Take the 2-minute tour ×
Information Security Stack Exchange is a question and answer site for Information security professionals. It's 100% free, no registration required.

After the announce that NSA conducts extensive metadata and traffic search on e-mail, it is clear that people looking for privacy and security would better abandon the system completely, since, by design, this kind of snooping cannot be avoided.

However, I do not know of any system that is meant to replace e-mail as a whole with a verifiably secure and private protocol. There are instant messaging systems and social networks who try to be secure by design, but I as far as I know no protocol that is meant to take the place of e-mail in everyday communication. PGP on top of traditional SMTP mail is being suggested as an alternative, but it is not immune to traffic analysis.

So, the question is: has anyone designed a secure protocol to replace e-mail with similar functionality? Is there a live project around that I can start using and supporting?

share|improve this question
    
Security must be defined in terms of Confidentiality, Integrity, and Availability. E-mail is a method of textual communication which make transit across a public network of networks. Do you wish confidentiality of sender identity, integrity of message, restriction of access of message content to a specified recipient? Confidentiality of recipient is beyond technical application under the current concept of e-mail. –  this.josh Aug 12 '13 at 6:26
    
Sorry for my poor knowledge of the technical terms. Integrity and restriction of access are easy to throw on top of any infrastructure with signatures and encryption, so they shouldn't be a challenge. Confidentiality of receiver is not possible within the current e-mail infrastructure, but since we are speaking about a new protocol designed from scratch, I don't see why we should bar it as impossible. –  Federico Poloni Aug 12 '13 at 7:45
add comment

8 Answers

If you ever look or read about the old spy tactics, they would pass messages in soda cans under a park bench, or do things that couldn't easily be found. From a traffic analysis and meta-data collection issue, encryption of the messages do not matter since the analysis is of the signals not the content. Signals analysis may look at things like between what two parties, frequency, etc.

One possibility may be using something like a TOR-like infrastructure, where the traffic is difficult to decipher without controlling a lot of nodes, though as recent news has shown, TOR is not a full proof way to get anonymity and can be compromised.

There are also anonymous BBS and BBS-type systems. You post the message encrypted and then someone else has to know where to go to find it an decrypt it. If you upload and download a lot, its difficult to tell the connections between individuals, making traffic analysis hard.

Focusing on meta-data or signals analysis part, you can always look at adding noise, such as the Chaffing and Winnowing system proposed by Ronald Rivest. You could also try to communicate over covert channels, split messages, etc. You could move around between various forums every few days, even if they are off topic (and thus less likely to be targeted in a search). You could communicate in the comments of random websites and have some way of telling your receiver how to find the comments. You could also implement stenography and post things in very public places.

In general, to make it easy and secure would be real challenge. If all traffic can be monitored at the wire level, then traffic analysis is always possible without sufficient noise or physically moving between networks (go to a new coffee shop every few minutes).

If you are not concerned about the meta data and the signals, but want privacy, any type of encryption where the service provider does not hold the keys will work. Ekiga and other tools provide encryption, and are relatively user friendly. Even outside of security, there are always arguments for replacing email, its an issue of what features are needed by the masses and how to get them to transition. The Internet is full of legacy hold backs which take a long time to be replaced (even WWW-browsing has lots of limitations and draw backs).

--Edit--

Just saw this today on HN: Secure Share. Some of their points:

We employ GNUnet for peer-to-peer routing and encryption (the new openssl of P2P) and PSYC to create the social trust graph (because it performs a dozen times better than XMPP or OStatus).

communication between parties cannot be measured as they may have none to several routing hops in-between. an observer never knows if a communication came where it came from and ends where it is going to.

May end up being something useful, if you could get people to agree to leave the web browser.

share|improve this answer
    
Secure share looks great, but it seems to be more focused on realtime Skype-like messaging rather than permanent asynchronous e-mail. –  Federico Poloni Aug 12 '13 at 7:48
add comment

It may be MUCH simpler to implement "(secure) email over (insecure) email".

  • PGP may provide end user to end user encryption.
  • Custom relay addresses (secure-relay@example.net) may provide a way to protect against "RCPT TO" analysis (also over encrypted connections) and message size analysis (adding random junk, random transission delays, batching messages). e.g. user@example.net -> secure-relay@example.net ->secure-relay@example.org -> user@example.org
share|improve this answer
    
Can you expand your answer a little more? –  Eric G Aug 11 '13 at 5:45
    
Which part? IMHO it can be expanded a lot but expanding it a little is not cost effective. It is more a hint than a complete answer. –  Andrzej A. Filip Aug 11 '13 at 8:04
add comment

E-mail is kind of a communication protocol, and there is nothing wrong with it (at least not from the perspective that you are writing about). The problem is that there is no de facto standard for encrypting it although there are some solutions like S/MIME or PGP but they are too complex for an average user.

