Re: Telling quote from Bernstein hearing

[Greg Broiles <gbroiles@netbox.com>]

Rich Burroughs wrote:

> I did not find his arguments very persuasive -- he seemed stuck on the idea
> that crypto is not speech, when that had already been ruled against by
> Judge Patel.

As I understand the government's argument, this is what they're trying to say:

Even though code can be speech, if it is executable code it is speech and
something else at the same time. The government is entitled to regulate or
prohibit the export of the "something else" without regard to the speech
component, as long as they are content-neutral with respect to the speech
component. For example, a cruise missile is subject to export control.
Painting words on the side of the missile won't change that. Executable
crypto code is subject to export control, and the fact that it has an
expressive component doesn't change that. The comments about "We don't know
or care what Mr. Bernstein is doing and we're not trying to control his
academic discourse" were intended to emphasize that the government is
content-neutral with respect to the expressive content of his crypto code.
Because the regulations are content-neutral with respect to the expressive
content, they are subject to an "intermediate scrutiny" standard (not the
"strict scrutiny" standard which would be applied if the restrictions were
content-based). The government believes that the AECA and the ITARs can
pass the "intermediate scrutiny" test.

In a nutshell, the government thinks that the expressive content of code is
the algorithm(s) expressed in it. For example, expressive content of PGP
2.x is RSA and IDEA and MD5 and some key management stuff. And they say
that their regulations won't be interpreted to cover versions of that
expressive content which don't take the form of machine-readable code. 

While I think that the entire export control scheme, as it is applied to
crypto today, is an exercise in dishonesty and futility*, I think that the
questions this case presents are difficult questions. I don't think it's
really comprehensible to separate the "expressive content" of source code
from its "functional content", because one dictates the other. Either the
government is able to control the functional content (in which case it
controls the expressive content), and the First Amendment is seriously
curtailed; or the government is unable to control the functional content,
and its ability to control the production and import/export of items of
politico/military significance is seriously curtailed. Historically, courts
have been deferential to the executive branch's ability to control military
and political/diplomatic matters. 

Either way, someone's pissed off, and someone thinks the court is sending
the country to hell in a handbasket. The "crypto is central to national
security" argument shouldn't be dismissed summarily - at least not if you
think Tim's predictions about cryptoanarchy are accurate. Crypto _is_
important. Crypto _is_ significant in the political and military venues.

(* It's dishonest and futile because it is insufficient to reach its stated
goals - if the purpose is to prevent the spread of strong crypto worldwide,
the government must control the expression of strong crypto algorithms and
techniques in all media and all forms. Such regulation would be
incompatible with the First Amendment and freedom generally. The present
regulations are so poorly matched to the apparent goal that I suspect they
have another purpose; and I suspect that purpose is to control the domestic
development of strong crypto. And attempting to do so while claiming to do
something else is dishonest. (That result is also incompatible with the
First Amendment and freedom generally.) And that's what I think of crypto
regulations, whether or not I appear to be agreeing with the government.) 
--
Greg Broiles                |  "We pretend to be their friends,
gbroiles@netbox.com         |   but they fuck with our heads."
http://www.io.com/~gbroiles |
                            |