On Fri, 18 Sep 1998, Robert Hettinga wrote:

Anyway, In light of more recent crypto-shenanigans from Billary, and the fact that this thing's a small crowd, I figured I'd ask if anyone on these lists had a question they wanted me to ask him.

"Was the encryption used by Bill Clinton to encrypt his deposition escrowed, and if so, with whom?"

One I want to ask, right off the top of my head is, "Given your recent successes in regulating foreign cryptography, what's your timetable for regulating domestic cryptography?"

Or banning outright. alan@ctrl-alt-del.com | Note to AOL users: for a quick shortcut to reply Alan Olsen | to my mail, just hit the ctrl, alt and del keys.

(I'm leaving the damned multiple lists cc:ed on this, as I have no idea where this thread came from. For those of you who send me notes saying I am not allowed to post on your list, plonk.) At 2:33 AM -0700 9/19/98, Alan Olsen wrote:

On Fri, 18 Sep 1998, Robert Hettinga wrote:

...

Anyway, In light of more recent crypto-shenanigans from Billary, and the fact that this thing's a small crowd, I figured I'd ask if anyone on these lists had a question they wanted me to ask him.

"With all the demands from Governent to escrow keys, what steps are being taken to protect the these keys and/or backdoors from misuse from people within the Government and those outside of the Government?"

"Would you use escrowed cryptography for your private communications? Who would you trust to hold those keys?"

Well, I believe that even asking these sorts of questions (and possibly getting answers from government) plays into the hands of the GAKkers. After all, suppose they give pretty good answers? What if, for example, they propose a committee consisting of the entire Supreme Court, the Director of the Sierra Club, and so on, and say that a _unanimous_ vote is required to gain access to a key? Does this at all change the fundamental unconstitutionality of telling me that I will face imprisonment if I speak or write in a manner which is not part of their "escrow" arrangement? If I keep a diary without depositing an escrowed key with this noble, careful, thoughtful committee of wise persons? Of course not. My speech, my writing, my private codes, my whisperings, my phone conversations...all of these...are not subject to govenmwental approval. "Congress shall make no law..." Doesn't say that Congress or the courts of the President get to declare illegal certain modes of speaking. (*) (* Please, I hope no one brings up "loud speech," "shouting fire," "obscene speech," "seditious speech," "slanderous speech," etc. This is well-trod ground, but clearly all of these apparent limits on speech, whether one agrees with them or not, have nothing to do with an overbroad requirement that speech only be in certain languages, that letters only be written on carbon paper with a copy filed with the government, and so on. The restrictions on _some_ kinds of speech are not a license to license speech, as it were, or to require escrow of papers, letters, diaries, and phone conversations and such.) My point about Alan's (and others') points is that if we get engaged in this kind of debate about how key escrow might work, we shift the debate from where it ought to be to where they _want_ it to be, namely, to issues of practicality. My view is that my writings are mine. They can try to get them with a search warrant or a court order, but they'd better not threaten me with imprisonment if I choose to write in some language they can't read. And that's all crypto really is, of course. Just another language. --Tim May (This space left blank pending determ. of acceptability to the gov't.) ---------:---------:---------:---------:---------:---------:---------:---- Timothy C. May | Crypto Anarchy: encryption, digital money, ComSec 3DES: 831-728-0152 | anonymous networks, digital pseudonyms, zero W.A.S.T.E.: Corralitos, CA | knowledge, reputations, information markets, Licensed Ontologist | black markets, collapse of governments.

On Fri, 18 Sep 1998, Tim May wrote:

(I'm leaving the damned multiple lists cc:ed on this, as I have no idea where this thread came from. For those of you who send me notes saying I am not allowed to post on your list, plonk.)

I usually leave most of the cc:s due to not knowing who is reading what anymore. (Gateways being as they are...)

...

"Would you use escrowed cryptography for your private communications? Who would you trust to hold those keys?"

Well, I believe that even asking these sorts of questions (and possibly getting answers from government) plays into the hands of the GAKkers.

Depends. With the right questions (assuming you get honest (Ha!) answers) you can make them look like control freaks, hypocrites, and/or worse. Hitting them with the unexpected can generate soundbites that can be used for the cause. Besides... If you don't challenge them in public, the public tends to get the impression that no one opposes this KRAP.

After all, suppose they give pretty good answers? What if, for example, they propose a committee consisting of the entire Supreme Court, the Director of the Sierra Club, and so on, and say that a _unanimous_ vote is required to gain access to a key?

I would respond that they were lying. I would also ask them if it was going to be like the current wiretap authorization commissions that just rubber stamp requests.

Does this at all change the fundamental unconstitutionality of telling me that I will face imprisonment if I speak or write in a manner which is not part of their "escrow" arrangement? If I keep a diary without depositing an escrowed key with this noble, careful, thoughtful committee of wise persons?

Of course not. And questions of constitutionality need to be asked of them directly to their faces. Without that direct confronation, they will do whatever they damn well please. (Not as if they don't now...)

Of course not. My speech, my writing, my private codes, my whisperings, my phone conversations...all of these...are not subject to govenmwental approval. "Congress shall make no law..."

Encrypt! Encrypt! OK!

