5 ms·

Spiegel might have done better to call it a GCHQ tool used by the NSA. You are suggesting that a tool that has been around for 10 years was picked up and used

by beezle 12y ago

Spiegel might have done better to call it a GCHQ tool used by the NSA. You are suggesting that a tool that has been around for 10 years was picked up and used by NSA/GHCQ. Even if we grant that, the sophistication level is a bit more than "simply modified". When is a fork no longer a fork because the code base has been modified/improved? With your argument, one could go back as far as they wished - NSA takes an idea from academia and implements it, then we just say 'oh well, it was the univeristy who came up with it, not NSA as they only used it'

cyphunk 12y ago
My question was concerning attribution. Though debatable lets assume still the original was not theirs and they remixed it. Even if they made massive changes turning into their own beast allowing us to call the Snowden version significant we are still left with the question of: Are all the attacks over the past 10 years attributed to this general varietal, their attacks? Without full knowledge attribution is like a blind taste testing of different vintages of wine. Also I think the remixed and implemented idea from academia applies less. It would be more like taking the paper from one journal, changing a few paragraphs and publishing it in another. And assuming these journals had a policy of publishing without names, how can we know who the author is in all prior journals when we eventually find a way to attribute one author to one journal?
drostie 12y ago
Basically what cyphunk is saying is: how do we know that this was used by the NSA? More precisely: how does Spiegel know to say "clear proof that Regin is in fact the cyber-attack platform belonging to the Five Eyes alliance"? We have something out in the wild (Regin) which contains a component (QWERTY) which was leaked in the Snowden documents, however it is quite possible that both descend from a common non-NSA source (Putative) such that: P (non-NSA) / \ Q R (non-NSA) (NSA) Furthermore it is possible that in fact P = R, and that Q was derived from Regin rather than the other way around. Lots of possibilities are out there. We don't really have proof that P = Q and that therefore R also belongs to the NSA. The issue is that we can attribute malicious attacks to R, and the article seems to be suggesting that therefore we should attribute these to the NSA. The reply by cyphunk is saying that this is a dangerous logical leap.

by beezle 12y ago

cyphunk 12y ago

My question was concerning attribution. Though debatable lets assume still the original was not theirs and they remixed it. Even if they made massive changes turning into their own beast allowing us to call the Snowden version significant we are still left with the question of: Are all the attacks over the past 10 years attributed to this general varietal, their attacks? Without full knowledge attribution is like a blind taste testing of different vintages of wine. Also I think the remixed and implemented idea from academia applies less. It would be more like taking the paper from one journal, changing a few paragraphs and publishing it in another. And assuming these journals had a policy of publishing without names, how can we know who the author is in all prior journals when we eventually find a way to attribute one author to one journal?

drostie 12y ago

Basically what cyphunk is saying is: how do we know that this was used by the NSA? More precisely: how does Spiegel know to say "clear proof that Regin is in fact the cyber-attack platform belonging to the Five Eyes alliance"? We have something out in the wild (Regin) which contains a component (QWERTY) which was leaked in the Snowden documents, however it is quite possible that both descend from a common non-NSA source (Putative) such that: P (non-NSA) / \ Q R (non-NSA) (NSA) Furthermore it is possible that in fact P = R, and that Q was derived from Regin rather than the other way around. Lots of possibilities are out there. We don't really have proof that P = Q and that therefore R also belongs to the NSA. The issue is that we can attribute malicious attacks to R, and the article seems to be suggesting that therefore we should attribute these to the NSA. The reply by cyphunk is saying that this is a dangerous logical leap.