They are also constantly evolving, and make changes to the algorithms that attempt to determine if a given site you're visiting is a phishing site. This means sometimes there's something that a given site (like DU) has been doing code-wise, perhaps for a long time, could suddenly be marked as "sketchy" because some new hack that hit the web leveraged similar code and the Guardio algorithm decides to take a broad, blunt force approach that ends up black-listing a lot of non-dangerous sites. I wouldn't take this particular incorrect report by the software to be one based on DU politics, though I suppose it's possible.
Also, sites run by small orgs with limited funds tend to get caught up in things like this more often than those run by big corporations with dedicated IT teams, simply because there's a LOT of updates that need to applied to the open-source code that your site is running, and programs like Guardio can detect software versions for certain client-side code (that runs in the user's browser). If you don't have staff to keep up with every update, you can start getting flagged by these sorts of detectors because you're pushing code from a older software from your site, one that's no longer considered 'safe'.
And sometimes these new security-compliant version updates aren't backwards compatible with your existing code, which means coming into proper security compliance can involve a big expense because you have to re-code a bunch of stuff, and you can be flagged by the web's various warning systems until you get your software versions up to date.
= new reply since forum marked as read