Database/AI/ML frameworks & serving
vLLM (hardcoded `trust_remote_code=True`): Two model implementation files force remote code execution regardless
CVSS 8.8CVE-2026-4944AI/ML frameworks & servingcurated
Impact
Two model implementation files force remote code execution regardless of operator setting
Who can reach it
Customer-supplied model repo
What to do
Upgrade past 0.14.1. Operator-level --trust-remote-code=false does not protect you — no config fix
References
Related entries
- vLLM (sparse tensor validation): Missing sparse-tensor invariant checks in multimodal embeddingsCVE-2026-56340 · vLLM (sparse tensor validation)High
- Ray (WebDataset reader): Unsafe deserializationCVE-2026-57516 · Ray (WebDataset reader)High
- Weaviate: RBAC role assignment does not verify the assigner holds the granted permissionsCVE-2026-59093 · WeaviateHigh
- LiteLLM Proxy: forged Authorization header reaches MCP tooling without a valid API keyCVE-2026-59822 · LiteLLM Proxy (MCP Streamable HTTP endpoint authentication)High
- NVIDIA OpenShell: a malicious gateway can inject OS commands into the connecting clientCVE-2026-65091 · NVIDIA OpenShell (gateway response handling)High
- NVIDIA NeMo: TabularTokenizer unpickles an attacker-supplied .pkl file, giving code executionCVE-2026-65179 · NVIDIA NeMo (TabularTokenizer pickle.load of an untrusted .pkl)High
This entry is curated: imported from vendor advisories with machine assistance, not yet individually verified. Confirm against your vendor's advisory before acting, and report anything wrong.