Re: real cryptographers - how safe would you be?


"RR" <newspaper.20.broom@xxxxxxxxxxxxxxx> wrote in message
news:%ijdi.15245$wH4.6102@xxxxxxxxxxxxxxxxxxxxxxxxxxxxx


- Does anyone (e.g. the CIA, NSA, Chinese) have the
ability to crack well
known encryption such as RC4, AES 128, AES 256?

Anyone in a position to know the answer to this would be
subject to the restrictions of at least a TS security
clearance. In the US, it would most likely be TS/SCI.
Someone with first-hand knowledge of this is not likely to
post it on USENET unless he's planning to retire to
Leavenworth.

- Is a brute force attack nowadays viable on any 128 or
256 key?

Since you haven't specified the cipher, and you seem to be
using terminology rather loosely, the answer is "maybe". In
fact, the answer is always "maybe" because of the answer to
the previous question. I'm assuming that you don't really
mean "brute force attack" in the precise meaning it has here
but rather some attack that recovers the key using
discovered weakness in the cipher.

- If "yes" to either of these questions, would any of
these organizations
provide the Zimbabwe regime with the technology? (A
political, question, I
know - feel free to treat as rhetorical.)

Probably not, but they might provide said government and
others with filtered information derived from the broken
messages.

- Would the fact that the regime has seen an encrypted
email from you simply
mean that they arrest you and perform some "rubber hose
cryptography"? In
which case, encryption is of no help and is probably a
really bad idea.

Obviously yes. When cryptography is used as a weapon (when
it becomes a threat), people and governments aren't
constrained to "play fair" (any more than they would be if
you threatened them with a pistol). Don't expect them to
limit their festivities to you; they will also go after your
friends, relatives, and the friends and relatives of whoever
it was you sent the message to.


- You know you can visit any number of websites that have
HTTPS (SSL)
connections, would you assume that an AES 256 connection
(using FireFox, for
example) would be pretty much unbreakable and not worth
monitoring (given
there are hundreds of thousands per day)? Therefore, such
a website which
allowed you to send a message via your browser would be
reasonably safe
(assuming you can trust the website owner).

Who can you trust, really? I have an applet on my website
that I claim uses AES-256 for encryption and SHA-256 for key
expansion and authentication. I believe that no practical
improvement is possible in the algorithm (so far as the
security of the end product is concerned).

Should you believe me? Should you trust me? Why? And what
about if your computer is compromised so that your
keystrokes and mouse movements are being quietly logged? It
doesn't matter how effective the encryption technology is if
it can be defeated by other means.

- If you couldn't use any of the above methods, what would
you do to safely
get information out? Would you setup a one-time pad
procedure on paper or,
maybe, fax with an outside party?
I guess this depends on how much
information you need to send and whether you could
disguish the transmission
in an innocuous method.

Perhaps a system of codewords put together so that they look
innocuous but convey your message. This is not, however,
cryptography. And you are right: the amount of information
you can transmit is limited.

- What about sending CDs or floppies that have a hidden
volume feature such
as TrueCrypt's? Plausible deniability may be the safest
means.


When you are talking about situations like this, "plausible
deniability" is moot. If you are suspected, your plausible
denials will be of no consequence whatsoever. Remember, you
can't play by the rules if there are no rules or if the only
rule is, "Might makes Right."



.

Quantcast