I'd agree that defending against malicious admins is really difficult. We have really little context to go by here, but I think there is important distinction to be made if the malicious actions (planting backdoors or whatnot) were done while the malicious actor was still employed or after their employment was terminated. Proper exit procedures protect against the latter, but generally are not that effective against the former.
> but in the end humans make mistakes
And it is useful for us outsiders to highlight the real mistakes so that we can learn from them, because that is really the biggest value of stories like this for the majority of people who are not directly impacted.
If I leave $20 on the sidewalk the thief is wrong for stealing it, but it is partially my fault for being stupid enough to leave $20 on the ground.
This can easily happen to anyone -
5PM Friday - $admin and $ceo have a fight
6PM Friday - $admin decides he's had enough with $company and $ceo, and wipes everything
7PM Friday - $admin is fired
Where are your exit procedures now? Are the current admins stupid for not having the foresight that $admin and $ceo's fight would have resulted in the worst?Note that I don't know the details and am making assumptions that may be wrong about the case in question, but in general, if you can't deny access quickly to any given account, you really want to fix that. Not just because of rogue ex-employees - what happens when $important_person's account is compromised?
No. It's easy to revoke access to a user. An admin is different - an admin can install whatever he wants to give him backdoor access. Or a timebomb.
Detecting unauthorized software from a rogue privileged user is a different problem with very different mitigations. It is a great topic that I'm personally interested in, given that I'm implementing controls for that, but I wasn't discussing that.