Re: 'Nother MIT talk on crypto...

From: hughes@ah.com (Eric Hughes)
> Micali's "fair" cryptosystem is a much better key surrender system
> than Clipper, but it still allows non-intended recipients for a
> message.
> 
> For this reason, I don't like it either.
> 
> Fight _all_ intrusions.
> 
> Eric

Right - here are some quotes from Micali's paper in the Crypto 92 proceedings.

"Abstract.  We show how to construct public-key cryptosystems that are _fair_,
that is, strike a good balance, in a democratic country, between the needs of
the Government and those of the Citizens.
[...]
"In this paper we show how cryptographic protocols can be successfully and
efficiently used to build cryptosystems that are fairer, that is, that strike
a better balance, in a democratic country, between the needs of society and
those of the individual."

Micali's system is basically a key escrow system that would be quite appealing
to those who love Clipper.  At least he has the honesty to make it clear that
such a system makes more sense if competing systems are made illegal:

"Of course, if using any other type of public-key cryptosystem were to be made
_illegal_, Fair PKC's would be most effective in guaranteeing both private
communication to law-obeying citizens and law enforcement.  (In fact, if a
criminal uses a phone utilizing a Fair PKC to plan a crime, he can still be
secured to justice by court-authorized line tapping.  If he, instead,
illegally uses another cryptosystem, the content of his conversations will
never be revealed even after a court authorization for tapping his lines, but,
at least, he will be convicted for something else: his use of an unlawful
cryptosystem.)  Nonetheless, as we shall discuss in section 4, Fair PKC's
are quite useful even without such a law."

When I first heard of this so-called "Fair" (one of the most misused words
in political debate) system, my reaction was to snort in derision.  But since
Clipper it starts to look like the lesser of two evils.  That just shows how
the terms of the debate can shift.  Eric is right that the best thing to do
is to remain firmly committed to free access to cryptographic technology for
everyone.

Hal

> Right - here are some quotes from Micali's paper in the Crypto 92
> proceedings.

> "Abstract.  We show how to construct public-key cryptosystems that
> are _fair_, that is, strike a good balance, in a democratic country,
> between the needs of the Government and those of the Citizens.
> [...]

In a democratic country, the needs of the government should be
identical to the needs of the citizenry.  One of the basic political
ideas upon which our society is founded is that government serves the
people and not the other way around.

We've seen Micali's point raised again and again.

The NSA spokesman quoted on this list some time ago, whose name I have
forgotten, made this point as well.  He also claimed that "government
money" had been used to pay for most recent cryptographic advances.
This is likely untrue, but it does raise an important philosophical
point: The government has no money of its own, it spends the money of
the citizenry.  The argument that the citizenry should be denied the
benefits of their expenditures is a weak one.

These points are likely to be obvious to most readers of this list,
but I haven't seen them raised.  I think that every time an NSA
official or White House spokesman or anyone else questions the
important political principle of "government of the people, by the
people, and for the people" we should point this out as often as we
can.  We must delegitimize NSA and their friends and show them for the
rascals they truly are.

Peter