3 ms·
This doesn't mean you should use unsecure methods instead and issue certificates to anyone capable to do MitM.
by codedokode 3y ago
This doesn't mean you should use unsecure methods instead and issue certificates to anyone capable to do MitM.
- blueflow 3y agoHow could Letsencrypt even verify a server setup if not via DNS/HTTP? Also, verify against what? The servers are basically random strangers without identity when they first talk to LE.
- codedokode 3y agoIn this case there has been a valid certificate for the site; this alone should raise suspicion. Also, if they cannot do secure validation then maybe they should stop issuing certificates for sites that already have a proper certificate.
- blueflow 3y agoThis happens all the time when a server is rebuilt from scratch - same cert using a different keypair.
- codedokode 3y agoSo should we conclude that SSL cert infrastructure is completely compromised and now any country can issue fake certificates?
- blueflow 3y agoNo, there is no reason to jump to such extremes.
- codedokode 3y agoThere are approximately 10 Tier-1 ISPs through which majority of Internet traffic passes, and unless I misunderstood something, they can issue valid certificates for almost any domain. To me it looks like "completely compromised".
- blueflow 3y agoEvery CA can issue valid certificates for every domain? And it always has been that way.
- codedokode 3y agoCA has a risk to get their root cert removed from browsers; ISP doesn't risk anything especially when asked by the govt.
- blueflow 3y agoThey risk having their peerings cancelled. Also it might be a crime in some countries.
- preisschild 3y agoOver the DNS challenge, of course.
- sigio 3y agoThis is exactly the same as all other CA's do this.... for DV-certificates you basically place a key/special file on the webserver, or receive a verficiation code via (plaintext) email. For EV certs there might be more validation, but users will never see the difference between EV and DV certificates.
- codedokode 3y agoSo SSL certificates are completely unreliable; we should only wait until Russian or Chinese comrades find a good use for this attack (e.g. temporary redirecting Western traffic using BGP to validate a Let's Encrypt's cert for Western site).
- bananapub 3y agoyes? this is a well-known problem, which is why CAA-ACME etc and certificate transparency logs exist.