Public Access to PDF Files in Nearline Storage via Vertex AI Search results

[CSRMedia]

I have set up a Vertex search inside a bucket that has Public Access enabled (allAuthenticatedUsers & allUsers set to Storage Object Viewer).  When I look at the Cloud Storage Bucket, I see " Public to internet".

For each individual object or file, I can copy and access the Public URL without any issues on several computers.

But when I search the same bucket in the Vertext search interface and click on the search results for the same files, I get 403 message when I try to preview or open the PDF file.  In the AI Applications settings, I have even deleted the Data (documents) as it was indexed before making the bucket public.  Then imported the data back in after making the bucket public but the search results still don't open.

When I logged in to my Google Account, I'm able to click on the link and access the files without any issues.

I just noticed that the URLs are slightly different:

This is the URL for a PDF file from the search results:

https://storage.cloud.google.com/sunnybrook_policies/6346032-2011-09-licensing-and-registering-your-...

This is a public URL from the bucket for the same file:

https://storage.googleapis.com/sunnybrook_policies/6346032-2011-09-licensing-and-registering-your-bu...

Any assistance you can provide would be helpful.

Thank you!

[CSRMedia]

Interesting...the links are working now.  But the preview window for each file still doesn't show up like it does when you're logged in.

You can see the demo here:
<URL Removed by Staff>

When you search say "sql" you can preview (click on the eye icon) and it should open the file in the bottom right preview pane.