CINDY A. COHN, ESQ.; SBN 145997
   McGLASHAN & SARRAIL
   Professional Corporation
   177 Bovet Road, Sixth Floor
   San Mateo, CA 94402
   Tel: (415) 341-2585
   Fax: (415) 341-1395
   
   LEE TIEN, ESQ.; SBN 148216
   1452 Curtis Street
   Berkeley, CA 94702
   Tel: (510) 525-0817
   
   M. EDWARD ROSS, ESQ.; SBN 173048
   STEEFEL, LEVITT & WEISS
   A Professional Corporation
   One Embarcadero Center, 30th Floor
   San Francisco, CA 94111
   Tel: (415) 788-0900
   
   JAMES WHEATON, ESQ.; SBN 115230
   ELIZABETH PRITZKER, ESQ.; SBN 146267
   FIRST AMENDMENT PROJECT
   1736 Franklin, 8th Floor
   Oakland, CA 94612
   Tel: (510) 208-7744
   
   Attorneys for Plaintiff
   Daniel J. Bernstein
   
   
                    IN THE UNITED STATES DISTRICT COURT
                                      
   
                  FOR THE NORTHERN DISTRICT OF CALIFORNIA
                                      
   DANIEL J. BERNSTEIN )
   ) C 95-00582 MHP
   Plaintiff, )
   ) PLAINTIFF'S MEMORANDUM OF
   v. ) POINTS AND AUTHORITIES IN ) SUPPORT OF MOTION FOR PARTIAL UNITED
   STATES DEPARTMENT OF ) SUMMARY JUDGMENT AND/OR
   STATE et al., ) SUMMARY ADJUDICATION OF ) ISSUES 
   Defendants. ) F.R.C.P. 56
   )
   ) Date: September 20, 1996
   ) Time: 12:00 p.m. ) Judge: Hon. Marilyn Hall Patel
   )
   
                               INTRODUCTION 
                                      
   
   When, if ever, may the U.S. Government require a private citizen to
   submit his ideas to a government agency for review and licensing prior
   to publication or public discussion? When, if ever, may the Government
   use such a process to censor publication of ideas about an entire
   subject area, namely the science of cryptography?
   Plaintiff PROFESSOR DANIEL BERNSTEIN ("PROF. BERNSTEIN") wishes to
   publish his own ideas about the science of cryptography in print, in
   the classroom and on the Internet. He seeks to engage in the free
   exchange of scientific ideas protected by both the freedom of speech
   and by academic freedom, "a special concern of the First Amendment."
   Keyishian v. Bd. of Regents, 385 U.S. 589, 603 (1967). The undisputed
   effect of the Arms Export Control Act, 22 U.S.C. §2778 et. seq.
   ("AECA") and its implementing regulations, the International Traffic
   in Arms Regulations ("ITAR") (hereinafter collectively referred to as
   the "ITAR Scheme"), is to require PROF. BERNSTEIN and others to submit
   their ideas about cryptography to the government for review, to
   register as arms dealers, and to apply for and obtain from the
   government a license to publish their ideas. Failure to do so results
   in severe civil and criminal penalties. In short, when ones ideas fall
   under the ITAR, instead of "publish or perish," the ITAR provides
   "publish and perish."
   In the first phase of this litigation, the Government argued that
   since Plaintiff's ideas were expressed, in part, in source code, they
   were not protected by the First Amendment. This Court rejected that
   argument and held that source code is speech for purposes of the First
   Amendment. Opinion on Motion to Dismiss, April 15, 1996 at 20:20-22.
   (hereinafter "Opinion"). But the ITAR Scheme controls more than just
   source code; its definition of software extends into algorithms and
   design structures, the very building blocks of mathematics and
   computer science. In addition, the Ninth Circuit recognized nearly 20
   years ago that the ITAR Scheme also controls much protected speech
   through its control of "technical data." United States v. Edler
   Industries, Inc., 579 F.2d 516, 520 (9th Cir. 1978). Accordingly,
   since it censors speech about cryptography both as software and as
   technical data, the ITAR Scheme must pass several stringent First
   Amendment tests.
   This Motion first seeks summary adjudication that the ITAR Scheme, to
   the extent it controls speech about cryptography, must meet stringent
   First Amendment scrutiny. Plaintiff argues first, that as to the prior
   restraints, the ITAR Scheme must pass the test provided in New York
   Times Co. v. United States, 403 U.S. 713, 730 (1971) (Stewart J., with
   whom White, J. joins, concurring)(hereinafter "Pentagon Papers").
   Second, as to application of the ITAR Scheme as subsequent punishment
   for speech about cryptography, Plaintiff submits that the strict
   scrutiny tests apply. E.g. Austin v. Michigan Chamber of Commerce 494
   U.S. 652, 654-655 (1990). This Motion also seeks partial summary
   judgment that there is no genuine issue of material fact that the ITAR
   Scheme, to the extent it controls speech about cryptography, fails
   three other standards of First Amendment jurisprudence: it lacks the
   procedural safeguards required by Freedman v. Maryland, 380 U.S. 51
   (1965); it is vague; and it is overbroad. The common principle which
   runs through these three doctrines, and which the ITAR violates, is
   the requirement that grants of administrative discretion be limited by
   clear standards and that judicial review be available. With no
   limitations on administrative action, the ITAR Scheme "readily lends
   itself to harsh and discriminatory enforcement," Thornhill v. Alabama,
   310 U.S. 88, 97 (1940) and "creates a threat of censorship that by its
   very existence chills free speech." Sec'y of State of Maryland v.
   Joseph H. Munson Co., 467 U.S. 947, 964 n. 12 (1984). Quite simply,
   the ITAR Scheme allows its administrative agencies to make
   inconsistent, incorrect and sometimes incomprehensible decisions
   censoring speech, all without the protections of judicial review or
   oversight.
   STATEMENT OF FACTS
   At the time of filing, Plaintiff was a Ph.D. candidate in mathematics
   at the University of California at Berkeley. He has since received his
   degree. He now teaches in the Department of Mathematics, Statistics
   and Computer Science at the University of Illinois at Chicago. PROF.
   BERNSTEIN has conducted research in the field of cryptography, the art
   and science of keeping messages secure. His work was developed on his
   own, without government funds or support, and is not classified. He
   intends to present his ideas, in all their forms, to the worldwide
   academic and scientific community. Despite this undisputed scientific,
   nonmilitary intention, Defendants have told Plaintiff that he cannot
   present his ideas.
   PROF. BERNSTEIN's work in cryptography included developing an
   algorithm called Snuffle for scrambling and descrambling
   communications. He wrote one description of the Snuffle algorithm in
   English using mathematical formulas (hereinafter "Paper"). In order to
   express it as precisely as possible, he also wrote a description of
   the algorithm in a computer language. He called the final versions of
   these computer language descriptions Snuffle.c and Unsnuffle.c.
   (Collectively referred to hereinafter as "Snuffle 5.0"). Bernstein
   Decl. ¶¶6-7. Professors Abelson and Sussman of the Massachusetts
   Institute of Technology, have explained why algorithms are often
   written in programming languages :
   Just as everyday thoughts are expressed in natural language, and
   formal deductions are expressed in mathematical language,
   methodological thoughts are expressed in programming languages. A
   programming language is a medium for communicating methods, not just a
   means for getting a computer to perform operations---programs are
   written for people to read as much as they are written for machines to
   execute.
   
