What would you do?: Assume all Public Key Exchange models in use failed overnight.

[TheCotMan]

I've wanted to start a kind of periodic "What if <catastrophe> happened?" for a while.
Let's see if you all like the idea.

Today's problem: public key exchange failure.

What if you woke up to find all public key exchange models in use (as used by things like pgp/gpg, and ssh) were found to have been broken in some fundamental way.

Not the keys, but the present model for which the present algorithms

What steps would you take, if any?

[The Dark Tangent]

Re: What would you do?: Assume all Public Key Exchange models in use failed overnight

I've wanted to start a kind of periodic "What if <catastrophe> happened?" for a while.
Let's see if you all like the idea.

Today's problem: public key exchange failure.

What if you woke up to find all public key exchange models in use (as used by things like pgp/gpg, and ssh) were found to have been broken in some fundamental way.

Not the keys, but the present model for which the present algorithms

What steps would you take, if any?

If RSA and DSS were broken there would be more problems in the world than just how to pick up mail, but..

IPSec for my VPNs would be fine, they use pre-shared secrets for phase 1 instead of IKE I believe...

SSH would be tricky, and passwords instead of certificates would only help half way.. might have to switch to IPSec tunnels.

Picking up email would suck too, no more pop tls or https to web forums. APOP, while unencrypted, would be the only fall back.. oh oh, I know! kpop.

Hmm.. must ponder.

[TheCotMan]

Re: What would you do?: Assume all Public Key Exchange models in use failed overnight

What about changes to firewall rules, or other restrictions to bandage system by denial of access?

Would you change the way you do things in life? Would you build new keys and share them privately? Pre shared secrets with people you know?

Would you assume ATM and CC transactions were safe from this? Would you stop using ATM or credit cards so much? How would you get your money? Only visit your own bank ATM?

Would you plan for looting in the streets? Prepare for possible violence?

[astcell]

Re: What would you do?: Assume all Public Key Exchange models in use failed overnight

I'm old enough to remember when credit cards not only had no magnetic strip, but the numbers on the cards were TYPED on it, not raised. Yes the world got along fine with couriers, registered mail, pencil and paper. More F2F meets and having to fly to another continent just to utter 5 words. Yup that was my world growing up.

[xor]

Re: What would you do?: Assume all Public Key Exchange models in use failed overnight

The short term would be less of a concern than the long term. If we all started using telnet, ftp, pop, http ...etc for the short term it would take time for m-hackers to adjust as well.

The long term is the real problem, first we would have to come with new crypto algorithms, standards and implementations; which take time understanding their dynamics, proving them out and testing.

There are many good already on the shelf algorithms, standards, that have already been rigorously tested. Switching to their adoption would be time consuming but faster that trying to do it from the ground up.

In fact the best case interim solution would be to move to one of those spec's since most firewalls support many.

xor

[xor]

Re: What would you do?: Assume all Public Key Exchange models in use failed overnight

Just to be clear are you talking about the infrastructure failing? Physical, or the model. If, the model, how is it broken. To say just broke isn't very clear. The people I support tell me their computer broke, it's how and what failed that is the real statement; what are the symptoms doctor cotman. I think you are saying that the keys aren't being recognized, so no exchange is happening, correct.

xor

[TheCotMan]

Re: What would you do?: Assume all Public Key Exchange models in use failed overnight

Just to be clear are you talking about the infrastructure failing? Physical, or the model. If, the model, how is it broken. To say just broke isn't very clear. The people I support tell me their computer broke, it's how and what failed that is the real statement; what are the symptoms doctor cotman. I think you are saying that the keys aren't being recognized, so no exchange is happening, correct.

xor

Well, if the model is broken, then the fundamental basis for the algorithms that implement the model are broken.

As for how broken the model would be, that is a matter of judgment. How broken would something have to be, in order to considered a "catastrophe" ?

(Topic of catastrophic failure in a model would not likely be something that could be repaired quickly or easily.)

catastrophe defined:
1. A sudden and widespread disaster: the catastrophe of war.
[chop]
1. A great, often sudden calamity.
2. A complete failure
[chop]

calamity defined:
1. An event that brings terrible loss, lasting distress, or severe affliction

[xor]

Re: What would you do?: Assume all Public Key Exchange models in use failed overnight

Personally I don't claim to be an expert in this subject or even educated. But since we are just talking I'll give it a try.

