From: "Vladimir Z. Nuri" <vznuri@netcom.com>
Date: Mon, 27 May 1996 06:36:10 +0800
To: Matt Blaze <mab@research.att.com>
Subject: Re: "Key Escrow without Escrow Agents"
In-Reply-To: <199605250132.VAA09818@nsa.tempo.att.com>
Message-ID: <199605252056.NAA28915@netcom7.netcom.com>
MIME-Version: 1.0
Content-Type: text/plain



the splitting of the keys among different *private*
sites, in such a way that key requests must be revealed publicly,
is an interesting idea.
the way that individual sites evaluate revelation requests 
might usefully be compared to "juries" in our society. 

in the case of the grand jury, the government invites a set
of citizens to determine if an indictment is justified. their
consensus opinion determines the decision.

in the case of conviction, we have a similar system.

the idea of "jury nullification" has a direct analogy as well.
in the key escrow scheme MB proposes, if a lot of sites refuse to release
keys based on the circumstances of the case, that would be
very similar to jury nullification.

I can see that you might create a code of law that determines
what procedures that these "distributed key juries"  are supposed
to follow. but like our legal system, the interpretation and
application is ultimately left up to them.

an interesting system, that is commendable for trying to
find a compromise between two seemingly irreconcilable polarities
(privacy and surveillance) but I doubt anyone in law enforcement
(with the mindset, "I can't be stopped from doing my job as I
see fit or criminals will get away") would go for it in the current form.