Back

I have never understood why the ability for users to generate their own key pairs, then upload the public key to the #CA for certification, was removed from all browsers I’m aware of NB2 This is not an issue for organisational users where their own organisation is their CA, since the organisation has many other (simpler) surveillance mechanisms if it chooses…

Finally: this is a huge improvement over unencrypted e-mail flying around, subject to mass surveillance anywhere it travels. But a weakness of relying on the security of an online service (rather than software running on your own PC/smartphone) is that it can be changed in subtle, targeted ways at any time by the provider, which is very difficult for you to detect.

Obviously, #Google (under order from the US government) could serve compromised “updates” at any time to individual users.

It would be technically possible to enable users to compare a “fingerprint” (hash) of security-critical plugins they are running, such as the one supporting Gmail’s #E2EE capability. If I can see I’m getting a different plugin for my OS/CPU to 99% of other users on the same platform, that’s a big warning sign. But I haven’t seen such software widely deployed (yet)

This would be analogous to the “reproducible builds” used to protect software against such targeted attacks. Although I don’t know widely deployed these are yet, either

Here is Google’s own documentation. It seems gmail users *will* use recipients’ own X.509 certificate where available. But there are all sorts of limitations which makes this a less than ideal “solution” https://support.google.com/mail/answer/13317990?sjid=9466913528060288184-NA

@1br0wn If Gmail rolled this out to their general end users, it would be a legally sufficient reason to reconsider the CPS designation

@gateklons And you could imagine some governments might even start using it as a secure, general communications mechanism with their citizens…