NSA Crypto Backdoor?

According to security experts, an algorithm for generating random numbers that is included in an official standard documented by the National Institute of Standards and Technology (NIST) could potentially include a backdoor planted by the NSA.

In a recent blog entry, cryptographer Bruce Schneier describes research that was presented by his colleagues Niels Ferguson and Dan Shumow at the CRYPTO 2007 conference this past August. The security researchers have raised concerns about a potential backdoor in the Dual_EC_DRBG algorithm, which is documented in NIST's 800-90 publication about deterministic random bit generators. Dual_EC_DRBG, which is based on elliptic curves, is said to be significantly slower to compute than the other algorithms in the standard and was supposedly only included at all because it has the strong support of the NSA.

Dual_EC_DRBG uses a seemingly arbitrary series of specific fixed numbers which are published in the standard to define the elliptic curve used for the algorithm. The origin of those numbers has not been revealed or explained but it is possible to use other numbers instead. The researchers realized that the fixed set of numbers used in Dual_EC_DRBG could have a mathematical relationship to a secret second set of numbers, which could then be used as a master key to decrypt content.

Source: ArsTechnica

Quite scary indeed to think there are static numbers in a encryption standard that only requires the determination of an additional set in order to basically nullify the encryption. This almost sounds like a replacement for the failed Clipper Chip which would have been just as detrimental to the internet as this will be if it turns out to be true. Sure seems like these days truth and fiction are sometimes a lot closer than people think when just such a story as this was spun in the Dan Brown book Digital Fortress.

If we cannot trust the NSA when contributing to such a project (and it seems kind of foolish to do so given their actions) then why are we bothering anymore, and why was the SELinux enhancement set so readily welcomed by the Linux community (who tend to be a paranoid lot)? Perhaps its time for a recompile to extract a few choice pieces of functionality...

[edit on 19-11-2007 by Helig]

I was curious about the SELinux too...

If Bruce Schneier is concerned, then everyone should be, he is one of the top guys.

The idea of using some unexplained constants could be of interest. From what I understand, when DES was created, it used some 'unexplained constants' (S box initial values). It turned out about 20 years later they had been chosen for a good reason... They made differential cryptanalysis more difficult. At the time DES was created though, only IBM and the NSA knew about differential cryptanalysis, and they kept it secret until academia 'rediscovered' it in the late 80's, early 90's.

Something to think about is that if the NSA are supporting a cryptographic method, it is probably strong enough for regular people to use, but just weak enough for the NSA to be the only people able to break it. It would go against the whole idea of the NSA to encourage people to use a code that they can't break, surely?

I'd feel very wary of using any sort of cipher devised by someone else, ESPECIALLY the NSA.

If you're going to protect your information and you're capable of doing so, come up with your own methods. That way you know if there's a key built into it and if there is you know just how well its hidden.

Never trust the NSA, because they'll never trust you.

reply to post by Helig
Thanks for the heads up. The very first thing that crossed my mind was the Clipper chip. That effort failed to materialize so I would suspect they would approach the issue from a different direction. This seems like a reasonable approach from their point of view I'd bet (guess any approach is reasonable in their eyes).

There have been back door entrance for US govt on software since the early 80's...the promis software scandel with Ed meese and the reagan crew, pretty much the same people messing things up now.

European govt and police forced purchased US software that allowed US govt to view their records, and more...it was a big mess in the late 80's. This is no different....same crooks, same crimes

About 3 years IIRC before Windows Vista was due to ship, Bill Gates had a secret meeting (but it was known he had a meeting) with the UK Government to discuss adding a master key to BitLocker so that the Government could decrypt anything encrypted using BitLocker technology in an instant, without have to guess passwords or crack the encryption (which obviously takes time).

IMHO this makes BitLocker encryption a broken encryption technology, because if the UK government can have a back-door key, then someone could conceivably generate said back-door key for total access.

I'm not quite sure how SELinux could be weakened due to the way it works (IIRC no encryption is included - it is simply a set of rules for file access, not much different to permissions, but much more flexible). EDIT: I guess a back-door could be put in here, and some kind of attack that breaks it and opens access?

I guess we've hit the ultimate problem with cryptology: TRUST. If we can't trust the programmer and/or designer of an encryption scheme to develop something secure, then there is no point even using it.

If I use x product for encryption, I'm trusting very much the developers and designers of that product to ensure there are no back-doors in the software, and that the encryption scheme itself is strong.

If I don't trust the encryption software you wrote, I implicitly do not trust you, and vice-versa.

The question is: who to trust??????

A great article! With our reliance on computers at an all-time high and accelerating, and increasing reports of data leaks, this is very serious stuff indeed, when we're questioning the very technology that is used to protect our data.

[edit on 19-11-2007 by mirageofdeceit]

[edit on 19-11-2007 by mirageofdeceit]

what was the name of the Russian spy caught in New England (DC?) for spying. allegedly they sent a 'voltage spike' down the power line to turn his computer on. How many of you buy that? Do you turn your power strip off after every use? Would it matter?

First thing I do when standing up a new server...disable SElinux.

Uk encryption law.

Having a law making it illegal to withold the password will only make the smalltime criminals hand over the password, The powers that be need a way to get into files if they really need to, so they just get backdoors included.

If they want your data they're gonna get it one way or another. Most people could only dream of being so important that the NSA or the like would trouble themselves to find out what you have encrypted.