I found it to be a security issue in our network. If I connected to a document in the “outside world”, still from our network, it would still fail. I then tried to access the “outside world” file from a different, non-corporate network, and it worked out fine.
I found that documents - even those on our corporate network - would open just fine in a webbrowser, and I thought of something I did years ago in another solution: pretending to be a browser.
Using a network traffic sniffing tool, Fiddler, I found out exactly how a http-request from a browser looked, especially the values in the http headers. I copied the values from the http header, and in UIPath I added those values to the Headers property in the HTTPRequest activity in UIPath.
With a little bit of work, the request sent by the HTTPRequest activity in UIPath looked enough like a webbrowser request, that our secutity policy let the traffic through, and I could retrieve the .xlsx with no problems.
I have since explained to our security team what I am trying to do, and they have since agreed to loosen the security for this specific type of traffic, of course still limited by IP address etc.
Hope it makes sense, otherwise ask