   Abelson and Sussman, Structure and Interpretation of Computer
   Programs, preface, page xv. (1985). See also Declaration of Harold
   Abelson, ¶ 3-17. This view is also shared by the author of the "Bible"
   of computer programming, Professor Donald E. Knuth of Stanford
   University, who has written:
   At first I thought programming was primarily analogous to musical
   composition -- to the creation of intricate patterns, which are meant
   to be performed. But lately I have come to realize that a far better
   analogy is available: Programming is best regarded as the process of
   creating works of literature, which are meant to be read.
   
   Literature of the program genre is performable by machines, but that
   it not its main purpose. Computer programs that are truly beautiful,
   useful, and profitable must be readable by people. So we ought to
   address them to people, not to machines.
   
   Knuth, Literate Programming, IX (emphasis in original). See
   Declaration of Carl M. Ellison ¶ 12. This view of computer programming
   languages as languages which fundamentally communicate between humans
   is widely recognized. For this reason, computer programs have been
   published in scientific journals for peer readership for over 25
   years. Ellison Decl. ¶¶15-23.
   PROF. BERNSTEIN wants to publish his cryptographic ideas as part of
   the normal process of academic, scientific and political exchange of
   ideas and information. This process lies at the heart of First
   Amendment as well as the scientific method, which requires that new
   ideas be continually tested and discussed in the "marketplace of
   ideas." Bernstein Decl. ¶¶46, 52; Appel Decl. ¶¶ 3-5, 17. Because the
   ideas developed in many areas of science involve computer code, the
   scientific method includes publication of such code. 
   Today, moreover, scientific publication includes Internet publication.
   PROF. BERNSTEIN also wishes to publish his ideas in "sci.crypt," an
   Internet discussion group about
   cryptography. The Internet has become a significant place to publish
   all types of academic and scientific ideas. Publishing computer code
   on the Internet enables not only the evaluation of ideas, but also
   their incremental improvement.
   Moreover, Plaintiff will teach a course on cryptography next year.
   Bernstein Decl. ¶¶53-63. Whether in class, during office hours, or on
   the Internet, he must be free to discuss cryptography with his
   students without a license. University classes in computer-related
   fields often have course syllabi, assignments, and materials available
   on the Internet; some require students to publish on the Internet.
   Miller Decl. ¶¶4-6. In addition, Plaintiff must be free to test out
   his ideas on colleagues before he finally decides to publish them.
   
                    STATUTORY AND REGULATORY FRAMEWORK 
                                      
   Section 38 of the AECA authorizes executive control of commercial
   "export" of "defense articles and defense services" through the power
   to designate "items" on the U.S. Munitions List ("USML") and to
   promulgate the ITAR. 22 U.S.C. § 2778(a)(1). The ITAR Scheme makes it
   unlawful "[t]o export or attempt to export from the United States any
   defense article or technical data or to furnish any defense service
   for which a license or written approval is required by this subchapter
   without first obtaining the required license or written approval" from
   the State Department ("State"). 22 C.F.R. § 127.1(a)(1); see 22 U.S.C.
   § 2778(b)(2). This basic premise is fairly straightforward. From this
   point, however, sorting out the various ITAR definitions and State
   interpretations of the text becomes difficult, if not impossible.
   A. Key Definitions in the ITAR Scheme As It Relates to Plaintiff's
   Claims
   The ITAR Scheme definitions which are most important to Plaintiff's
   claims are those of "defense articles and defense services,"
   "software," "technical data," and "export." It is through the
   interactions of these that protected expression is restrained and that
   the problems of excessive discretion, vagueness and overbreadth are
   most obviously manifest. For example, key definitions such as "defense
   article," "technical data" and "defense services" refer to each other
   in circular fashion. Yet these very labels determine what activities
   constitute "export." 
   1. Defense Articles and Defense Services
   "'[D]efense articles and defense services' means, with respect to
   commercial exports subject to [22 U.S.C. §2778], those items
   designated . . . pursuant to" the USML. 22 U.S.C. § 2794(7). The USML
   includes "software with the capability of maintaining secrecy or
   confidentiality of information or information systems." 22 C.F.R.
   §121.1 (XIII)(b)(1). Thus cryptographic software is apparently a
   defense article. Technical data is also included as a defense article.
   22 C.F.R. §120.6; 123.1(b) and §123.1(e). Finally, defense services
   are defined as "furnishing assistance" or technical data to foreign
   persons. 22 C.F.R. §120.9(1) and (2).
   2. Software
   Software "includes but is not limited to the system functional design,
   logic flow, algorithms, application programs, operating systems and
   support software for design, implementation, test, operation,
   diagnosis and repair." 22 C.F.R. § 121.8(f). Thus, contrary to most
   reasonable expectations, software is not limited to computer code. It
   also includes system designs, algorithms and logic flows, the very
   building blocks of many mathematical concepts and scientific ideas,
   and by its own language "is not limited" to just those.
   3. Technical Data 
   "Technical data" appears several places throughout the ITAR Scheme,
   creating confusing circularity. As mentioned above, the definition of
   defense articles includes technical data, thus making technical data
   on its face is a subset of defense articles. Furthermore, defense
   articles include "technical data recorded or stored in any physical
   form," so a book or diskette on which technical data is recorded is
   apparently a defense article.
   Technical data itself includes several different types of ideas and
   information. The relevant definitions of technical data are:
   
