CVE-2026-30824
Published: 07 March 2026
Description
Flowise is a drag & drop user interface to build a customized large language model flow. Prior to version 3.0.13, the NVIDIA NIM router (/api/v1/nvidia-nim/*) is whitelisted in the global authentication middleware, allowing unauthenticated access to privileged container management and…
more
token generation endpoints. This issue has been patched in version 3.0.13.
Mitigating Controls (NIST 800-53 r5)AI
AC-14 requires explicit identification and authorization of actions permitted without identification or authentication, directly preventing improper whitelisting of sensitive NVIDIA NIM endpoints in the authentication middleware.
AC-3 enforces approved authorizations for logical access to system resources, addressing the middleware's failure to block unauthenticated access to privileged container management and token generation functions.
AC-6 applies least privilege to restrict access to only necessary functions, mitigating the scope of damage from unauthenticated exploitation of privileged endpoints.
Security SummaryAI
CVE-2026-30824 is a critical authentication bypass vulnerability (CWE-306) in Flowise, an open-source drag-and-drop user interface for building customized large language model (LLM) flows. In versions prior to 3.0.13, the NVIDIA NIM router endpoints (/api/v1/nvidia-nim/*) are whitelisted in the global authentication middleware, enabling unauthenticated access to privileged container management and token generation functions. The vulnerability carries a CVSS v3.1 base score of 9.8 (AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H), reflecting its high severity due to complete lack of authentication controls on sensitive endpoints.
Any unauthenticated attacker with network access to a vulnerable Flowise instance can exploit this issue remotely with low complexity and no user interaction required. Successful exploitation grants access to privileged operations, including container management and token generation, potentially allowing full compromise of the confidentiality, integrity, and availability of the affected system through arbitrary code execution, data exfiltration, or persistent control.
The issue was addressed in Flowise version 3.0.13, as detailed in the project's release notes and security advisory. Security practitioners should upgrade to this version or later to mitigate the vulnerability, with further details available in the GitHub security advisory (GHSA-5f53-522j-j454) and release tag (flowise@3.0.13).
Flowise's focus on LLM orchestration introduces AI/ML relevance, as exploited instances could disrupt or hijack model inference pipelines integrated with NVIDIA NIM containers. No public evidence of real-world exploitation has been reported as of the CVE publication on 2026-03-07.
Details
- CWE(s)
Affected Products
AI Security AnalysisAI
- AI Category
- Other AI Platforms
- Risk Domain
- N/A
- OWASP Top 10 for LLMs 2025
- None mapped
- MITRE ATLAS Techniques
- None mapped
- Classification Reason
- Matched keywords: large language model
MITRE ATT&CK Enterprise TechniquesAI
Why these techniques?
The vulnerability is an authentication bypass in a public-facing web application (Flowise API endpoints), enabling unauthenticated remote exploitation (T1190). It directly grants access to privileged container management functions, facilitating container administration commands (T1609).