CVE-2026-46432: CWE-94: Improper Control of Generation of Code ('Code Injection') in InternLM lmdeploy
LMDeploy is a toolkit for compressing, deploying, and serving large language models. In versions 0.12.3 and prior, LMDeploy is vulnerable to arbitrary code execution through hardcoded "trust_remote_code=True" in multiple HuggingFace model-loading call sites. At time of publication, there are no publicly available patches.
AI Analysis
Technical Summary
CVE-2026-46432 affects InternLM's lmdeploy toolkit for compressing, deploying, and serving large language models. In versions 0.12.3 and prior, the product uses hardcoded 'trust_remote_code=True' in multiple HuggingFace model-loading call sites, which leads to improper control of code generation (CWE-94). This vulnerability enables arbitrary code execution under conditions requiring local access. There is no official patch or remediation available at the time of publication.
Potential Impact
Successful exploitation allows an attacker with local access to execute arbitrary code with the privileges of the lmdeploy process. This can lead to full compromise of the affected system, including confidentiality, integrity, and availability impacts as indicated by the CVSS score of 7.8 (high).
Mitigation Recommendations
No official patch or fix is currently available. Users should avoid running vulnerable versions (<=0.12.3) in untrusted environments and restrict local access to trusted users only. Monitor vendor advisories for updates regarding patches or mitigations.
CVE-2026-46432: CWE-94: Improper Control of Generation of Code ('Code Injection') in InternLM lmdeploy
Description
LMDeploy is a toolkit for compressing, deploying, and serving large language models. In versions 0.12.3 and prior, LMDeploy is vulnerable to arbitrary code execution through hardcoded "trust_remote_code=True" in multiple HuggingFace model-loading call sites. At time of publication, there are no publicly available patches.
CVSS v3.1
Score 7.8high
Affected software
Run on your own infrastructure? Check whether these packages are installed with threat-finder — our free open-source scanner.
Weaknesses
AI-Powered Analysis
Machine-generated threat intelligence
Technical Analysis
CVE-2026-46432 affects InternLM's lmdeploy toolkit for compressing, deploying, and serving large language models. In versions 0.12.3 and prior, the product uses hardcoded 'trust_remote_code=True' in multiple HuggingFace model-loading call sites, which leads to improper control of code generation (CWE-94). This vulnerability enables arbitrary code execution under conditions requiring local access. There is no official patch or remediation available at the time of publication.
Potential Impact
Successful exploitation allows an attacker with local access to execute arbitrary code with the privileges of the lmdeploy process. This can lead to full compromise of the affected system, including confidentiality, integrity, and availability impacts as indicated by the CVSS score of 7.8 (high).
Mitigation Recommendations
No official patch or fix is currently available. Users should avoid running vulnerable versions (<=0.12.3) in untrusted environments and restrict local access to trusted users only. Monitor vendor advisories for updates regarding patches or mitigations.
Technical Details
- Data Version
- 5.2
- Assigner Short Name
- GitHub_M
- Date Reserved
- 2026-05-13T22:18:22.830Z
- Cvss Version
- 3.1
- State
- PUBLISHED
- Remediation Level
- null
Threat ID: 6a28a1018dd33fbd85928648
Added to database: 06/09/2026, 23:25:53 UTC
Last enriched: 06/17/2026, 08:41:34 UTC
Last updated: 07/31/2026, 19:22:59 UTC
Views: 61
Community Reviews
0 reviewsCrowdsource mitigation strategies, share intel context, and vote on the most helpful responses. Sign in to add your voice and help keep defenders ahead.
Want to contribute mitigation steps or threat intel context? Sign in or create an account to join the community discussion.
Actions
Updates to AI analysis require Pro Console access. Upgrade inside Console → Billing.
External Links
Need more coverage?
Upgrade to Pro Console for AI refresh and higher limits.
For incident response and remediation, OffSeq services can help resolve threats faster.
Latest Threats
Check if your credentials are on the dark web
Instant breach scanning across billions of leaked records. Free tier available.