Initially and most likely life would go on in an insecure way until new secure models could be developed or existing alternatives could be implemented. Most crypto is so complicated the average person still would not be able to do anything with it. Data would only be insecure/accessible to a few. It would be a tremendous blow and set back to the perception of trust by the people of internet commerce. As reliant as we are on technology and the fact the most people don't secure anything unless it's done for them again life would go on. The internet has too much going for it for a little thing like trust to get in the way. The internet is an unstoppable juggernaut, short of a solar flare, gamma ray burst, an EMP weapon detonated over the US, nuclear holocaust, the Taliban sacking Washington DC, or us exhausting all energy in the world the internet will be with us insecure or secure for the foreseeable future.

There was a guy in the wireless village at defcon that was pushing a box that performed realtime decryption of ssl webpages. Just cause it can be done doesn't mean he has my banking info or is reading my mail.

I personally don't see how this could happen, please elaborate on the possibilities. Though this is an extreme analogy and no doubt a stretch it's sort of like saying what would happen if all the windows boxes one day suddenly didn't boot(wishful thinking). We would all learn MAC OSX, LINUX or FREEBSD.

xor

[TheCotMan]

Re: What would you do?: Assume all Public Key Exchange models in use failed overnight

Personally I don't claim to be an expert in this subject or even educated. But since we are just talking I'll give it a try.

The beauty of this topic, is level of skill is not a requirement for posting a reply. Anyone can provide a list of things they might do differently. :-)

Even a forum member who is a luddite, that uses carrier pigeons to another human as a forum proxy can reply. Maybe they would say, "The only thing that I would do differently is say,'I told you so!nyah nyah!'"

I personally don't see how this could happen, please elaborate on the possibilities. Though this is an extremely analogy and no doubt a stretch it's sort of like saying what would happen if all the windows boxes one day suddenly didn't boot(wishful thinking). We would all learn MAC OSX, LINUX or FREEBSD.

If I provide a specific cause, then there is an attempt to solve the specific problem and not answer the question. (I'd rather not provide any specifics on where the model might fail.) However, you provide an answer to this question by asking this question-- your decision would be to allocate your time and resources to looking at the failure in the models.

Even if a new model was found, which appeared to "solve" the present catastrophe, the implementations using the broken models would still exist in live systems, and perhaps exist in firmware, only upgradeable by physical replacement. The only thing a new model would change would be how long the catastrophe would last.

[renderman]

Re: What would you do?: Assume all Public Key Exchange models in use failed overnight

I assume this topic came about because of the recent announcement of 2 separate groups both making leaps in quantum computing...

I think that what would change depends on the type of person on an individual level, and at what level of society on a mass level

On an individual level you'd find credit cards and most common commerce functions would be suspect and you'd see something akin to the old 'run on the bank'.

Large segments of society would grind to a halt as suddenly large banking transactions cannot be verified nor trusted because of the inevitable number of jackasses who would begin to screw with things or just create enough noise in the system to create chaos.

In a smaller bubble beyond socio-economic meltdown, I would see alot of communication remain the same. Most people recieve email in plain text and few if any average home users use crypto beyond banking/commerce functions.

I think that for those of us who use crypto as much as we can (I like un-necessary crypto) our trust would not be immediately broken however our paranoia and suspicion would increase and new communication would be vetted very closely. Habits would change and less un-necessary communications would occur.

It's an interesting question to think about because it's a very real one that may occur. I personally tend to be a pessimist and assume the worst will occur. Society tends not to deal with paradigm shifts too well and things would get very ugly very quickly.

At least it would'nt be boring.....

[Schuyler]

Re: What would you do?: Assume all Public Key Exchange models in use failed overnight

I'd stop clicking on the red box.

[xor]

Re: What would you do?: Assume all Public Key Exchange models in use failed overnight

A better question to ask would be what would happen if an EMP device were detonated over the US. An attack of our infrastructure and technology not killing people directly mite not merit a direct military response. Can you really justify killing people over the destruction of critical electronic infrastructure.

Imagine one day we are part of a global village; suddenly and without warning we are reduced to getting news from our neighbor. No TV, cars, radios, computers, the US goes dark. All unprotected electronics destroyed.

Could you justify war at this point or would you first have to put the country back together which could take a decade or more?

Even without an attack a solar flare from the sun directed at the planet could be absolutely catastrophic. In fact it has already happened in the past and will happen again.

thx-1138

[Schuyler]

Re: What would you do?: Assume all Public Key Exchange models in use failed overnight

A better question to ask would be what would happen if an EMP device were detonated over the US. An attack of our infrastructure and technology not killing people directly mite not merit a direct military response. Can you really justify killing people over the destruction of critical electronic infrastructure.