The most important part in this game are the users who should have some level of understanding of the e-mail infrastructure and cryptography in order to communicate securely. To solve this problem we should develop user-friendly solutions with end-to-end cryptography. A promising initiative is the MailPile project:

http://www.indiegogo.com/projects/mailpile-taking-e-mail-back

But please note that even if we encrypt all our e-mails (or other communication) with strong crypto on our own equipment and use techniques to anonimize our traffic, the government will still have the ability to hack our workstations just as we saw in the case of Tor.

Since you asked about everyday communication I don't think we will ever have a system that hides traffic metadata simply because vast majority of the users just don't care: For example the fact that I communicate with my friends and business partners at certain times only holds that much information that I already willingly shared on social media.

However, OPSEC and anonymizer technologies like mixers (which can be applied as an extension to the existing e-mail infrastructure) can improve the situation. Tor can also be used as an underlying layer for e-mail (+ TLS wrapped protocols, and PGP or S/MIME).

PS.: Actually, Google Wave came to my mind as a supposed e-mail alternative that would work on private servers using HTTPS, but I don't really think that it will have much impact...

share|improve this answer
1  
OP noted that encryption is the problem, but the meta data / traffic analysis / social network problem is not alleviated by encryption alone. –  Eric G Aug 10 '13 at 19:10
    
Is it possible to route PGP-encrypted messages over the Tor network? (Take, for example, two individuals running TAILS, Thunderbird + Enigmail and the email provider of their choosing.) I feel that this would address both privacy and anonymity. –  nitrl Aug 10 '13 at 23:42
    
@nitrl Yes, it is possible, added to the edit. –  buherator Aug 11 '13 at 9:25
add comment

There is a solution called valeso.com That is super-safe, with individual settings like 'destroy after reading', etc. It comes with a pretty nifty e-mail client and will satisfy most if not all of the secure e-mail needs one might have. Downside of valeso.com is, the server is in USA, the country with an on/off switch for civil rights. If you are not using encryption to prepare/commit a crime, you have nothing to worry about, as long as the rest of the legal system in the U.S. remains functional.

I believe, however, that most people's privacy concerns are simply, that they would not like to send bank-mail, lawyer-mail, love-letters or other personal secrets as postcards, so that every postman can read it. I believe GNUPG.org has meanwhile simplified pgp encryption to the level, that almost everyone can handle. For iPhone there is secumail, it can handle PGP messages. PGP should become a default feature of all e-mail clients.

Downside to all and any encryption is, if your smartphone or pad is not be able to handle and will show a lot of e-mails as encrypted gibberish.

share|improve this answer
add comment

Bitmessage would be a good solution as it does not leak metadata. Of course the people you are emailing would need to be using Bitmessage.

share|improve this answer
add comment

A lot depends on the frequency and volume of communications. If I wanted to pass a few discreet messages, I would post an ad on Craigslist.

share|improve this answer
add comment

You may use Openpgp but if you are not an expert I suggest Mail1Click.

Mail1Click is a quite famous secure email provider, they are using tls and RSA Asymmetric algorithm. Their servers aren't In USA, the company is from UAE and they have a web interface or the app (windows, Android, etc..).

share|improve this answer
add comment

IPV6 holds the answer for 'direct B to B window writes' as opposed to all that is involved and 'at risk' within the current IPv4 IP/DNS/protocol stack.. its the mechanism of email transportation itself that's the problem!! We need a paradigm shift in cognition to eliminate E to E mail altogether and out of the cyclic trap of staged transportation... as opposed to just writing it upon the destination directly. Just stop and think for a moment how obvious and simple that alternative really is given the complication that we currently suffer on account of the system as it is right now... Most every modern comms devices is capable of secure windowing, only then can transportation vulnerabilities be circumvented... thus if you wana hack me; then your gonna have to hack my device and whatever protection I've put in place, so now just stop and think again about the decentralization of messaging and the ramifications thereof... Steveo Reedo nextfriend@live.co.uk

share|improve this answer
add comment

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.