When has it become normal practice to intercept (as in #MitM attack!) connections secured by a #letsencrypt certificate? That's outrageous!

For years and years I didn't have a single such issue with my selfhosted service, but recently I had the "pleasure" - twice! - to deal with such malicious networks. It's beginning to be a huge annoyance. (Some of my services are relevant for my work.)

Do we have to go back to paid certificates??

#selfhosting #cybersecurity #TLS

@hambier
How exactly did you think this was accomplished? I dont understand what youre claiming? are you saying the root certs are compromised?
@dlakelan I'm saying that the browser is throwing the typical scary warnings and when inspecting it's not the certificate that my server is actually presenting, but rather one issued by Fortinet or some other corporate security software that has been substituted for it.
@hambier
Ah right. Corporations install their own certs on their client machines and then MiTM themselves so as to monitor everything employees do. its a shitty practice. best to stay away from those networks!
@dlakelan @hambier Right. This has nothing to do with using any particular CA oneself for one's services.

@mkj @dlakelan Well, in the first case I mentioned it is (IMO) an incompetent misconfiguration since it's a public(!) Wifi and hence the endpoints are not supposed to be equipped with additional CAs.

And in the second case (workplace) it turned out that some recent config update of the endpoint antivirus suite added this misbehaviour. (Certificate substituted, but no CA added for it...)

In both case Letsencrypt sites seem zo be particularly targeted. Other certs are not substituted.

@mkj @dlakelan LOL. I just noticed that even our very own (workplace) website and intra/extranet are affected. 😂

@hambier @mkj

is it related to some kind of captive portal? Maybe once you sign into that portal the problem goes away? The captive portal would probably need to be accessed via http

@dlakelan @mkj It seems to have been a temporary misconfiguration/misbehaviour of the endpoint security. 🤷‍♂️
After a couple of hours and a few reboots it seems back to normal.