CVE-2026-76841: Xinference model loading executes remote code from Hugging Face directories

Summary

Xinference loads Hugging Face models with trust_remote_code unconditionally enabled, and versions before 2.12.0 provide no setting to disable it. A caller with model launch access can register a model with an arbitrary path, causing AutoTokenizer.from_pretrained to import and execute Python declared in the model's tokenizer_config.json auto_map, with the privileges of the worker process.

Mitigation

Fixed in 2.12.0, which gates every site behind allow_trust_remote_code and the XINFERENCE_TRUST_REMOTE_CODE setting, permitting remote code only for bundled built-in models.

Related items

• MediumCVE-2026-121936: easyVoice server-side request forgery through openaiBaseUrl in /api/v1/ttsSimilar attack · NVD/CVE Database
• MediumCVE-2026-118628: MLflow get-artifacts server-side request forgery information disclosureSimilar attack · NVD/CVE Database
• HighCVE-2026-118625: MLflow Tracking Server privilege escalation via get_endpointsSimilar attack · NVD/CVE Database
• InfoWhen trust becomes the payload: Inside the threats, etc quarterly reportSimilar attack · Elastic Security Labs
• MediumCritical Langflow flaws enable authentication bypass and RCESimilar attack · CSO Online