   Information, other than software as defined in § 120.10(d) [sic],
   which is required for the design, development, production,
   manufacture, assembly, operation, repair, testing, maintenance or
   modification of defense articles. This includes information in the
   form of blueprints, drawings, photographs, plans, instructions and
   documentation. . . .
   
   Software as defined in § 121.8(f) of this subchapter directly related
   to defense articles[.]
   
   22 C.F.R. § 120.10(a)(1), (a)(4). Because software is defined as
   technical data, cryptographic software appears to be both a defense
   article and as technical data, and the "furnishing" of software is a
   defense service.
   4. Export
   The AECA does not define "export." In the ITAR, despite the overlap
   between defense articles and technical data noted above, the meaning
   of "export" for each differs significantly. "Export" of defense
   articles is defined as "sending or taking a 'defense article' out of
   the United States in any manner, except by mere travel outside of the
   United States by a person whose personal knowledge includes technical
   data." "Export" of technical data includes disclosing orally or
   visually or transferring technical data to any foreign person, whether
   abroad or in the United States. 22 C.F.R. § 120.17(a)(1), (a)(4).
   (Emphasis added)
   B. How The ITAR Scheme Was Applied To PROFESSOR BERNSTEIN
   Vaguely aware of the Defendants' restrictions on cryptography,
   Plaintiff contacted State to learn whether he needed a license to
   publish his ideas. He was told to submit a Commodity Jurisdiction
   ("CJ") request per former provision 22 C.F.R. § 120.4, which he did.
   Bernstein Decl. ¶17. In response to this first CJ, Defendants stated
   that a license was required. Plaintiff then engaged in a protracted,
   frustrating and ultimately unsuccessful attempt to learn how the
   Defendants interpreted the ITAR Scheme. Bernstein Decl. ¶¶16-42. See
   Miller Decl. ¶¶ 14-45; Junger Decl. ¶¶ 8-16; Johnson Decl. ¶¶22-33;
   Zimmermann Decl. ¶¶17-24 and NRC Report at 4-14 to 4-18, 4-30 to 4-33
   and 4-47.
   Eventually Plaintiff submitted a second round of five separate CJ
   Requests, each of which asked if he could publish a different item
   directly related to Snuffle. Plaintiff sought to give the Defendants
   the opportunity to consider each item separately. Bernstein Decl.
   ¶¶34-36. Defendants informed Plaintiff that all of the items were
   "subject to the licensing jurisdiction of the Department of State."
   Bernstein Decl. ¶37.
   Over two years later and after having been sued, Defendants issued a
   "clarification" letter directly rejecting the clear meaning of their
   previous responses. Defendants belatedly indicated that the source
   code items were defense articles, the Instructions "appeared to be"
   technical data and the Paper, over which they had first asserted
   control in 1992, "appeared" not to be controlled at all. Bernstein
   Declaration, ¶41.
   C. Impact of the ITAR Scheme
   The facial reach of Defendants' determinations is staggering. PROF.
   BERNSTEIN may not take or send Snuffle 5.0 source code outside of the
   U.S. in any manner. He may not present Snuffle 5.0 to a conference
   abroad or communicate it privately to an overseas colleague (even a
   U.S. citizen). See also Junger Decl. ¶¶ 18-23. He may not publish
   Snuffle 5.0 on the Internet in any form. See also Demberger Decl. ¶¶
   2-13 and Miller Decl.¶¶ 14-17, 20. Although the regulations are vague
   on this point, it appears that he may not even publish it in print,
   since published journals or books are inevitably sent abroad. See also
   Junger Decl. ¶¶ 23-28. In short, he is prevented from placing his
   source code into the "marketplace of ideas."
   As to the Instructions, which Defendants have apparently belatedly
   designated as "technical data," the restraint is similarly broad.
   Disclosing technical data to a foreign person anywhere is "export,"
   sweeping in face-to-face teaching of a foreign student, phone calls
   with foreign colleagues, presentations at conferences and electronic
   or print publication. See also Junger Decl. ¶¶ 18-28. In each
   instance, technical data would inevitably be disclosed to a foreign
   person. Even displaying technical data in this Court's public
   courtroom as part of this case is "export" if a foreign person is
   seated in the gallery.
   Finally because the labels of defense article and technical data are
   not mutually exclusive, Snuffle 5.0 source code is facially subject to
   the "technical data" export provisions as well, while the Instructions
   are facially subject to the "defense article" export provisions, and
   any communication of either is a "defense service". In sum, regardless
   of whether scientific information is labeled as a "defense article,"
   "technical data," or "defense service," the ITAR significantly
   restricts its communication.
   
                                 ARGUMENT 
                                      
   I. THE ITAR SCHEME REGULATES SPEECH. 
   A. The Controlled Items Are Speech
   Since this Court has held that source code is speech for First
   Amendment purposes, the Snuffle 5.0 source code at issue here is
   speech. There is no dispute that the Instructions, now evidently
   declared "technical data" by Defendants, are speech. Finally, the
   Snuffle Paper was recognized as speech by this Court. Opinion
   14:17-23. Since this Motion attacks the facial overbreadth, vagueness
   and lack of procedural protections in the ITAR Scheme, each of these
   items is part of the "speech." Indeed, the fact that Defendants
   improperly applied these regulations to Paper, an obvious speech item,
   is a perfect example of why the ITAR is so constitutionally offensive.
   
