Access Control Flaw in Open WebUI AI Platform by Open WebUI
CVE-2026-44556

7.1HIGH

Key Information:

Vendor

Open-webui

Status
Open-webui
Vendor
CVE Published:
15 May 2026

What is CVE-2026-44556?

Open WebUI, a self-hosted artificial intelligence platform, has a security flaw in its /responses endpoint that permits any authenticated user to send requests directly to upstream LLM providers without proper model-specific access restrictions. This flaw arises because the endpoint only verifies user session validity, lacking comprehensive checks for model ownership and access permissions that are in place for the primary chat completion endpoint. As a result, this vulnerability exposes various AI models to unauthorized interactions by any authenticated user through a simple POST request, potentially leading to misuse of AI capabilities. The issue has been addressed in version 0.9.0.

Affected Version(s)

open-webui < 0.9.0

References

https://github.com/open-webui/open-webui/security/advisor...
x_refsource_CONFIRM

CVSS V3.1

Score:
7.1
Severity:
HIGH
Confidentiality:
Low
Integrity:
None
Availability:
Low
Attack Vector:
Network
Attack Complexity:
Low
Privileges Required:
Low
User Interaction:
None
Scope:
Unchanged

Timeline

  • Vulnerability published

  • Vulnerability Reserved

.