4 ms·
I've quickly skimmed through the paper and the ProtonMail whitepaper the author supposedly based their analysis on. Sorry to be harsh, but the conclusions of th
by m000 8y ago
I've quickly skimmed through the paper and the ProtonMail whitepaper the author supposedly based their analysis on. Sorry to be harsh, but the conclusions of the analysis are pretty much BS. The reason is that the threat model defined by the author is obviously not the one ProtonMail defends against.
The author makes-up a straw-man threat model which is not based on anything that ProtonMail has published:
> A crucial security assumption, based on ProtonMail's self-professed security goals in its specification documents (2.1), is that the ProtonMail server P is untrusted.
AFAIK, ProtonMail only claims that your email are unencryptable while you don't access the service. Since they have to receive and relay emails to regular SMTP servers and they provide access via IMAP to regular mail programs, obviously their servers at some point will touch your unencrypted email. Therefore, you put a certain amount of trust on them.
So, yes, under the defined straw-man threat model, ProtonMail is not secure. Thank you Captain Obvious!
- lorenzhs 8y agoUnless you count the promises on their front page. "All emails are secured automatically with end-to-end encryption. This means even we cannot decrypt and read your emails. As a result, your encrypted emails cannot be shared with third parties." Obviously that's not possible when receiving mail via SMTP and providing access via IMAP. But that's not clear to non-technical users. I think it's fair to say that putting the above quote on the landing page is misleading. EDIT: I was wrong on IMAP, which requires users to install a bridge application that handles decryption. That doesn't change anything about SMTP, though.
- freeflight 8y ago> Obviously that's not possible when receiving mail via SMTP and providing access via IMAP. Afaik on desktop all those involve the use of the ProtonMail Bridge application [0]? At least that was my understanding when I couldn't get my Thunderbird setup to check ProtonMail accounts, it requires the Bridge application. The paper is weird in that context because the Bridge application is mentioned only once in the whole paper, as a side note: "For simplicity, we group it with smartphone applications." But from what I understand their whole Outlook client setup wouldn't even work without it, so desktop clients are just as secure as the smartphone applications when they use the Bridge? [0] https://protonmail.com/support/knowledge-base/imap-smtp-and-pop3-setup/ https://protonmail.com/support/knowledge-base/imap-smtp-and-...
- lorenzhs 8y agoOh, looks like you're right on IMAP. But they still receive mail via SMTP from other servers, there's no way around it if you want to be able to receive email from people not using your service. You need to be confident that their receiving servers encrypt your email without storing an unencrypted copy elsewhere.
- m000 8y agoYes, for not technical users this can be misleading. Protonmail could have been more up-front.
- pmoriarty 8y ago"ProtonMail only claims that your email are unencryptable while you don't access the service." Why would they be unencryptable when ProtonMail controls the javascript that "encrypts" the message in the first place? The main point of Kobeissi's analysis is that ProtonMail could, at their whim, serve whatever javascript they want to the client, including javascript that compromises the encryption or hands over the keys to ProtonMail. If they do this, then it's pretty obvious that the message on ProtonMail's servers would be decryptable by them, no?