Could you link some good explanation? My understanding is that Grover<span style=“font-size:4px”>not Furr</span>'s algorithm basically turns AES-256 into AES-128
Personally I am still paranoid tho about what might happen in the future.
I think it’s reasonable to be in some cases where you are sharing sensitive data that you can’t afford to possibly be broken by governments or whoever in a decade or more from now.
The thing to do with truly important data is to cascade algorithms. That is encrypt it using multiple algorithms so a failure in one in a cryptographic sense means they still need a failure in another and if the combined failures cannot shave off enough bits they still can’t get the data. AES is fairly strong and proven so I would use it as one of those encryption schemes. Ideally you’d do this with ciphers from multiple mutually hostile governments (one from Russia, one from US/NATO).
Depends on what schemes you mean. Stuff like RSA which relies on unsolveable math problems, yes that’s fucked so in theory a lot of HTTPS web traffic will in future be deciphered by the NSA if they’ve been storing it (they have been for a while now). But things like AES-256 as someone mentioned, certain other schemes should be quantum safe.
Put it this way, a good strong password on AES-256 encryption can put you at a cracking time of hundreds of millions of years. Shaving off even 95% of that time with quantum computing wouldn’t be helpful because you’d still be looking at millions of years of cracking effort. In practice anything that puts cracking time beyond two decades or so of effort is something that protects the data because nothing but a formula for time travel is going to be of that much value to governments to spend such an amount of time dedicating a massive array just to cracking it. In practice LEO will usually give up on cracking something after perhaps 12-16 months of efforts and declare it failed. NSA and so on have more or less given up on cracking actually strong encryption so much as trying to subvert and weaken implementations with bad math and much more broadly just moving to straight up hacking victims and putting malware to grab the data before its encrypted.
AES has certain problems with it, there are methods for shaving off a certain amount of bits but even these when considered against a strong implementation (in terms of passwords we’re talking >=20 characters, upper/lower case, digits, symbols) don’t really put it within the realm of truly broken in a reasonable time frame as I mentioned above.
Are they near breaking encryption yet?
I’m far from an expert, but apparently some symmetric algorithms like AES-256 are considered “quantum-safe”
I’ve done a little bit of reading. I don’t trust AES-256 in the long term, they need to upgrade the standards.
Could you link some good explanation? My understanding is that Grover<span style=“font-size:4px”>not Furr</span>'s algorithm basically turns AES-256 into AES-128
Well reading this it tells that the key is very hard to guess, and the algorithm is solid enough that the key cannot be guessed by looking at the data… https://www.n-able.com/blog/aes-256-encryption-algorithm
And the government and institutions use it so… https://www.clickssl.net/blog/256-bit-encryption
Personally I am still paranoid tho about what might happen in the future. I think it’s reasonable to be in some cases where you are sharing sensitive data that you can’t afford to possibly be broken by governments or whoever in a decade or more from now.
The thing to do with truly important data is to cascade algorithms. That is encrypt it using multiple algorithms so a failure in one in a cryptographic sense means they still need a failure in another and if the combined failures cannot shave off enough bits they still can’t get the data. AES is fairly strong and proven so I would use it as one of those encryption schemes. Ideally you’d do this with ciphers from multiple mutually hostile governments (one from Russia, one from US/NATO).
nah nah nah just roll your own crypto
(Only half joking: worthless in a targeted attack but effective over unmanned dragnet)
Yea I figure if you have anything truly important or prone to being targeted then… It doesn’t cost you much to over do it compared to not.
Depends on what schemes you mean. Stuff like RSA which relies on unsolveable math problems, yes that’s fucked so in theory a lot of HTTPS web traffic will in future be deciphered by the NSA if they’ve been storing it (they have been for a while now). But things like AES-256 as someone mentioned, certain other schemes should be quantum safe.
Put it this way, a good strong password on AES-256 encryption can put you at a cracking time of hundreds of millions of years. Shaving off even 95% of that time with quantum computing wouldn’t be helpful because you’d still be looking at millions of years of cracking effort. In practice anything that puts cracking time beyond two decades or so of effort is something that protects the data because nothing but a formula for time travel is going to be of that much value to governments to spend such an amount of time dedicating a massive array just to cracking it. In practice LEO will usually give up on cracking something after perhaps 12-16 months of efforts and declare it failed. NSA and so on have more or less given up on cracking actually strong encryption so much as trying to subvert and weaken implementations with bad math and much more broadly just moving to straight up hacking victims and putting malware to grab the data before its encrypted.
AES has certain problems with it, there are methods for shaving off a certain amount of bits but even these when considered against a strong implementation (in terms of passwords we’re talking >=20 characters, upper/lower case, digits, symbols) don’t really put it within the realm of truly broken in a reasonable time frame as I mentioned above.