Re: The New GAK-Clipper Thing will Fail

[jim bell <jimbell@pacifier.com>]

At 05:41 AM 10/4/96 -0700, Bill Stewart wrote:
>At 03:07 PM 10/3/96 -0800, jim bell <jimbell@pacifier.com> wrote:
>>The point, of course, is NOT to encourage these companies to support Clipper 
>>IV.  Rather, goal is to suggest to them a "poison pill" which would make 
>>their cooperation meaningless in the end, while at the same time giving them 
>>a 2-year free 56-bit export.  Think of it as a monkey-wrench they can throw 
>>into the works.
>
>A really _fine_ post!  I'm also impressed by the way they announced it
>just _after_ Congress ended its session, while they're busy losing the export
>level in court.

There's an uncomfortable "tactics" problem, associated with my previous 
suggestions.  To say something like, "we think you ought to change that GAK 
proposal by [fill in the blank]", at least IMPLIES that it will somehow be 
"acceptable" should that change be made.   Well, to me, no GAK will ever be 
acceptable.  Even so, I think it would still be tactically useful to help 
sabotaging GAK by "improving it to death."  

I start by assuming that most of the companies who signed onto the Axis 
("alliance") would really have preferred to NOT see GAK, everything else 
being the same.  They want the goodies; they don't want the shit.  I think 
they should be approached by pointing this out, and suggesting that if they 
want to limit the negative publicity they'll surely get from this plan, 
while at the same time collecting the goodies the government is offering, 
they can conveniently and publicly "interpret" their rights broadly, 
announce that they'll structure their systems in the least 
government-friendly way possible.

All this should be possible, because of the fact that this proposal isn't 
really even settled.  In fact, it doesn't even ask the participants to show 
their plan immediately, merely after two years or so.

Insisting that the government pay for all legal challenges is a good start, 
and refusing to do any GAK transfer without a court challenge.


Another thing they could do is to insist that 56-bits of key always remain 
non-GAK.  (perhaps increasing at the rate of two bits every three years.)  
This would make GAK essentially useless for that drift-net fishing that's 
often talked about, because even a sudden policy change forcing "key-escrow" 
people to give up all their keys will still make decrypting a message a 
pain.  It would also make it easier to use super-encryption, because finding 
that needle in that 2**56 haystack would make it impossible to prove which 
particular decrypt contained a further-encrypted message: Even if 
super-encryption were outlawed, it would become essentially impossible to 
prove that none of the other (2**56)-1 messages were not some valid, 
non-super-encrypted message.


Jim Bell
jimbell@pacifier.com