A high-severity cross-site scripting vulnerability in Open WebUI allows an authenticated attacker to perform account takeover via malicious terminal port previews.

The application improperly configures iframe sandboxing within the PortPreview component, allowing malicious scripts to execute with the same origin as the Open WebUI platform. This flaw can be triggered by an authenticated user who hosts malicious content on a shared terminal server, which is subsequently rendered when a victim views the preview.

Successful exploitation allows an attacker to gain full control over a victim's account, leading to unauthorized access to sensitive data, platform configurations, and AI conversation history. With a CVSS score of 8.7, this vulnerability represents a significant risk to organizational confidentiality and integrity, particularly in environments where Open WebUI is used for collaborative or shared AI tasks.

Immediate Action: Upgrade the Open WebUI installation to version 0.11.1 or later to apply the necessary security patches for the PortPreview component.

Proactive Monitoring: Monitor application access logs for unusual activity or unexpected redirects originating from the terminal proxy or preview components.

Compensating Controls: Restrict access to terminal sharing features to trusted users only and implement a strong Content Security Policy (CSP) to mitigate the impact of cross-site scripting attacks if an immediate update is not feasible.

Public Exploit Available: Unknown

Given the potential for complete account takeover and the identified existence of proof-of-concept material, this vulnerability poses a severe risk to the security of your AI infrastructure. Organizations should prioritize updating their Open WebUI instances to version 0.11.1 immediately to eliminate the underlying cross-site scripting flaw and protect user accounts from unauthorized access.