CVE-2026-76841
Gravedad CVSS v4.0:
ALTA
Tipo:
CWE-94
Control incorrecto de generación de código (Inyección de código)
Fecha de publicación:
24/08/2026
Última modificación:
24/08/2026
Descripción
*** Pendiente de traducción *** Xinference loads models with Hugging Face remote code execution unconditionally enabled, and before version 2.12.0 exposes no setting to disable it. Six loader call sites pass trust_remote_code=True as a literal or as an unconditional default: RerankModel._get_tokenizer in xinference/model/rerank/core.py, SentenceTransformerRerankModel.load in xinference/model/rerank/sentence_transformers/core.py, SentenceTransformerEmbeddingModel.load in xinference/model/embedding/sentence_transformers/core.py, FlagEmbeddingModel.load in xinference/model/embedding/flag/core.py, and two sites in xinference/model/llm/transformers/core.py where PytorchModel._sanitize_model_config and PytorchModel._get_components default the value to True. Because a caller with model launch access can register a model whose type is unknown and supply an arbitrary model path, the server reaches _auto_detect_type and then AutoTokenizer.from_pretrained, which imports and executes Python declared by the model directory's own tokenizer_config.json auto_map, running attacker-supplied code with the privileges of the worker process. Version 2.12.0 gates every site behind allow_trust_remote_code and the XINFERENCE_TRUST_REMOTE_CODE setting, permitting remote code only for bundled built-in models.
Impacto
Puntuación base 4.0
8.70
Gravedad 4.0
ALTA
Puntuación base 3.x
8.80
Gravedad 3.x
ALTA
Referencias a soluciones, herramientas e información
- https://github.com/xorbitsai/inference
- https://github.com/xorbitsai/inference/blob/v2.11.0/xinference/model/rerank/core.py
- https://github.com/xorbitsai/inference/issues/5023
- https://github.com/xorbitsai/inference/pull/5027
- https://www.vulncheck.com/advisories/xinference-through-remote-code-execution-via-hardcoded-trust-remote-code-in-model-loaders