   B. On Its Face the ITAR Scheme Regulates Scientific Speech
   Scientific speech is entitled to the highest degree of First Amendment
   protection. The ITAR Scheme seeks to license scientific expression in
   three ways. First, the technical data provision reaches scientific
   expression by including "information in the form of blueprints,
   drawings, photographs, plans, instructions and documentation" relating
   to defense articles. 22 C.F.R. § 120.10(a)(1). Second, the regulation
   of software includes speech since computer software in source code
   form is speech. Opinion at 20:20-22. Third, the ITAR Scheme seeks to
   control more than code in its definition of software, including
   "design", "logic flow" and "algorithms" as part of the definition and
   thus sweeping in the very building blocks of much mathematics and
   computer science.
   Defendants have argued that they only seek to restrict ideas written
   in programming languages. While Plaintiff disputes this, even if it
   were true, speech is burdened. In conditioning publication of ideas
   upon their not being written in a programming language, Defendants
   seek to prescribe the form and content of expression. Expressing
   algorithms in a computer language increases precision of
   communication.
   Each step of an algorithm must be precisely defined; the actions to be
   carried out must be rigorously and unambiguously specified for each
   case. The algorithms of this book will hopefully meet this criterion,
   but since they are specified in the English language, there is a
   possibility the reader might not understand exactly what the author
   intended. To get around this difficulty, formally defined 'programming
   languages' or 'computer languages' are designed for specifying
   algorithms, in which every statement has a very definite meaning. Many
   of the algorithms in this book will be given both in English and in a
   computer language.
   
   D. Knuth, 1 The Art of Computer Programming: Fundamental Algorithms 5
   (1968).
   To write an algorithm as a computer program, therefore, is to make a
   choice among various ways of saying something. The Government has no
   authority to make that choice for private citizens. "[T]hat a speaker
   has the autonomy to choose the content of his own message" is "the
   fundamental rule of protection under the First Amendment." Hurley v.
   Irish-American Gay, Lesbian and Bisexual Group of Boston, 115 S.Ct.
   2338, 2347 (1995).
   C. On its Face and as Applied to PROF. BERNSTEIN, the ITAR Scheme
   Infringes Academic Freedom 
   Academic freedom "has long been viewed as a special concern of the
   First Amendment." Regents of the University of California v. Bakke,
   438 U.S. 265, 312 (1978); Keyishian, 385 U.S. at 603. Academic freedom
   protects more than just what Prof. Bernstein wants to teach in the
   classroom; it protects his right to publish books and articles and to
   exchange the fruits of his work and research with others. See Gordon &
   Breach Science Publishers v. AIP, 859 F.Supp. 1521, 1541 (S.D.N.Y.
   1994) ("debate . . . in academic journals" is "near the core of the
   First Amendment"). In Bakke, the Supreme Court noted:
   It is the business of a university to provide that atmosphere which is
   most conducive to speculation, experiment and creation. It is an
   atmosphere in which there prevail the "four essential freedoms" of a
   university - to determine for itself on academic grounds who may
   teach, what may be taught, how it shall be taught, and who may be
   admitted to study.
   
   Bakke, 438 U.S. at 312 quoting Sweezy, 354 U.S. at 263 (Frankfurter,
   J., concurring).
   Scientific and technical exchange of information is also protected as
   part of academic freedom. "The classroom is peculiarly the
   'marketplace of ideas.'" Keyishian, 385 U.S. at 603. "Teachers and
   students must always remain free to inquire, to study and to evaluate,
   to gain new maturity and understanding." Sweezy v. New Hampshire, 354
   U.S. 234, 250 (1957)(quotation and citation omitted); see Rosenberger
   v. Univ. of Virginia, 115 S.Ct. 2510, 2520 (1995) (university setting
   "at the center of our intellectual and philosophic tradition").
   In a 1981 memorandum, the Justice Department Office of Legal Counsel
   ("OLC") warned that the ITAR scheme's coverage "is so broad that the
   revised provisions could be applied to persons who are not directly
   connected or concerned in any way with any foreign conduct carrying
   dangerous potential for the United States . . . [such as] to
   communications of unclassified information by a technical lecturer at
   a university." Tien Decl., Exhibit A, OLC memo for Robinson at 212
   (Bates stamp page 60057); see also Bates stamp pages 60007-8 and
   600017-18. That scenario is before this Court. Prof. Bernstein's work
   is neither classified nor directly connected with any such foreign
   conduct, yet it is restricted by the ITAR Scheme.
   The ITAR Scheme violates all four freedoms raised in Bakke. "The heart
   of the system consists in the right of the individual faculty member
   to teach, carry on research, and publish without interference from the
   government." Dow Chemical Co. v. Allen, 672 F.2d 1262, 1275 (7th Cir.
   1982) (quotation and citation omitted). Prof. Bernstein is not free to
   teach cryptography by teaching about the development and analysis of
   cryptographic computer programs; he needs Defendants' approval and a
   government license which he cannot practically obtain and which is
   revocable at whim.
   D. Foreign Persons Have First Amendment Rights to Speak and Receive
   Information Which Are Abridged by the ITAR Scheme's Discrimination
   Against Them 
   The ITAR Scheme violates two more of Bakke's "four essential freedoms"
   by restricting who may teach and who may be admitted to study. By
   doing so, it reaches overbroadly into the speech rights of foreign
   persons. See American-Arab Anti-Discrimination Committee, et al v.
   Reno, 70 F.3d 1045, 1063 (9th Cir. 1995). Under the ITAR Scheme, no
   foreign person, not even a foreign Nobel prize winner teaching at
   Stanford, may obtain an export license. 22 U.S.C. §2778(g). Moreover,
   because a U.S. speaker "exports" by disclosing technical data to a
   "foreign person," foreign persons' First Amendment rights to receive
   information are restricted. As a result, PROF. BERNSTEIN might have to
   exclude foreign students from his cryptography course in order to
   engage in the open exchange of cryptography ideas in the classroom.
   Yet the Supreme Court long ago held that "where a [willing] speaker
   exists . . . the protection afforded is to the communication and to
   its source and recipients both." Virginia Pharmacy Bd. v. Virginia
   Consumer Council, 425 U.S. 748, 756 (1976) (rights of consumers to
   receive price advertising); See United States v. National Treasury
   Employees Union, 115 S.Ct. 1003, 1015 (1995). ("NTEU")
   E. By Restricting the Ability of People to Protect the Privacy of
   Their Communications, the ITAR Scheme Chills Speech 
   
