Has the NSA already broken bitcoin?

[Cryddit]

That's because you're only looking at the NSA, and the people who don't see anything wrong are looking at what the code actually does.

Yes, the NSA is a bad actor in several senses.  But you're talking about something which works in a known way that we can see and analyze.  What you're doing is sort of like claiming that there must be a hole in a steel pot because the guy who made the pot is an unethical businessman.  Well, unethical he might be.  He may have made some pots with holes in them.  But it does not mean that every pot he ever made has a hole in it.

People can still look at the pot - inspect it carefully even, test it as a pressure vessel - and find that it does not leak.  Similarly, scores of professional cryptographers and math Ph.D's have analyzed every aspect of SHA2 looking for ways to attack it, and found no leaks. 

If we could not see how it worked, or we didn't know how to search for problems, etc, then you'd have a point, but you'd also be meeting with no disagreement.  People would immediately and rightfully reject anything whose workings and structure they could not verify and analyze, notwithstanding whether they trust or do not trust its source.

[marcus_of_augustus]

That's because you're only looking at the NSA, and the people who don't see anything wrong are looking at what the code actually does.

Yes, the NSA is a bad actor in several senses.  But you're talking about something which works in a known way that we can see and analyze.  What you're doing is sort of like claiming that there must be a hole in a steel pot because the guy who made the pot is an unethical businessman.  Well, unethical he might be.  He may have made some pots with holes in them.  But it does not mean that every pot he ever made has a hole in it.

People can still look at the pot - inspect it carefully even, test it as a pressure vessel - and find that it does not leak.  Similarly, scores of professional cryptographers and math Ph.D's have analyzed every aspect of SHA2 looking for ways to attack it, and found no leaks. 

If we could not see how it worked, or we didn't know how to search for problems, etc, then you'd have a point, but you'd also be meeting with no disagreement.  People would immediately and rightfully reject anything whose workings and structure they could not verify and analyze, notwithstanding whether they trust or do not trust its source.

Not only that, he hasn't described any conceivable method that a "back-doored" sha256 could be used to attack bitcoin in such a way as to cause damage or loss before it could be replaced if necessary.

Total troll, with repeating the same question ad-nauseum, ad-hominems and using willful ignorance to avoid specifics. No-name, throwaway newbie account. Yawn.

[hhanh00]

It boggles my mind how, after Snowden and all the other information that has come out lately, someone can be called a troll for saying they do not trust the integrity of NSA crypto.

What would the NSA have to do to lose support from you people?

And if sha2 turns out to be some NSA sneakiness will you say "wow, totally shocked"?

No crypto is trusted whether it comes from the NSA or not. Let's say SHA-3 gets preferred treatment because it was not from the NSA. Well, who can be sure that they were not involved at all? With the spy stuff going on, it's better to stick to the math.

[RodeoX]

This is all BS.

[cypherdoc]

no ice:

i'll tell you what's more likely given all the evidence we've seen over the years re: crypto, the NSA, and Bitcoin.

it's more likely you're a gvt/bank hired troll assigned to come here and inject fear into the Bitcoin community and to try and scare off any new adopters.  i say that b/c everything you've said is hearsay and is based on lack of evidence.  you appear not to understand the first thing about cryptography.  it's pure fear mongering.

[spazzdla]

stop feeding the troll

It boggles my mind how, after Snowden and all the other information that has come out lately, someone can be called a troll for saying they do not trust the integrity of NSA crypto.

What would the NSA have to do to lose support from you people?

And if sha2 turns out to be some NSA sneakiness will you say "wow, totally shocked"?

You are a troll or WIDLY over estimate how smart you are.. sorry to break it too you.  Like widly over estimate..

Do you even know what a cipher is ?  DO you know what encryption even does?


It is like trying to explain the Imaginary number to a 7yo that just learned how to add.

[justusranvier]

There's an industry devoted to building machine that break double SHA256 as rapidly and efficiently as possible.

If there's an explotable flaw in SHA256, they'll eventually find it and incorporate it into their products.

Then all the manufactures will copy the technique and the network difficulty will adjust upward to compensate for the attack and things will be right back normal, giving us plenty of time to upgrade the network to a stronger hash function.

[becoin]

Not only that, he hasn't described any conceivable method that a "back-doored" sha256 could be used to attack bitcoin in such a way as to cause damage or loss before it could be replaced if necessary.

Well here is one. A message "Bitcoin was created by a special team authorized by US government" published and signed with Satoshi's PGP key. Although some altcoins will flourish that will be the end of bitcoin.

[marcus_of_augustus]

Not only that, he hasn't described any conceivable method that a "back-doored" sha256 could be used to attack bitcoin in such a way as to cause damage or loss before it could be replaced if necessary.

Well here is one. A message "Bitcoin was created by a special team authorized by US government" published and signed with Satoshi's PGP key. Although some altcoins will flourish that will be the end of bitcoin.

