Browse / AI Framework and Platform Flaws / CVE-2026-44558
Open WebUI's Channel Access Grants Bypass filter_allowed_access_grants
- Class
- AI Framework and Platform Flaws (filed by a model that read the source)
- Technique
- Authentication and access control
- Affected
- open-webui (PyPI)
- Gets past
- Authorisation check
- CVSS
- CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:L/I:L/A:N
- EPSS
- 0.2% chance of exploitation in 30 days
How it works, in short
Channel create and update endpoints store access grants without the permission filter other resources apply, so users can set public wildcard grants. (Our summary, written by a model. The source text is below.)
Details
Channel Access Grants Bypass filter_allowed_access_grants
Affected Component
Channel creation and update endpoints:
- backend/open_webui/routers/channels.py (lines 291-340, create_new_channel)
- backend/open_webui/routers/channels.py (lines 617-638, update_channel_by_id)
- backend/open_webui/models/channels.py (lines 825-826, set_access_grants call without filtering)
Affected Versions
Current main branch (commit 6fdd19bf1) and likely all versions supporting user-created group channels with access grants.
Description
All resource routers in Open WebUI (knowledge, models, notes, prompts, tools, skills) call filter_allowed_access_grants() before persisting access grants. This function strips principal_id: "*" wildcard grants from users who lack the relevant sharing.public_* permission, and strips individual user grants from users who lack access_grants.allow_users permission.
The channel router does not call filter_allowed_access_grants on either create or update paths. A non-admin user who can create group channels (or who owns a channel) can submit arbitrary access grants — including public wildcard grants — and those grants are stored verbatim, bypassing the admin's permission framework.
# channels.py — access_grants from form data flow directly into persistence
# No call to filter_allowed_access_grants() anywhere in these paths.
# Compare with knowledge.py / models.py / notes.py / prompts.py / tools.py / skills.py,
# all of which do:
# form_data.access_grants = filter_allowed_access_grants(user, form_data.access_grants)
# before creating or updating.
Attack Scenario
- Admin configures permissions so that regular users do NOT have
sharing.public_channels— public sharing of channels is intended to be admin-only. - Attacker (a regular user) creates or owns a group channel.
- Attacker sends: