PatchSiren cyber security CVE debrief
CVE-2026-46432 InternLM CVE debrief
LMDeploy, a toolkit for compressing, deploying, and serving large language models, is vulnerable to arbitrary code execution. This vulnerability is due to hardcoded 'trust_remote_code=True' in multiple HuggingFace model-loading call sites in versions 0.12.3 and prior. The vulnerability has a CVSS score of 7.8 and is classified as HIGH severity.
- Vendor
- InternLM
- Product
- lmdeploy
- CVSS
- HIGH 7.8
- CISA KEV
- Not listed in stored evidence
- Original CVE published
- 2026-06-10
- Original CVE updated
- 2026-06-10
- Advisory published
- 2026-06-10
- Advisory updated
- 2026-06-10
Who should care
Users of LMDeploy versions 0.12.3 and prior should be aware of this vulnerability and take necessary actions to mitigate it.
Technical summary
The vulnerability is caused by hardcoded 'trust_remote_code=True' in multiple HuggingFace model-loading call sites. This allows for arbitrary code execution, which can lead to severe consequences.
Defensive priority
HIGH
Recommended defensive actions
- Update to a patched version of LMDeploy as soon as available.
- Review and restrict model loading configurations to prevent arbitrary code execution.
Evidence notes
At time of publication, there are no publicly available patches.
Sources and references
Verified primary and authoritative sources
-
CVE-2026-46432 CVE Program record
Publisher, destination, and source semantics verified
URL: https://www.cve.org/CVERecord?id=CVE-2026-46432
CVE Program - Official CVE Program record with source-provided CVE metadata.
-
CVE-2026-46432 NVD vulnerability detail
Publisher, destination, and source semantics verified
URL: https://nvd.nist.gov/vuln/detail/CVE-2026-46432
NIST National Vulnerability Database - Official NIST NVD detail page and source-specific vulnerability assessment.
Supplemental references
-
Source reference
Unverified legacy reference
URL: https://github.com/InternLM/lmdeploy/security/advisories/GHSA-m549-qq94-fvhg
Methodology and review provenance
AI-assisted synthesis based on stored public vulnerability evidence. System validation, approval state, and publication status do not by themselves establish human review of this revision. PatchSiren helps prioritize defensive review and does not prove exposure or remediation on any system.