topics

Inference infrastructure

Summary

Mitigation

Affected packages in the Exposure Registry

Topics

Related items