... and what would that have to with a back-doored sha256?

[BurtW]

Ultimately it's a moot point

I agree with you wholeheartedly!

Bitcoin is not going to change the hashing algorithm used based upon feelings.  If a weakness is ever discovered there will be plenty of time to switch to another algorithm.

[BurtW]

Not only that, he hasn't described any conceivable method that a "back-doored" sha256 could be used to attack bitcoin in such a way as to cause damage or loss before it could be replaced if necessary.

Well here is one. A message "Bitcoin was created by a special team authorized by US government" published and signed with Satoshi's PGP key. Although some altcoins will flourish that will be the end of bitcoin.

... and what would that have to with a back-doored sha256?

Not only that but doesn't PGP uses RSA, not ECC?

[becoin]

Not only that, he hasn't described any conceivable method that a "back-doored" sha256 could be used to attack bitcoin in such a way as to cause damage or loss before it could be replaced if necessary.

Well here is one. A message "Bitcoin was created by a special team authorized by US government" published and signed with Satoshi's PGP key. Although some altcoins will flourish that will be the end of bitcoin.

... and what would that have to with a back-doored sha256?

Not only that but doesn't PGP uses RSA, not ECC?

My point is there is no need to craft complex 'conceivable' methods to attack bitcoin. What attacker needs to focus on is breaking a single PGP key (was it 2048 bit or less?). This is why it is important not to sit and wait until this key is broken but move on and make bitcoin truly independent of its creator's will.

If you insist on RSA vs ECC argument, then same message might be signed with the key Satoshi used to sign network genesis block.

[J. J. Phillips]

no ice:

i'll tell you what's more likely given all the evidence we've seen over the years re: crypto, the NSA, and Bitcoin.

it's more likely you're a gvt/bank hired troll assigned to come here and inject fear into the Bitcoin community and to try and scare off any new adopters.  i say that b/c everything you've said is hearsay and is based on lack of evidence.  you appear not to understand the first thing about cryptography.  it's pure fear mongering.

I think this is the most sensible interpretation of the thread.

[hhanh00]

No crypto is trusted whether it comes from the NSA or not. Let's say SHA-3 gets preferred treatment because it was not from the NSA. Well, who can be sure that they were not involved at all? With the spy stuff going on, it's better to stick to the math.

That's an excellent point and I understand it.

The problem is that cryptography is a special kind of subject, like physics or math. There are almost no real experts and very few people who are good at it, but there are loads and loads of people who will tell you they are experts or good at it.

It is a pretty safe bet that the NSA has cash to hire the more qualified cryptographers so it seems like they might be qualified to introduce a flawed algorithm that could get past public scrutiny. Have they dfone it before? It's what they do.

The NSA has deep pockets for sure. They are the #1 employer of mathematicians in the US and their budget though classified is estimated at around 50 billion US$.

But have they managed to push a flawed encryption standard through? I don't think we can know. They were caught red handed once but it could be trick too. Get caught on a small lie so that the bigger lie goes undetected.

I see you don't want to touch the NSA with a 10-ft pole. What are the alternatives?

* You choose another hash scheme. I already said that it would be hardly possible to prove that the NSA  was never involved in its development. Even if they weren't, they could still know a way to crack it.
* You choose a 'provably secure hash function'. Well - they are just as secure as another problem deemed to be hard. Then again, the NSA could have solved it.

In short, no one knows what they can do and can't do.

So, we use blind tasting.

What the community has done is to pick a few hash functions: SHA-2, RIPEMD-160 and apply them several times. Each of these functions has had ample public analysis. To keep a weakness secret, they would have to design/find a flaw that is so crafty that no other person can see it. They have many enemies in the world, therefore I think that if there was such a flaw someone else would have pointed it out.
Even if they managed, well, in bitcoin you hash the hash.

The flaw would need to be gigantic.
Much bigger than the MD-5 weakness - and in several unrelated hash functions - and somehow every mathematician in the world is part of a conspiracy of silence.

I don't trust the NSA, but I think that the fact that they were the creator of SHA-2 doesn't impact its applicability in bitcoin.

So, no - I don't think the NSA has broken bitcoin.

PS: I intentionally didn't use any jargon. I believe the concern that the OP has is not related to cryptography details.

[BurtW]

^^^  Great post.   Thanks.

[X7]

I think the NSA created bitcoin... and they created SHA256

[BurtW]

I think the NSA created bitcoin... and they created SHA256

And they created this forum.

[no-ice-please]

Who created SHA256?

[jaxxx]

I think the NSA created bitcoin... and they created SHA256

And they created this forum.

They also have friends who created several dark markets an dodgy exchanges. They are busy little bumble bee's aren't they?

[5970]

Who created SHA256?

NSA. Says wikipedia  http://en.wikipedia.org/wiki/SHA-2