Can you elaborate more? In what sense? They are certainly vulnerable if you compromise the system which has them saved. Are they vulnerable to a brute force breach? Likely not, brute force is always a slow process, no matter how you speed it up. Is X vulnerable should always be phrased as Is X vulnerable to Y.
I'm not sure I follow. A brute force attack generates a key which can then be tested against the target. This isn't magic, in fact, the more detailed answer above made the same argument essentially. Thought they provide a different argument suggesting that it may be possible to increase the ability to guess keys correctly.
I would think that if the rate of guesses is only one per second it would take eons to find the correct one. I was just wondering if there is an offline method. Just wondering how safe we can be when the news comes out.
I assume you are correct. I also assume you and I are somewhat safe. You have to consider the likely hood of an individual getting a hold of one of these devices in the near future. Then again you really have to ask if there is any reason you in particular will be targeted by someone with this technology.
Eventually everyone will be targeted. I have an unannounced web server serving a domain in the middle of internet nowhere and I get script attacks all the time. If some how this gets commercialized with no regulation criminals will open up shop for business practically the day after. Its like CRISPR: sure it can do some powerful things but so can your adversaries to you. I hate to think of the internet as a cloud of adversaries but that is really what it has become.
CRISPR fears are really overrated. It is nowhere near as mature as it needs to be. Further, you're less likely to make anything useful with it and more likely to just kill the organism you are trying to create. I work in a biotech lab, and we hardly ever use it and instead opt for manual plasmid knock in/down or tissue engineering techniques instead. The extent of work you have to put in to actually get something useful from CRISPER is prohibitive of it being used to target you and I.
If your ssh key relies on RSA, then yes, you're relying on the difficulty of factoring numbers; which Shor's algorithm can perform efficiently. But only if the quantum computer has enough qubits. The needed number of qubits needed to factor is a multiple of the rsa key length. Unlike a regular computer, where you can finagle two computers with 2 gigs of memory to work like a single machine wiht 4 gigabytes of memory, so you can't just merge two 1024 qubit computers to get a 2048 qubit computer. And we're a long way off to having anything close to that.
The simplest solution is simply to increase the size of the key beyond that of the most advanced quantum computer (which is essentially today's approach ;) ). Also, you just need to exchange a symmetric key; Shor's algorithm is a special case, you don't get the exponential -> polynomial speedup[1] to encryption algorithms in general. So once you deliver the symmetric key in a secure way, using a quantum resistant algorithm, you can resume business as usual.
Though the question is how relevant this is. Likely SSH will have moved to something more secure years before any quantum computer with sufficient computing power will become available.
Because that's the thing these scare-stories don't tell you: While the progress in quantum computing is moving very slowly, in recent years the progress in post quantum cryptography has been very fast.
Two examples off the top of my head:
1. Instantaneous stock transactions from anywhere in the world. (quantum communication/entanglement) Offers a substantial RTT improvement on current trading systems.
2. High frequency trading based on quantum prediction algorithms. (Quantum information theory) May provide for HFT algorithms capable of producing predictions based on an exponential number of information streams. E.g. imagine making future investment decisions based on _all_ trades that happen in a span of time.
@ me if you decide make a company to do either of these things.
But it only reduces the effective key length by 2. Which, although amazing, has a fairly straightforward response. Double your key length beyond what was previously viewed as 'intractable'.
They break all widely used public key encryption systems (RSA, DH, ECC-based). There are post-quantum systems, but none of them are in widespread use yet.
Symmetric encryption systems are weakened, but can easily be made quantum safe by increasing the key size.
Unlike nuclear weapons, a quantum computer is only dangerous if people don't know you have one.
There are a number of ways of dealing with the security issues posed by quantum computers; most of them involve simply increasing the key length of your encryption keys.
However, if people don't think you have a quantum computer, or you're able to make it better faster than expected, they won't change their behavior and so you have a fixed target.
Second sentence:
"Now that the technology is moving from research to practice"
No such thing is happening. We're still in the state of "we're waiting for someone to actually show that quantum computers can do anything at all faster than a classical computer". Yeah, it's probably gonna happen in the not so far future, yeah, there's progress. But saying anything is moving "to practice" is absurd at this point.
It's all fluff, vague nonsense about what quantum computers might supposedly do, and mostly wrong. The implied assumption is that a quantum computer can solve any NP problem in polynomial time, and that doesn't seem likely. And then, all of a sudden, the purpose of this nonsense becomes clear:
> For its part, ISARA has developed a new method of creating a compound or hybrid digital certificate
This is followed by lots of marketing buzzwords. Disgusting, really.
33 comments
[ 3.9 ms ] story [ 44.5 ms ] threadThe simplest solution is simply to increase the size of the key beyond that of the most advanced quantum computer (which is essentially today's approach ;) ). Also, you just need to exchange a symmetric key; Shor's algorithm is a special case, you don't get the exponential -> polynomial speedup[1] to encryption algorithms in general. So once you deliver the symmetric key in a secure way, using a quantum resistant algorithm, you can resume business as usual.
[1] Though in brute force searches, you can reduce the effective key length by 2. https://en.wikipedia.org/wiki/Grover%27s_algorithm
Though the question is how relevant this is. Likely SSH will have moved to something more secure years before any quantum computer with sufficient computing power will become available.
Because that's the thing these scare-stories don't tell you: While the progress in quantum computing is moving very slowly, in recent years the progress in post quantum cryptography has been very fast.
2. High frequency trading based on quantum prediction algorithms. (Quantum information theory) May provide for HFT algorithms capable of producing predictions based on an exponential number of information streams. E.g. imagine making future investment decisions based on _all_ trades that happen in a span of time.
@ me if you decide make a company to do either of these things.
...???
However, it can reduce the difficulty of a generic brute force search
https://en.wikipedia.org/wiki/Grover%27s_algorithm
But it only reduces the effective key length by 2. Which, although amazing, has a fairly straightforward response. Double your key length beyond what was previously viewed as 'intractable'.
Symmetric encryption systems are weakened, but can easily be made quantum safe by increasing the key size.
You can't remotely disarm nuclear weapons or protect an entire industry by switching algorithms.
They are a problem, and we should prepare for them, but they aren't apocalyptic.
There are a number of ways of dealing with the security issues posed by quantum computers; most of them involve simply increasing the key length of your encryption keys.
However, if people don't think you have a quantum computer, or you're able to make it better faster than expected, they won't change their behavior and so you have a fixed target.
No such thing is happening. We're still in the state of "we're waiting for someone to actually show that quantum computers can do anything at all faster than a classical computer". Yeah, it's probably gonna happen in the not so far future, yeah, there's progress. But saying anything is moving "to practice" is absurd at this point.
It's all fluff, vague nonsense about what quantum computers might supposedly do, and mostly wrong. The implied assumption is that a quantum computer can solve any NP problem in polynomial time, and that doesn't seem likely. And then, all of a sudden, the purpose of this nonsense becomes clear:
> For its part, ISARA has developed a new method of creating a compound or hybrid digital certificate
This is followed by lots of marketing buzzwords. Disgusting, really.