   Freedom to speak would mean little if one could not control who could
   overhear the speech. Thus, the First Amendment includes the right to
   speak confidentially. See United States v. United States District
   Court, 407 U.S. 297, 314 (1972). It prevents compelled speech. Hurley
   v. Irish-American Gay Group of Boston, 115 S.Ct. 2338 (1995). It
   protects anonymous speech. McIntyre v. Ohio Elections Com'n, 115 S.Ct.
   1511, 1516 (1995). It prevents compelled disclosure of those with whom
   one associates and speaks. NAACP v. Alabama ex rel. Patterson, 357
   U.S. 449, 462 (1958). It extends to a person's right to communicate
   with foreigners. Bullfrog
   Films, Inc. v. Wick, 842 F.2d 502, 509, n.9 (9th Cir. 1988) (remanded
   on other grounds, 959 F.2d. 778 (9th Cir. 1992)).
   For electronic speech, encryption is like an "envelope" which may only
   be opened by authorized recipients. There can be little doubt that a
   statute or regulation requiring persons using the U.S. mails to write
   on postcards or restricting the availability of envelopes to licensed
   individuals would significantly affect the content of mailed
   expression. Without envelopes, people would only express facts or
   ideas that they were willing to let anyone read. Thus, cryptography,
   like paper and ink, is inherently imbued with First Amendment
   significance. Cf. Minneapolis Star & Tribune Co. v. Minnesota Comm'r
   of Revenue, 460 U.S. 575, 585 (1983) (invalidating "use tax" on paper
   and ink products consumed in newspaper publication).
   II. THE ITAR SCHEME IS A PRIOR RESTRAINT WHICH FAILS THE HEAVY
   PRESUMPTION AGAINST CONSTITUTIONAL VALIDITY THAT THE FIRST AMENDMENT
   DEMANDS
   
   Any scheme that requires a permit or license for speech which may be
   granted or withheld in the discretion of an official makes the
   exercise of First Amendment freedoms contingent upon the uncontrolled
   will of an official and is a . . . prior restraint upon the enjoyment
   of such freedoms. Staub v. Baxley, 355 U.S. 313, 322 (1958). Those
   schemes represent "the most serious and least tolerable infringement
   on First Amendment rights." (Nebraska Press Ass'n v. Stuart, 427 U.S.
   at 539, 559 (1976). As a result, any system of prior restraint of
   expression comes bearing a heavy presumption against its
   constitutional validity. Pentagon Papers, 403 U.S. at 714 (quotation
   and citations omitted)(emphasis added). This heavy presumption cannot
   be overturned and publication cannot be prevented -- even as to
   matters involving military affairs -- unless it is shown that
   publication will "surely result in direct, immediate, and irreparable
   damage to our Nation or its people." Id., at 730 (Stewart, J.
   concurring). In addition, licensing schemes requiring prior government
   approval of communication are facially void if impermissibly vague or
   excessive in reach. Lovell v. City of Griffin, 303 U.S. 444, 452-53
   (1938).
   The ITAR, which requires a license prior to publication, is just such
   a prior restraint. It fails to overcome the "unusually heavy
   presumption against constitutional validity" required by the First
   Amendment.
   A. THE SCHEME CONTAINS AT LEAST FOUR PRIOR RESTRAINTS
   1. The CJ Process is a Prior Restraint 
   As noted above, figuring out what one can and cannot do under ITAR
   involves entering a circular maze of definitions. See supra at pages
   5-8. Because the reach of the ITAR scheme is so unclear, its
   requirements so onerous, and liability both criminal and civil so
   severe, individuals have a powerful incentive to invoke the CJ
   procedure if there is even the slightest possibility that the
   Government could later determine that their work is, or is "directly
   related to" an item on the USML. See Pittsburgh Press Co. v.
   Pittsburgh Comm'n on Human Relations, 413 U.S. 376, 390 (1973) (a
   "special vice" of prior restraint is "inducing excessive caution in
   the speaker"). PROF. BERNSTEIN, like many others, was unsure about the
   reach of the ITAR Scheme. When he asked Defendants whether he could
   publish his ideas, PROF. BERNSTEIN was advised to use the Commodity
   Jurisdiction process. Later, when he submitted his items, he was
   advised that a license was required. However, even if PROF. BERNSTEIN
   had learned that no license was needed, he would still have
   experienced significant pre-publication delay. Thus the CJ process
   itself is a prior restraint, whether or not something is ultimately
   deemed controlled, because it delays publication. Freedman, supra, 380
   U.S. at 57 (delay is a form of unbridled discretion).
   2. The Registration and Fee System Is a Prior Restraint. 
   When Defendants determined that he must seek a license, PROF.
   BERNSTEIN was first required to register as an arms exporter.
   Registration amounts to a license to seek a license; one may not apply
   for a license without first registering (22 C.F.R. §§122.1(c),
   120.4(b)) and paying fees. 22 C.F.R. §§ 122.2(a), 122.3(a). This
   payment requirement alone creates a prior restraint. See Forsyth
   County v. Nationalist Movement, 505 U.S. 123, 130 (1992). At best,
   registration adds even more delay; at worst, Defendants have
   unreviewable discretion in deciding whether to accept registration
   (see discussion, infra, at 18).
   Also, registration is expressly "a means to provide the U.S.
   Government with necessary information about who is involved in certain
   manufacturing and exporting activities." 22 C.F.R.§ 122.1(c). This
   requirement makes no sense for Plaintiff, a math professor, and acts
   as a prior restraint by infringing upon his right to publish his ideas
   without telling the government about it. See McIntyre 115 S.Ct. at
   1516.
   3. The Licensing Decision Is a Prior Restraint. 
   Plaintiff did not apply for a license to publish. An application would
   have been futile. An export license application for defense articles
   must designate the "country of ultimate destination;" the exporter
   "must ascertain the specific end-user and end-use prior to submitting
   an application," 22 C.F.R. § 123.9(a); See also id. at § 123.21(a).
   This cannot be done when one publishes source code in a book, journal
   or online discussion group, since one does not know who will receive
   it. As a practical matter, then, it is impossible to comply with the
   licensing requirement when one wishes to publish.
   Equally important, no standards control Defendants' licensing
   decision. "[A]ny application for an export license or other approval
   under this subchapter may be disapproved . . . whenever" the State
   Department "deems such action to be in the furtherance of world peace,
   the national security or the foreign policy of the United States, or
   is otherwise advisable." 22 C.F.R. §126.7(a)(1). Thus, CJ and export
   license determinations are prior restraints under Staub (355 U.S. at
   322), as they are completely discretionary.
   4. The ITAR's Reporting/Recordkeeping Mandate Is a Prior Restraint.
   
