Open WebUI Cache Misconfiguration: Potential Cross-User Model Exposure
An early warning has been issued regarding a potential vulnerability in Open WebUI's get_all_models handlers, which may expose user model lists due to a cache misconfiguration.
What happened
Reportedly, the get_all_models handlers in routers/openai.py and routers/ollama.py were intended to cache permission-filtered model lists per user. However, the @cached decorator appears to have been misconfigured, resulting in a shared cache entry for all users. This misconfiguration is under investigation for potentially serving one user's model list to another authenticated user. The recommended action is to review and update the caching configuration in routers/openai.py and routers/ollama.py to use key_builder instead of key in the @cached decorator. For more details, consult the primary sources.
How 0Day mitigates this
open_webui is anywhere in your dependency tree, the engineers who own the affected repositories get a push alert the moment it is flagged — no manual audit to remember to run.Read how this differs from waiting on a scanner to catch a known advisory, or see the exact, read-only access 0Day needs to do this for an organization.