sentence-transformers Arbitrary Code Execution on Local Model Load Despite trust_remote_code=False
Key facts
- Attack vector
- Network
- Privileges required
- None
- User interaction
- None
- Attack complexity
- Low
- Attack requirements
- None
- Impact type
- Code execution
- Confidentiality · integrity · availability
- High · High · High
- Fix status
- Partial Red Hat: 3 fixed, 5 without a fix
Attack profile from CVSS 4.0 · CNA. Fix status from CNA, GitHub and Red Hat. Impact type from CWE-94.
Summary
sentence-transformers contains a security control bypass vulnerability that allows attackers to achieve arbitrary code execution by exploiting a logic flaw in the import_module_class helper within sentence_transformers/util/misc.py, where the guard condition includes an 'or os.path.exists(model_name_or_path)' clause that satisfies the trust gate whenever the supplied path exists on the local filesystem, regardless of the trust_remote_code=False argument. Attackers who can control or influence the contents of a model directory on disk can place malicious Python files such as modeling_*.py referenced via modules.json, causing the code to execute at import time when an application loads the model with SentenceTransformer(path, trust_remote_code=False), bypassing the documented security contract and achieving code execution within the loading process.
Why it matters now
SSVC decision points
CISA ADP · SSVC 2.0.3
- Exploitation
- Poc
- Automatable
- Yes
- Technical impact
- Total
The decision (Track, Attend, Act) also depends on your mission and well-being impact; see CISA's SSVC guide.
Affected products
| Vendor | Product | Affected versions | Fixed in | Status | Source |
|---|---|---|---|---|---|
| huggingface | sentence-transformers | up to 5.5.1 | — | No fix listed | CNA |
| — | sentence-transformers (PyPI) | before 5.6.0 | 5.6.0 | Patch available | GHSA-jhr6-gm9c-rqjv |
- sentence-transformers Vendorhuggingface Affected up to 5.5.1 Fixed in — No fix listed CNA
- sentence-transformers (PyPI) Vendor— Affected before 5.6.0 Fixed in 5.6.0 Patch available GHSA-jhr6-gm9c-rqjv
Remediation
- Fix · GHSA-jhr6-gm9c-rqjv sentence-transformers (PyPI) Fixed in 5.6.0 Advisory →
Red Hat 3 advisories
- RHSA-2026:87410 Released 2026-10-21 · Red Hat rating CriticalRHOAI 2.25.12 - Red Hat OpenShift AI
- Red Hat OpenShift AI 2.25 registry.redhat.io/rhoai/odh-built-in-detector-rhel9, registry.redhat.io/rhoai/odh-caikit-nlp-rhel9, registry.redhat.io/rhoai/odh-codeflare-operator-rhel9 +47
- RHSA-2026:84734 Released 2026-10-19 · Red Hat rating CriticalRHOAI 3.4.5 - Red Hat OpenShift AI
- Red Hat OpenShift AI 3.4 registry.redhat.io/rhoai/odh-ai-gateway-payload-processing-rhel9, registry.redhat.io/rhoai/odh-automl-rhel9, registry.redhat.io/rhoai/odh-autorag-rhel9 +47
- RHSA-2026:73987 Released 2026-09-30 · Red Hat rating CriticalRHOAI 3.3.7 - Red Hat OpenShift AI
- Red Hat OpenShift AI 3.3 registry.redhat.io/rhoai/odh-built-in-detector-rhel9, registry.redhat.io/rhoai/odh-dashboard-rhel9, registry.redhat.io/rhoai/odh-data-science-pipelines-argo-argoexec-rhel9 +47
Vendor assessments
Red Hat
Red Hat rating: Important. Red Hat's page for this CVE →
| Product | Status | Plan or fix |
|---|---|---|
| Lightspeed Core | Affected | No fix yet |
| OpenShift Lightspeed | Affected | No fix yet |
| Red Hat Ansible Automation Platform 2 | Affected | No fix planned: Will not fix |
| Red Hat Enterprise Linux AI (RHEL AI) 3 | Affected | No fix yet |
| Red Hat OpenShift AI (RHOAI) | Affected | No fix yet |
| Red Hat OpenShift AI 2.25 | Fixed | RHSA-2026:87410 |
| Red Hat OpenShift AI 3.3 | Fixed | RHSA-2026:73987 |
| Red Hat OpenShift AI 3.4 | Fixed | RHSA-2026:84734 |
Not affected: 5 products
- OpenShift Lightspeed
- Red Hat OpenShift AI (RHOAI)
- Red Hat OpenShift AI 2.25
- Red Hat OpenShift AI 3.3
- Red Hat OpenShift AI 3.4
Timeline
- 2026-10-02 Threatist record created from CVE-2026-68770
- 2026-10-03 Alias added: GHSA-jhr6-gm9c-rqjv
- 2026-10-07 Fix available: RHSA-2026:73987 · Red Hat
- 2026-10-20 Fix available: RHSA-2026:84734 · Red Hat
- 2026-10-20 Fix available: Red Hat OpenShift AI 3.4 · Red Hat
- 2026-10-22 Fix available: RHSA-2026:87410 · Red Hat
- 2026-10-23 Fix available: Red Hat OpenShift AI 2.25 · Red Hat
References
Patch 2
Other 6
- GitHub Issue
- Product Repository
- https://github.com/huggingface/sentence-transformers/pull/3935
- https://github.com/huggingface/sentence-transformers/commit/626eb602b088878a8173bcab69225f68411b99e1
- https://github.com/huggingface/sentence-transformers/releases/tag/v5.6.0
- https://github.com/huggingface/sentence-transformers/releases/tag/v6.0.0
More huggingface vulnerabilities
- TI-2026-20914 LeRobot Unsafe Deserialization Remote Code Execution via gRPC CVSS 4.0CRITICAL 9.3
- TI-2026-04577 Policy Bypass in LightGlue Nested Config Resolution in huggingface/transformers CVSS 3.1CRITICAL 9.6
- TI-2025-04589 Sandbox Escape Vulnerability in huggingface/smolagents CVSS 3.1CRITICAL 10.0
- TI-2026-79938 Pickle deserialization in Hugging Face Datasets allows code execution CVSS 4.0HIGH 8.7
- TI-2026-32710 Diffusers: None.py Trust Remote Code Bypass CVSS 3.1HIGH 8.8
- TI-2026-00400 Unbounded External Image Fetch in Validation Leads to Resource-Exhaustion DoS in huggingface/text-generation-inference CVSS 3.0HIGH 7.5
Sources
- CVE-2026-68770 · CVE Program
- CVE-2026-68770 · NVD
- FIRST EPSS
- GHSA-jhr6-gm9c-rqjv · GitHub Advisory Database · CC BY 4.0
- CVE-2026-68770 · Red Hat security data · CC BY 4.0
This product uses the NVD API but is not endorsed or certified by the NVD. Sources and licences