[2159] in cryptography@c2.net mail archive
Re: A Geodesic Society?
daemon@ATHENA.MIT.EDU (Rick Smith)
Mon Feb 23 15:32:57 1998
In-Reply-To: <v04003a0fb113507726b0@[139.167.130.248]>
Date: Mon, 23 Feb 1998 13:59:13 -0600
To: Robert Hettinga <rah@shipwright.com>, cryptography@c2.net
From: Rick Smith <rsmith@securecomputing.com>
It's a terrific read, but there's one nit I'd like to see discussion on:
the notion that you can have both a large amount of personal privacy and a
reputation for being highly creditworthy. In practice I think you'll have
to trade privacy for lower cost credit, or gain privacy by paying more for
credit.
In order to earn favorable rates on your digital bearer bonds, you'll have
to make information about your successfully redeemed bonds as public as
possible. This produces the moral equivalent of book entries (ephemeral,
possibly, but quite visible to the tax collectors). If you try to split
your transactions among two or more different digital IDs, then you produce
two or more shorter credit histories instead of a single longer (and more
appealing) one. So from the very start there is a strong bias against
privacy.
Perhaps in some idealized world there's no reason why a given digital ID
will always correspond to a given individual or other taxable entity. In
practice, we will probably find a variety of traffic analysis type attacks
that can backtrace transactions based on digital IDs. This is especially
true if we talk about the volume of transactions a typical individual might
perform. How do I scrub my paycheck deposit twice a month? What about the
way I buy groceries at the same store every week? I want to do these
routine transactions as cheaply as possible because I do so many of them.
It's the mundane things that leave the widest trail.
If someone could come up with a formal security policy specification that
captures the nuances of this problem, I think the analysis would show that
you can get privacy or low cost credit, but not both.
So the extremes of publicity and personal privacy are faces of the same
cryptographic coin.
Rick.
rsmith@securecomputing.com Secure Computing Corporation
"Internet Cryptography" at http://www.visi.com/crypto/ and bookstores