   Even when an export license is granted, the ITAR impose numerous
   requirements which cannot practically be met when the speaker wishes
   to publish. In other cases, the export need not be licensed, but the
   regulations nonetheless burden communication.
   For example, shipments of defense articles generally require a
   Shipper's Export Declaration. 22 C.F.R. § 123.16(a) ("SED"). If
   Plaintiff wishes to mail a copy of Snuffle 5.0 or the Instructions to
   an overseas colleague, he must show when and how the export took
   place. 22 C.F.R. § 123.22(e). The SED must be authenticated by either
   the postmaster or Plaintiff, and he must return the SED to State when
   the mailings are completed. 22 C.F.R. § 123.24. Also, State has the
   discretion to require that a nontransfer and use certificate be filed
   in any export. This certificate must be executed by the foreign
   consignee and foreign end-user, prior to issuance of a license and
   license applicant. State may also, in its discretion, require that the
   "appropriate authority" of the foreign end-users own government sign
   the certificate before a license is issued. 22 C.F.R. § 123.10(c).
   Even if a license is granted, the ITAR's recordkeeping requirements
   present another prior restraint. Exporters must maintain records about
   their activities for five years after expiration of a license. 22
   C.F.R.§ 122.5(a). This recordkeeping requirement includes not only
   exports but the "disposition" of defense articles. Thus, PROF.
   BERNSTEIN must keep logs of his academic work and specify with whom
   any developments or refinements were discussed. Obviously it is a
   severe burden to comply with such onerous requirements in the context
   of academic publication.
   B. THE ITAR SCHEME MUST MEET PENTAGON PAPERS'
   MANDATE OF EXACTING SCRUTINY
   
   The "chief purpose" of the First Amendment is "to prevent previous
   restraints upon publication." Near v. Minnesota ex rel. Olson, 283
   U.S. 697, 713 (1931). Because the ITAR Scheme as extended to speech
   about cryptography is just such a previous restraint upon publication,
   it must pass exacting scrutiny to be upheld. See CBS, Inc. v. U.S.
   District Court, 729 F.2d 1174, 1178 (9th Cir. 1983) (standard by which
   prior restraints on speech are reviewed is an "extraordinarily
   exacting" one). Thus, Defendants must prove that in every
   communication restrained by the ITAR, including PROF. BERNSTEIN's
   speech: (1) there exists a real, articulable likelihood of "direct,
   immediate, and irreparable damage" (Pentagon Papers at 730)(Stewart J.
   concurring) and (2)) the regulation at issue is necessary to prevent
   the injury so established. Burson v. Freeman, 504 U.S. 191, 199
   (1992). Unless both mandates are satisfied, the ITAR Scheme cannot be
   applied to speech about cryptography.
   1. Defendants Must Show That "Direct, Immediate And Irreparable Damage
   Will Surely Result" If Publication Is Not Restrained. 
   
   The First Amendment's heavy presumption against the validity of prior
   restraints requires that Defendants offer more than averments that
   publication may result in an alleged harm. The Government must instead
   come forward with credible, factual evidence that publication of
   academic and scientific speech about cryptography, including source
   code, will surely result in direct, immediate, and irreparable damage
   to our Nation or its people. Pentagon Papers at 730 (Stewart, J.
   concurring). See Turner Broadcasting Systems, Inc. v. FCC, 114
   S.Ct.2445, 2470 (1994) ("When the Government defends a regulation on
   speech as a means to. . .prevent anticipated harms, it must do more
   than simply 'posit the existence of the disease sought to be cured.' .
   . .It must demonstrate that the recited harms are real, not merely
   conjectural."). This high standard is applied despite the recognized
   fact that:
   Much speech is dangerous. Chemists whose work might help someone
   build a bomb, political theorists whose papers might start political
   movements
   that lead to riots, speakers whose ideas attract violent protesters,
   all these and
   more leave loss in their wake.
   
   American Booksellers Ass'n, Inc. v. Hudnut, 771 F.2d 323, 333 (7th
   Cir. 1985), aff'd mem., 475 U.S. 1001, reh'g denied, 475 U.S. 1132
   (1986). In cases involving claims of national security, the Supreme
   Court long ago explained the nature of speech which might be properly
   restricted under this test as including "the sailing dates of
   transports or the number and location of troops" during wartime. Near,
   283 U.S. at 716.
   In Pentagon Papers a former Pentagon employee, Dr. Daniel Ellsberg
   gave to the media secret, classified government reports called the
   Pentagon Papers. The Papers recorded U.S. involvement in the Vietnam
   War and, in June 1971, the New York Times decided to publish them. The
   Government sought to enjoin publication, asserting that the release of
   that information would harm U.S. foreign intelligence gathering by:
   [m]inimiz[ing] our chance of successful interception. Cutting down
   successful interception by our communication intelligence will
   directly affect our military operations. Signal intelligence now gives
   direct support to our troops today, and saves many lives.
   
