Hi, I'm seeing that the keyAgreement KU bit is not explicitly forbidden for 
ECDSA TLS certificates (e.g. in the CAB Forum Baseline Requirements or the 
Mozilla Root Store Policy), but why not considering that this is an enabler 
for the KCI-based MitM attack described in https://kcitls.org?. 

Now, I'm looking that there were already some discussions (e.g. 
https://archive.cabforum.org/pipermail/public/2016-February/023207.html) on 
forbidding it, but I'm not really sure of the motivations. Anyway, why it 
didn't get forbidden back then?.

In the other hand, I'm not really sure if there are still some TLS 
implementations vulnerable to this attack (specially the ones that support 
fixed (EC)DH client authentication), but given that there might exist some 
outdated setups, isn't it an unnecesary risk to allow this KU bit for ECDSA 
certificates?.

Finally, if the KU extension is not set at all in the certificate, this 
attack is still possible so the fact that the BRs make the KU optional 
might be problematic.

-- 
You received this message because you are subscribed to the Google Groups 
"[email protected]" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To view this discussion on the web visit 
https://groups.google.com/a/mozilla.org/d/msgid/dev-security-policy/4b7eac35-6865-4500-8a6f-dae78e95ad02n%40mozilla.org.

Reply via email to