I cannot imagine that situation happening without incurring a very specific loss of life in both the immediate and in the following days and weeks. I'd imagine we would be responding as much to the loss of our infrastructure as we would to the loss of life.

As far as going to war at that point, it would depend on how protected our military electronics were. Personally I would have the military on alert and on our shores at that point. National defense first and foremost. I know a lot of people would rather see us marching to war to do something about it, but I think it would be ignorant not to expect more after an event like that. If there was good, verified, actionable intelligence saying "yes, there is more, and it's going to come from X" I'd see if we couldn't get an ally in the area (Say Isreal, if it's the middle east) to launch a preemptive counterstrike to whatever worse was about to come.

[TheCotMan]

Re: What would you do?: Assume all Public Key Exchange models in use failed overnight

I assume this topic came about because of the recent announcement of 2 separate groups both making leaps in quantum computing...

Well, there are 3 or 4 spaces I know about, probably more, that might have a chance in this realm. (None of these are really new, and have been investigated for a long time.) One is a difficult problem in mathematics that has yet to be proven. It is related to an assumption made in math that many think might be true, but has not been proven to be true. Another is quantum computing. However, quantum computing might be the weakest of these, as access to build faster computing system at this level could provide us with fast, reliable quantum cryptography too.

Much of present high-speed crypto (done on computers today) that use Public Key, or Public Key exchange, rely on very basic assumptions. Many non-public key-exchange are not necessarily at risk. For example, a vernam (Simple XOR, OTP, no key reuse, *) cipher with non-reused "keys" exchanged in an implied secure manner allow for what is considered by many to be one of the most "secure" (from attack by only watching the data stream, not physical security, eavesdropping, etc.)

As DT mentioned, cases where there are *only* secret keys (such as pre-shared secret keys in VPNs) are often considered to not be at *new* risk to attacks (specific) to PKI, by matter of definition and key secrecy.

I wanted to have the problem be generic enough, so as to not be limited to just the spaces where fundamental assumptions required for present Public Key systems are being attacked, but also include spaces that I've not considered, or that might develop in the future.

I think that what would change depends on the type of person on an individual level, and at what level of society on a mass level

Right. I was hoping to see an interesting cross section of responses from members here, from "new person" to, "I am a director/manager and would do this," and everyone in between.

On an individual level you'd find credit cards and most common commerce functions would be suspect and you'd see something akin to the old 'run on the bank'.

Right. The banking systems have everything they need to pass information (OTP/XOR, or a private/secret key exchange) from one place to another using physical security. They also have laws, and physical security measures to have physical security in place to decrease risk of loss of private key exchange/storage. It wouldn't be a pretty transition, but they would probably be able to shift away from any present public-key systems faster than other businesses.

In a smaller bubble beyond socio-economic meltdown, I would see alot of communication remain the same. Most people recieve email in plain text and few if any average home users use crypto beyond banking/commerce functions.

Good point.

It's an interesting question to think about because it's a very real one that may occur. I personally tend to be a pessimist and assume the worst will occur. Society tends not to deal with paradigm shifts too well and things would get very ugly very quickly.

At least it would'nt be boring.....

Heh heh. No. No it wouldn't. It would be very exciting, actually. If someone posted, "What would I do differently? I would buy a new camera," then we would have a new undercover reporter spotted.

I'd stop clicking on the red box.

Hahahahahah! Clever. :-)

[xor]

Re: What would you do?: Assume all Public Key Exchange models in use failed overnight

I cannot imagine that situation happening without incurring a very specific loss of life in both the immediate and in the following days and weeks. I'd imagine we would be responding as much to the loss of our infrastructure as we would to the loss of life.

As far as going to war at that point, it would depend on how protected our military electronics were. Personally I would have the military on alert and on our shores at that point. National defense first and foremost. I know a lot of people would rather see us marching to war to do something about it, but I think it would be ignorant not to expect more after an event like that. If there was good, verified, actionable intelligence saying "yes, there is more, and it's going to come from X" I'd see if we couldn't get an ally in the area (Say Isreal, if it's the middle east) to launch a preemptive counterstrike to whatever worse was about to come.

Actually since we moved away from the cold/possible nuclear war stance a lot of military electronics don't contain the $5.00 transorbs. I think things would be so messed up after an attack like the one mentioned above that we would only be able to mount something akin to the Doolittle Raid.

thx-1138