IP: Wash Post editorial on Encryption

[Dave Farber <farber () cis upenn edu>]

> http://www.washingtonpost.com/wp-srv/WPcap/1999-05/10/022r-051099-idx.html
>
> Encryption and Free Speech
> Monday, May 10, 1999; Page A22=20
>
> A FEDERAL APPEALS court has declared unconstitutional government
> regulations designed to restrict the export of certain types of encryption
> software. The decision on this arcane issue is important. It strikes a
> serious blow against the government's export-control regime, which has been
> increasingly controversial as encryption programs have become crucial to
> computer security, online commerce and data-integrity applications. The
> government regards encryption algorithms as munitions and -- to the dismay
> of high-tech industry and privacy activists -- restricts the export of
> strong cryptography. The current case arose when Daniel J. Bernstein, then
> a graduate student in California, sought to publish the source code for a
> cryptographic system he called "Snuffle." The government, however, insisted
> he needed a license.
>
> Computer source code is a form of writing that has no easy analogue in
> First Amendment law. It is text written by people that can be transformed
> by computers into the executable files that we call programs; it is speech
> that actively does things. Source code is primarily a vehicle for people to
> instruct computers, a dialect that programmers can speak and that computers
> can understand. But it also is undeniably expressive of scientific ideas.
> Indeed, source code is routinely shared by people to communicate how
> certain computer-related tasks can best be accomplished. Is source code
> speech or is it a device?
>
> The government's view is that source code on paper is expressive speech but
> that when on a disk (or on paper that can be scanned by a computer), it
> becomes a device whose purpose is not communicative but functional. This
> distinction is more than a bit strained, and a divided 9th Circuit panel
> now has rejected the argument that source code's functionality makes it
> regulable. It held that the export controls, at least as applied to source
> code, are "an impermissible prior restraint on speech" that attack
> scientific expression.
>
> This may well be analytically correct, but it is, nonetheless, troubling.
> The export-control regime has all kinds of problems -- not the least of
> which is that encryption is now so widely available that the Pandora's box
> is open.  But the government's interest in controlling the spread of strong
> encryption is a real one that cannot be dismissed blithely, nor can an
> export-control system that excludes source code be very meaningful. One of
> the panel's judges, Myron Bright, suggested in a frustrated concurrence
> that "the importance of this case suggests that it may be appropriate for
> review by the United States Supreme Court." Indeed, it would be well worth
> the high court's considering where the precise line between speech and
> machine really is.=20
>
>                            =A9 Copyright 1999 The Washington Post Company