   Sims, Triangulating the Boundaries of the Pentagon Papers, 2 Wm. &
   Mary Bill of Rights J. 341, 449 (1993) (reprinting declassified
   portions of the government's brief in Pentagon Papers).
   However, while acknowledging that release of some information in the
   reports may harm national security concerns, the Court, on expedited
   review, declined the requested injunction, concluding that the
   Government had failed to establish a direct link between publication
   and a resulting immediate and irreparable injury to the U.S. or its
   citizenry:
   . . . I cannot say that disclosure of any of [the documents at issue]
   will surely result in direct, immediate, and irreparable damage to our
   Nation or its people. That being so, there can under the First
   Amendment be but one judicial resolution of the issues before us. I
   join the judgments of the Court.
   
   Pentagon Papers at 730 (Stewart, J. concurring).
   Here, too, Defendants have asserted that the government interest in
   the control of speech about cryptography by the ITAR is "the need to
   control the availability of cryptography from the United States so
   that foreign intelligence gathering functions are not harmed." Def.
   Memorandum In Support of Motion to Dismiss, at 13. There is, as yet,
   nothing in the record to establish a direct and immediate link between
   publication of academic and scientific speech relating to cryptography
   and irreparable damage to U.S. intelligence gathering. However, to
   survive exacting scrutiny, Defendants must also show that such efforts
   will surely result in direct, immediate and irreparable injury to our
   Nation or its people.
   Indeed, the Government arguably faces an even greater burden here,
   because it is not restraining specified and classified materials
   directly concerning an ongoing war, as the case in Pentagon Papers.
   Here the Government claims the general power to license or censor
   information on certain scientific subjects, including protected
   domestic academic activity and publication. Further, it claims this
   power when the nation is not at war. Keeping in mind the Supreme
   Court's admonition that "[T]he validity of a restraint on speech in
   each case depends on careful analysis of the particular
   circumstances," Speiser v. Randall, 357 U.S. 513, 521 (1958), this
   Court must determine whether the Government's justifications here are
   sufficient to justify a scheme with such widespread peacetime impact.
   See NTEU, 115 S.Ct. at 1014 (1995) ("widespread impact...gives rise to
   far more serious concerns than could any single [ ] [decision and]
   chills potential speech before it happens.")
   2. The ITAR Must Be Narrowly Drawn And
   Alleviate The Alleged Harm.
   
   Even if Defendants are able to satisfy Pentagon Papers' causation and
   "direct, immediate and irreparable damages" mandates, they must
   further "demonstrate. . .that the [ITAR scheme] will in fact alleviate
   these harms in a direct and material way" (Turner, 114 S.Ct. at 2450),
   and that the harm alleged "cannot be militated by less intrusive
   measures." CBS, Inc. v. Davis, 114 S.Ct. at 914 (Blackmun, J., in
   chambers).
   
   III. THE ITAR SCHEME IS ALSO A CONTENT-BASED SPEECH RESTRICTION WHICH
   MUST SURVIVE THE STRICT SCRUTINY THAT THE FIRST AMENDMENT DEMANDS
   
