CVE-2026-41523: CWE-94: Improper Control of Generation of Code ('Code Injection') in vllm-project vllm
vLLM versions prior to 0.22.0 contain a code injection vulnerability due to an assert-based security check bypass when running in Python optimized mode. This allows unauthenticated attackers to execute arbitrary code by publishing a malicious HuggingFace model. The issue is fixed in version 0.22.0.
AI Analysis
Technical Summary
CVE-2026-41523 is a code injection vulnerability (CWE-94) in the vLLM inference engine for large language models. The vulnerability arises from an assert-based security check in the activation function loading mechanism that can be bypassed when vLLM runs with Python optimizations enabled (python -O or PYTHONOPTIMIZE=1). This allows an unauthenticated attacker to achieve arbitrary code execution on the server by supplying a malicious HuggingFace model. The vulnerability is resolved in vLLM version 0.22.0.
Potential Impact
An attacker can execute arbitrary code on the server running vulnerable versions of vLLM (<0.22.0) without authentication, potentially leading to full system compromise. The vulnerability affects confidentiality, integrity, and availability of the affected system.
Mitigation Recommendations
Upgrade vLLM to version 0.22.0 or later, where this vulnerability is fixed. No other mitigations are indicated. Patch status is confirmed fixed in 0.22.0.
CVE-2026-41523: CWE-94: Improper Control of Generation of Code ('Code Injection') in vllm-project vllm
Description
vLLM versions prior to 0.22.0 contain a code injection vulnerability due to an assert-based security check bypass when running in Python optimized mode. This allows unauthenticated attackers to execute arbitrary code by publishing a malicious HuggingFace model. The issue is fixed in version 0.22.0.
CVSS v3.1
Score 7.5high
Affected software
Run on your own infrastructure? Check whether these packages are installed with threat-finder — our free open-source scanner.
AI-Powered Analysis
Machine-generated threat intelligence
Technical Analysis
CVE-2026-41523 is a code injection vulnerability (CWE-94) in the vLLM inference engine for large language models. The vulnerability arises from an assert-based security check in the activation function loading mechanism that can be bypassed when vLLM runs with Python optimizations enabled (python -O or PYTHONOPTIMIZE=1). This allows an unauthenticated attacker to achieve arbitrary code execution on the server by supplying a malicious HuggingFace model. The vulnerability is resolved in vLLM version 0.22.0.
Potential Impact
An attacker can execute arbitrary code on the server running vulnerable versions of vLLM (<0.22.0) without authentication, potentially leading to full system compromise. The vulnerability affects confidentiality, integrity, and availability of the affected system.
Mitigation Recommendations
Upgrade vLLM to version 0.22.0 or later, where this vulnerability is fixed. No other mitigations are indicated. Patch status is confirmed fixed in 0.22.0.
Technical Details
- Data Version
- 5.2
- Assigner Short Name
- GitHub_M
- Date Reserved
- 2026-04-20T18:18:50.682Z
- Cvss Version
- 3.1
- State
- PUBLISHED
- Remediation Level
- null
Threat ID: 6a39b9b1eed863c81e85ff8f
Added to database: 06/22/2026, 22:39:45 UTC
Last enriched: 06/22/2026, 22:54:09 UTC
Last updated: 06/23/2026, 00:18:17 UTC
Views: 6
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.
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.