New Global Directory of OpenPGP Keys

Gemini writes "The PGP company just announced a new type of keyserver for all your OpenPGP keys. This server verifies (via mailback verification, like mailing lists) that the email address on the key actually reaches someone. Dead keys age off the server, and you can even remove keys if you forget the passphrase. In a classy move, they've included support for those parts of the OpenPGP standard that PGP doesn't use, but GnuPG does."

Comments

whitelists?

[essreenim]

Sounds like a good way to make a global whitelist!
Allow incomming mail only from such valid e-mail accounts that are using the service. Could be useful for spam. Or will spam endure as it always has done... ;/

Re:whitelists?

[wwest4]

> Or will spam endure as it always has done... ;/

Or only allow incoming mail that's signed. This won't prevent spam, but it will complicate the spammers' lives a bit, at least for a while.

Re:whitelists?

[Tenebrious1]

Sounds like a good way to make a global whitelist!

It won't be any different from individuals creating their own whitelist, since you can't implement whitelists at the ISP level since most people do not use PGP and cannot be forced to use it.

It wouldn't stop spammers at all though, since spammers could still create legitimate keys, send out a billion spam then delete those email accounts and move on. It may slow it down a bit until some smart spammer creats a program to automate the process of creating, registering, and authenticating the key, but I doubt it will take too much time and effort.

Backdoors?

[gmknobl]

Are there backdoors? And if there are not, what will Homeland Security or the like try to do about it?

Can they do anything about it, realistically?

Have I completely misunderstood this (a common event, unfortunately) or will this be one of the few ways of having as close to true privacy as we can realistically get?

Re:Backdoors?

[rdieter]

Doesn't matter. This is a directory for public (ie, the non-private portion of) OpenPGP keys, which are/should-be publically available anyway. Else, why use public/private pgp keys at all?

Re:Backdoors?

[Anonymous Coward]

The problem is that you need to make sure the public key belongs to the recipient/sender you are communicating with. Anyone who can intercept traffic to/from this server can put himself in the middle of your supposedly private conversation. The web of trust is a way of eliminating/reducing this threat, but that means people have to actually go out and have their keys signed in real life. Encryption with authentication is useless.

Re:Backdoors?

[JimDabell]

Are there backdoors?

It doesn't matter. Keyservers are merely a method of distributing keys, not establishing trust. You can establish trust by a number of methods, such as manually verifying the fingerprint with the person yourself using a trusted medium (e.g. face to face) or having somebody you trust sign the key (after verifying their key, of course).

The real danger to public key cryptography taking off is that it will become commonplace to simply trust keys without verifying them. Everyone will feel more secure, but the security will be an illusion.

Re:Backdoors?

[JimDabell]

It matters a lot if, let's say, you encrypt a sensitive email with a fake public key not belonging to the person you think it is.

No, it doesn't matter in the slightest how you got the key. PGP operates under the assumption that it's not practical to always use a trusted medium to exchange keys. It doesn't trust keys by default.

PGP uses the concept of a "web of trust" to decide whether you should trust a key or not. If you can securely verify the legitimacy of a public key, then you can sign it, so that people who trust your judgement can also trust the key. In reverse, you can trust keys that people you trust have signed.

How the keys are transferred is completely irrelevent to this mechanism. You could download a public key from Gnutella or Usenet, and as long as it's been signed by somebody you trust, or you can verify the fingerprint over a secure medium, it's trustable.

So, your scenario would play out as follows:

1. Download "trojan" public key.
2. Not signed by somebody you trust? Throw the key away.
3. Signed by somebody you trust, but the signature is invalid? Throw the key away.
4. Signed by somebody you trust, and the signature is valid? The key is trustworthy.

The balance between how practical and how secure your web of trust is depends on how much trust you place in others. It doesn't depend on the medium you transfer keys under in the slightest. That is why it doesn't matter if there are backdoors in the keyserver. No amount of tampering with it could make the web of trust any less secure.

