How Dangerous is Crypto?

[Jim Windle]

In this debate about restrictions on crypto after Tuesday's events I
think it important to address the issues on logical and empirical
rather than emotional or ideological grounds.  I have made a number of
posts but would like to try to summarise.  These are my thoughts, your
mileage may vary.  When the terms crypto or strong crypto are used
what is being referred to?  Presumably it is user friendly  version of
cryptographically strong Public key systems.  How dangerous are these
programs?  And in related fashion what happens if they are banned or
restricted, for terrorists, cryptography enthusiasts and the general
public.  Is it even practical?  And if so what alternative systems are
available for "legitimate" uses and how good are they.  A few points
on these questions.

First, while it is logical to me a educated and reasonably technically
sophisticated westerner that the bin Laden terror network uses these
public key systems because its how I would do it if I were them, I
have see no hard evidence that in fact they use public key
cryptography or even the internet.  I have seen press reports claiming
without substantiation that this is the case but no real evidence of
it.

Second, if we assume for a minute that the terrorist use public key
systems and that they could be removed from and kept out of terrorist
hands (and also their ability to reproduce a working system on their
own) what would they do?  Would they use a one time pad system that is
logistically more difficult but potentially more secure
cryptographically if used properly.  If they want to use crypto and
public key systems are not available this seems likely to me, and I
don't see how this makes the NSA's job any simpler.  Whatever the
terrorist do, I doubt strongly is will be communicating in the clear
over channels easily accessible to the NSA.

Third, assuming public key systems could be removed from and kept out
of terrorist hands it would mean eliminating these existing sytems
which we know to work pretty well from private and commercial uses.
If removed the affects on institutional financial markets and
individuals who purchase products otherwise unavailable to them over
the internet would be devastating.  If existing public key systems are
replaced with a some sort of escrowed key system there would certainly
be disruptions during its adaptation, to the extend there were
widespread doubts about its efficacy there would be lessening of
electronic commerce dependant upon crypto.  Who would bear the costs
of removing and replacing the existing public key systems and what
would those costs be?  This I can't answer except to say they are
likely to be very high.  Finally is a key secrow system even workable?
And if so what do we do until it is developed in a reliable form?

Considering these points, I have to conclude that it would be
impossible to deny strong crypto in the former of an existing system,
a sytem they develop themselves or a simple one time pad system, to
terrorist groups and there is no gain to be had by "denying" them
these systems, even if they use them at all.  Furthermore I conclude
that the attempt to do so if made would likely be very costly to both
the financial services industry and individual consumers even if a
viable replacment system based on key escrow is available.  Finally I
think, with much less certainly though, that a workable key escrow
system is likely to be difficult to develop and successfully implement
and if develped it is likely to impose a number of direct and indirect
costs.

Jim Windle


Join 18 million Eudora users by signing up for a free Eudora Web-Mail account at http://www.eudoramail.com



---------------------------------------------------------------------
The Cryptography Mailing List
Unsubscribe by sending "unsubscribe cryptography" to majordomo at wasabisystems.com