I have been reading the ROPEMAKER exploit whitepaper by Mimecast. In the article they are mentioning of sending a safe HTML(which may contain evil text, but as long as it is a plain HTML its safe with respect to the protective mechanisms at the recipients end) to the recipient and do the exploit by selectively displaying the necessary fields in the HTML using remotely loadable CSS.
I don't really get the idea of remotely loadable CSS for emails. How can an attacker/user send CSS along with the email body and load it remotely at the recipient end?
Is this actually possible? If yes, how?