If you think about your line of reasoning, if what you said were true, PGP would be pretty damn insecure to begin with, as you'd necessarily be trusting an external entity (the PGP keyserver admins) with all your communications.

PGP's defaults are the real problem.

[nlinecomputers]

Every PGP new user has done it. Created a brand new key while learning the program and forgot the passphrase. There are hundreds of unused keys that was created and never used but can never be deleted because they don't expire.

Had PGP's defaults been for a 1 year key instead of infinite this wouldn't be an issue.

I always create 1 year keys but I've got a couple of key out there over 10 years old that I FUBAR'd that'll never go away.

Re:Widespread Crypto Revolution?

[Luigi30]

Yes... until some government makes encryption illegal because it evades wiretaps (they're trying, believe me...).

Re:Is there a future for PGP?

[Ashe Tyrael]

There is a problem with this though. Several ISPs, for good and legitimate reasons (spam and virii) don't allow certain types of e-mail attachment. Which means if I sign an e-mail, the fact I've signed it gets filtered by the receiving ISP.

Nothing wrong with the standard itself, just a lack of support and clue by ISPs.

Oh great, spammer heaven

[phr1]

Fantastic, a global database of cryptographically authenticated email addresses that have been tested to reach a real person.

We need a new key format, that doesn't have a live email address but instead has a hash of one. You'd send the address separately so it could be compared against the hash. There'd be salting to stop brute force searches. The database server could then still verify all the addresses (by sending emails out) but the actual email addresses would stay unpublished.

Re:Is there a future for PGP?

[jludwig]

Its missing what I call the "grandmother" factor. I can explain it to most technical people I encounter (but can't convince any to use it), but its way too complex an implementation for most average users to handle - my mother or grandmother. Its not that they can't understand it, but the computer is already overwhelming and they need something that "just works(tm)". The Web of trust concept "just makes my head want to explode(tm)"

Unfortunately I can't see a good way to make things more transparent and invisible to the end user. Most folks don't pick good passwords, yet that is absolutely essential for PGP private key security. Also, a yearly drive reformat is not uncommon, so lost keys are a huge issue. This technology partially address that issue but I shouldn't need to check to see if someone updated there key every message, plus theres the trust issue with a constantly rotating keyset.

Jeff

Re:Centralization ??

[jimbro2k]

Good point, but this just provides a central option . You can still do a private(?) exchange of public keys with your friends & not friends, or do both..

Can a central repository bring security?

[cesarbremer]

A central repository of public keys can bring problems, for example, if the central repository is located in USA and the FBI want to do a man-in-the-middle attack? How can you be assured that the public key from the guy you want to send a encrypted message is realy the correct public key? I don't know better solution than having a lot of servers in different countries, under different governments controls and laws, and when the user do a search, he can do the search in a lot of servers. How about having servers in USA, China, France, Germany, China, Finland, North Corea......, and the user can search the user public key in all these databases? When storing the public keys, why not the user store his keys in these distributed servers? Can you really believe that storing your keys under one company control can bring security?

Re:Encrypted Spam?

[I confirm I'm not a]

So if I'm willing to post my public key and verify every 6 months that I'm the same live email responder at the other end, then what assurance do I have that encrypted email sent to me isn't spam?

Another way of looking at it is from the "cost" of spamming - encrypting a spam "costs" the spammer, hence recent suggestions for charging mail-senders in CPU-cycles. Additionally, you'd be able to verify whether you held the spammer's public key on your keyring, and very easily "process" (ie. delete with extreme prejudice) encrypted emails from unknown senders.

A Big Step...

[shaneh0]

Perceived Value is very closely tied to percieved scarcity. As people begin to *realize* that their privacy is as scarce as it actually is, people will begin to value their privacy ergo encryption.

Feeding that will be dirt simple encryption applications that make it so EASY to encrypt and decrypt that you might as well do it. (Like, for example, the application I'm finishing right now but refuse to plug until it's released)

The biggest problem now is that if a developer wants to include Public Key encryption abilities in has app he has to create an entire key management system and force users to gather the keys of all their contacts manually because there's just no other way. How many users are going to do that for a program that they only kinda think they need?

