From: Mike_Spreitzer.PARC@xerox.com
Date: Thu, 21 Jul 94 16:49:59 PDT
To: cypherpunks@toad.com
Subject: "Key Escrow" --- the very idea
Message-ID: <94Jul21.164935pdt.14430(3)@alpha.xerox.com>
MIME-Version: 1.0
Content-Type: text/plain


(1) I'm not an anarchist.  Does that make me out of place here?  I'm willing to
live with some amount of government, as long as us owners stand a chance of
controlling or overthrowing it.  My biggest problem with Capstone is that it
changes the balance of power too much.

(2) I think crimes can be committed in cyberspace.  Substantially, if not
entirely, in cyberspace.  Maybe not so many now.  But I think it's
intellectually dishonest of us who understand the growing importance of
cyberspace to claim there won't be any social contracts there that could be
violated.  I accept the terms of the 4th ammendment: search and siezure allowed
when due process followed.  "Key escrow" is an attempt to implement the
cyberspatial analog of search.

(3) The Feds must know they can't prevent modestly well funded, educated, and
motivated folks from using unbreakable cryptography amongst themselves.  The
argument for doing key escrow anyway is that by installing a breakable
infrastructure, they'll make enough investigations cheaper and more effective
to be worth it.  Note that's a comparison of their money and success rate
against our privacy; no wonder they got it so wrong.

(4) If you accept points (1) and (2) above, you're left wanting a way to
implement searches in cyberspace when due process is followed.  I hope
anarchists won't be the only people opposing changing the balance of power
greatly in the government's favor (by poorly designed key escrow).  What are
the rest of us left to answer with?  Perhaps a much better key escrow design.
One that integrates the search with the due process in a cryptographically
strong way; one that can't be subverted by a few people in a few organizations.
For example, who says an escrowed key must have only two parts?  Why not a
whole lot of parts, distributed to a whole lot of people/organizations?  If
there are only 1000 legal wiretaps in a year, and they're already fairly
expensive, we can add a fair amount to the cost before it gets significant.
And again, remember where we're weighing money against freedom.  It may be that
we just have to spend more to stay a reasonably free society.  Also, it's worth
debating just how strong the protections have to be.  Will we need them to be
stronger than those against physical searches?  How few people does it take to
subvert the current protections against illegal searches?  Do we feel that
needs to be changed?  How much are we willing to spend on it?