Doesn't say that Congress or the courts of the President get to declare illegal certain modes of speaking. (*)

(* Please, I hope no one brings up "loud speech," "shouting fire," "obscene speech," "seditious speech," "slanderous speech," etc. This is well-trod ground, but clearly all of these apparent limits on speech, whether one agrees with them or not, have nothing to do with an overbroad requirement that speech only be in certain languages, that letters only be written on carbon paper with a copy filed with the government, and so on. The restrictions on _some_ kinds of speech are not a license to license speech, as it were, or to require escrow of papers, letters, diaries, and phone conversations and such.)

I am willing to argue that most of those restrictions are against what the founding fathers meant when they wrote the document. But currently we live in a land where "consensus doublethink" is the nature of the law.

My point about Alan's (and others') points is that if we get engaged in this kind of debate about how key escrow might work, we shift the debate from where it ought to be to where they _want_ it to be, namely, to issues of practicality.

What needs to be pointed out is that it CANNOT work. By asking them if they would use it, you point out that the system is so untrustworthy that they themselves would not even use it. (And if you say they would, you just start laughing.)

My view is that my writings are mine. They can try to get them with a search warrant or a court order, but they'd better not threaten me with imprisonment if I choose to write in some language they can't read. And that's all crypto really is, of course. Just another language.

The current administration has taken the tact that ANY argument, no matter how irrational, to advance their legal position, must be used. That has to colapse opon itself after a while. Whether it be through the total errosion of their reputation capitol (What little is left.) or through judges who cry "enough" and toss out the entire specious line of reasoning. alan@ctrl-alt-del.com | Note to AOL users: for a quick shortcut to reply Alan Olsen | to my mail, just hit the ctrl, alt and del keys.

At 8:38 PM -0800 9/17/98, Robert Hettinga wrote:

Anyway, In light of more recent crypto-shenanigans from Billary, and the fact that this thing's a small crowd, I figured I'd ask if anyone on these lists had a question they wanted me to ask him.

Probably a bit late, but I would ask, "If the government is going to require GAK, how is it going to demonstrate that all the keys is accessed were accessed according to legal proceedings?" Another way of asking is, "How do we ensure that all government accesses to keys become publicly known within a reasonable amount of time." (I would define reasonable as much less than 5 years.) (I've been reading David Brin's "The Transparent Society".) ------------------------------------------------------------------------- Bill Frantz | If hate must be my prison | Periwinkle -- Consulting (408)356-8506 | lock, then love must be | 16345 Englewood Ave. frantz@netcom.com | the key. - Phil Ochs | Los Gatos, CA 95032, USA

Okay, so the short precis on Magaziner's answer to my question about encryption controls, foriegn or domestic, is he's agin it.

He says that controlling foriegn encryption is impossible, and controlling domestic encryption is, at the very least, unconstitutional.

This is way beyond what I expected. If I had expected him to go beyond what he said at the MIT Media in Transition forum earlier this year I would have made every effort to attend. Essentially at MIT I tried to put words in his mouth by stating what I knew he had said in private and adding that if he agreed then Freeh would insist that one or the other of them go. He was not at that time able to say anything more, although he did not leap to the opportunity to push the Freeh line.

He says that the reason the administration's encryption policy is so convoluted is that the law enforcement and the "economic" encryption camps, anti, and pro, evidently, is that the two sides are at loggerheads.

This statement in and of itself is not the type of thing I would expect unless either Magaziner is planning to jump ship or he knows that he can get away with it without retribution. If it is the latter it would indicate that Freeh has very much less influence than he did 6 months ago. Not exactly a ringing endorsement of our position but something that can be used against spokesmen for the FBI party line. All in all I would prefer allowing the crypto export laws to time out, becomming progressively less relevant until they disappear than have bills appear in Congress. However good the bill that goes into congress the result will be at best a compromise. Phill

On Sat, Sep 19, 1998 at 02:13:39PM -0700, Tom Weinstein wrote:

...

Arnold G. Reinhold wrote:

One question I'd like asked is whether the US Gov will approve 56-bit RC-4 for export on the same terms as 56-bit DES. That would allow export versions of web browsers to be upgraded painlessly, making international e-commerce 64 thousand times more secure than existing 40-bit browsers. (56-bit DES browsers would require every merchant to upgrade their SSL servers and introduce a lot of unneeded complexity.)

Actually, it wouldn't be any easier to deploy 56-bit RC4 than DES. Either would require roughly the same changes to both clients and servers.

Not easier technically but "easier" maybe politically. Key length seems to be held (probably wrongly) as a rough measure of crypto "strength" by journos and those in power. 40bit RC4 is weak. How strong would 56bit RC4 be? -- pgp 1024/D9C69DF9 1997/10/14 steve mynott

Tom Weinstein writes:

...

Arnold G. Reinhold wrote:

One question I'd like asked is whether the US Gov will approve 56-bit RC-4 for export on the same terms as 56-bit DES. That would allow export versions of web browsers to be upgraded painlessly, making international e-commerce 64 thousand times more secure than existing 40-bit browsers. (56-bit DES browsers would require every merchant to upgrade their SSL servers and introduce a lot of unneeded complexity.)

Actually, it wouldn't be any easier to deploy 56-bit RC4 than DES. Either would require roughly the same changes to both clients and servers. And from a protocol perspective, it would be worse, at least for SSL, since SSL doesn't have a 56 bit RC4 mode at all.

-Ekr -- [Eric Rescorla ekr@rtfm.com]