If you want the answer to that question, look at the percentage of users who currently encrypt any large part of their communication (SSL excluded?)

Re:Encrypted Spam?

[Frank T. Lofaro Jr.]

Spammers won't sent you encrypted mail.

It is way too computationally expensive.

Spam programs are designed to work extremely fast, using very little CPU to send a message.

That is why things like hashcash would work, they'd make it economically unfeasible for spammers.

Encryption takes quite a bit of work (just less than unauthorized decryption :)

Re:Encrypted Spam?

[TheLoneCabbage]

Asymetricly encrypted emails are rarely actually encrypted. They are signed. which is that I mearly provide an encrypted hash of the email, to prove that whoever sent it, has access to the private key.

The keys themselves can be signed by a master key, by o' say PGP's new website. (this does not require the PGP website to have a copy of the private key)

What this meens is they could give the signing service away for free to individuals, in order to create a defacto standard. But then charge legitimate bulk emailers for the privlege of their service. PGP becomes the arbiter of who is spam and who is not. In exchange they get to charge for permission to send bulk/commercial mail.

Sounds like a good buisness plan.

Of course, I'll have to RTFA once the /.'ing stops.

One word

[lildogie]

> Is there any way to acutally prove that a message is encrypted,
> as opposed to being just random garbage data that two people
> happened to mail to each other?

Torture.

Re:Is there a future for PGP?

[Greyfox]

If companies would sign their corrispondance with a PGP key, it could eliminate (Or at least siginificantly reduce) phishing. More so if common mail clients were to support PGP and PGP signatures better.

Re:FPCP

[Anonymous Coward]

So, some poor bastard gets spammed by your C-R system, because don't want to deal with your own spam. He does what C-R systems expect humans to do, and replies to confirm that he's real, thus letting spam into your mailbox. Hell, he may even have automated that process for common C-R systems, to deal with idiots who turn on C-R and don't understand what they're doing, thus never seeing your challenge; after all, when he got hit by this joe job, he had better things to do than read each C-R in turn and only answer the genuine ones, not the spam induced ones.

And as a result of him doing WHAT YOU ASKED HIM TO, and thus causing you to see ONE piece of spam, you feel entitled to let him in for huge amounts of the crap? Maybe he should be entitled to take $100 from you for each challenge you send him. It would at least give him an incentive not to answer your challenges unless they're replies to messages he's sent, and it's a damned sight easier to cope with losing the odd $100 than to get yourself off huge numbers of mailing lists.

** APPLAUSE **

[Anonymous Coward]

Well said. Anyone who thinks a C-R system is a good idea simply doesn't understand what they are doing. I also do what the GP does - respond to C-Rs that I get due to joe-jobbing or the virus du jour.

And in case any C-R users wish to respond, here in a nutshell is why C-R is explicitly worse than useless : You receive a bunch of mail. Some of it may be whitelisted, some of it may be blacklisted. Some of it may be rejected outright due to eg SpamAssassin. Some of it may not be accepted in the first place due to RBLs. Whatever, at the end of all that, you have a body of messages for which you have to decide what to do. Instead of just facing up to that burden and delivering the message (or not), the C-R user passes that burden back to the purported sender. Most all of the time this is an innocent third party. So a C-R user's burden may go down, but only at the expense of the wider net community. It's ignorant and wasteful, and is little different than the modus operandus of spammers : let other people bear the cost of my own selfish actions.

If you're using a C-R system you are hardly any better than a spammer.

Re:This presents problems with the trust path.

[Anonymous Coward]

Not true, it is imperative that it be possible to revoke trust in keys. When an imposter is detected, the key needs to be both removed as well as any trusts they created to other keys. Not only that, but you don't want ancient trusts lying around forever, trust is a dynamic thing, and trust networks change. Forgetting old trust only means that it must be re-established every few years. That's a good thing.