   Even if the requirements of Pentagon Papers are met, the subsequent
   punishment provisions of the ITAR, which also impact speech about
   cryptography, must pass constitutional scrutiny. Determining whether a
   governmental regulation of speech is permissible under the First
   Amendment depends, in part, on whether the restriction is
   content-based or content-neutral. A content-based restriction is
   subjected to strict scrutiny and is constitutional only if the
   government shows the "regulation is necessary to serve a compelling
   state interest and that it is narrowly drawn to achieve that end."
   Perry Educ. Ass'n. v. Perry Local Educators' Ass'n., 460 U.S. 37, 45
   (1983). See Burson v. Freeman, 504 U.S. at 197-198; Simon & Schuster,
   Inc. v. N.Y. Crime Victims Bd., 502 U.S. 105, 116-118 (1991).
   As explained below, the ITAR Scheme as it is applied to cryptography
   is the very definition of a content-based speech restriction. It must,
   therefore, withstand strict scrutiny to survive.
   A. THE ITAR SCHEME IS CONTENT-BASED
   1. The ITAR Scheme is Content-Discriminatory On Its Face
   A restriction against speech may be content-based either because it
   favors or disfavors certain viewpoints, or because it discriminates
   among various subjects. Consolidated Edison Co. v. Public Service
   Comm'n., 447 U.S. 530, 537 (1980). See Burson v. Freeman, 504 U.S. at
   197. On its face, the ITAR Scheme restrictions on cryptography treat
   both software and academic and scientific speech differently based on
   their content in at least three key respects.
   First, the ITAR Scheme is a content-based regulation because, at its
   most fundamental level, the scheme restricts academic and scientific
   discussion of the entire topic of cryptography. See Consolidated
   Edison, supra, 447 U.S. at 537. Whether defined as a defense article
   or technical data, cryptographic software, together with its essential
   academic and scientific building blocks (e.g., source code,
   algorithms, and logic flows), and instructions as to its use, are all
   swept within the ITAR Scheme's prohibition. As a result, as explained
   on pages 9-10, supra, one may not discuss these items in any form on
   the Internet, at an international academic conference, privately with
   a foreign colleague, or on its face even in the more traditional
   printed format. The ITAR Scheme is content-based because it removes
   discussion on the entire topic of cryptography from the "marketplace
   of ideas." "[A]bove all else, the First Amendment means that
   government has no power to restrict expression because of its message,
   its ideas, its subject matter, or its content." Police Department of
   Chicago v. Mosley, 408 U.S. 92, 95 (1972). By picking and choosing by
   subject matter that which is allowed and that which is not, the ITAR
   scheme is antithetical to the First Amendment.
   Second, as a related matter, the ITAR Scheme requires that Defendants
   to review software and academic and scientific information and
   appraise its possible uses, strength, and effectiveness, to determine
   whether or not it is subject to the ITAR Scheme. Such inquiries are
   inherently content-driven. In City of Cincinnati v. Discovery Network,
   Inc., 507 U.S. 410 (1993), Cincinnati sought to defend, against a
   First Amendment challenge, an ordinance banning news-racks
   distributing commercial publications from public sidewalks, while
   permitting similar news-racks purveying newspapers. The Court
   concluded that by any commonsense understanding of the term, the ban
   in this case is 'content-based' because whether any particular
   newsrack falls within the ban is determined by the content of the
   publication resting inside that newsrack. Id., 507 U.S. at 429. The
   same is true here. Under the ITAR Scheme, Defendants must review the
   content of the various expressions to determine whether or not they
   are prohibited.
   Finally, by prohibiting encryption software, the ITAR Scheme
   indirectly restricts a particular mode of expression -- private
   speech. Thus, in the same way as prohibiting whispering would force
   one to announce a private declaration of love to the world, or
   preventing the use of envelopes would compel an attorney to
   communicate with her foreign client by postcard, the ITAR Scheme
   restricts the sending of electronic communications in a mode that
   limits who can read them. However, our nation's courts have
   consistently held that singling out particular modes of expression
   amounts to content-based discrimination under the First Amendment. See
   McIntyre 515 U.S. at 1516. (First Amendment extends to anonymous
   speech); Yniguez v. Arizonian for Official English, 69 F.3d 920,
   934-936 (9th Cir. 1995) (en banc), cert granted, 64 U.S.L.W. 3639
   (U.S. Mar. 25, 1996) (No.95-974) (communications in language other
   than English held a protectible "mode of expression" under the First
   Amendment).
   2. The ITAR Scheme Contains the Facial Indications of Viewpoint
   Discrimination 
   As a related matter, the ITAR also contains facial indicia of
   viewpoint discrimination in at least two respects. First, defendants
   have singled out cryptography for differential treatment. 22 C.F.R. §
   121.8(f). This is so despite the existence of considerable interest
   in, and continuing public debate and governmental policy-making on,
   the subject of cryptography. See generally NRC Report, attached as
   Exhibit E to Tien Decl. at Preface, vii-xiii. Thus, export controls
   are a tool by which the government controls public views about
   cryptography. See id., at 4-1. It has long been held, however, that
   "power in government to channel the expression of views is
   unacceptable under the First Amendment." Bellotti, 435 U.S., at 785
   (footnote omitted). Viewpoint-based suppression presents a risk even
   more serious than content discrimination. R.A.V. v. City of St. Paul,
   505 U.S. 377, 391 (1992) (prohibiting "fighting words" on disfavored
   topics "goes even beyond mere content discrimination, to actual
   viewpoint discrimination"). The government is not neutral in this
   policy debate; it champions one type of encryption, "key escrow." NRC
   Report at xi, see generally Froomkin, The Metaphor is the Key:
   Cryptography, the Clipper Chip, and the Constitution, 143 U.Pa.L.Rev.
   709 (1995). As the Supreme Court noted in Bellotti, "[E]specially
   where . . . suppression of speech suggests an attempt to give one side
   of a debatable public question an advantage in expressing its views to
   the people, the First Amendment is plainly offended." Bellotti, 435
   U.S. at 785-789 (footnote omitted).
   Second, the ITAR permits dissemination of cryptography to the extent
   that discussion encompasses only "public domain" information and
   "information concerning general scientific, mathematical or
   engineering principles commonly taught in schools, colleges and
   universities." 22 C.P.R. § 120.10(a)(5). At the same time, however,
   the ITAR Scheme restricts discussion on the same topic of new ideas,
   technological advances, original work, and minority views not
   "commonly taught in schools, colleges, and universities." As such,
   from a scientific standpoint the ITAR is analogous to a regulatory
   scheme permitting the once-commonly taught view that the sun revolved
   around the earth, while at the same time prohibiting publication of
   Copernicus' new and different ideas.
   
   3. Defendants Cannot Recast The Content-Based ITAR As Content- Neutral
   Under The "Expressive Conduct" Doctrine of U.S. v. O'Brien
   Defendants attempt to transmogrify the content-based ITAR into
   content-neutral regulation aimed at expressive "activity" or "conduct"
   requiring less rigid constitutional analysis under United States v.
   O'Brien, 391 U.S. 367 4(1968) (burning draft card for expressive
   reasons). The assertion is based on Defendants' argument that
   Plaintiff's cryptographic software, its underlying academic and
   scientific foundations (e.g., source code, algorithms, and logic
   flows), and instructions as to its use, are functional "conduct," as
   opposed to pure speech. However O'Brien applies only when
   "[a]pplication of a facially neutral regulation . . . incidentally
   burdens speech." United States v. Albertini, 472 U.S. 675, 687 (1985).
   Plaintiff has already shown that the ITAR Scheme is not facially
   content neutral; that alone renders O'Brien inapplicable.
   In addition, Defendants' argument has already been rejected by this
   Court:
   Whether source code and object code are functional is immaterial to
   the analysis at this stage. Contrary to Defendants' suggestion, the
   functionality of a language does not make it any less like speech. . .
   . Thus, even if Snuffle source code, which is easily compiled into
   object code for the computer to read and easily used for encryption,
   is essentially functional, that does not remove it from the realm of
   speech. Instructions, do-it-yourself manuals, recipes, even technical
   information about hydrogen bomb construction, see United States v.
   Progressive, Inc., 467 F.Supp. 900 (W.D. Wisc. 1979), are often purely
   functional: they are also speech. Opinion, at pp.16-18.
   
   Moreover, the ITAR Scheme directly burdens speech because it is
   directed at communication by licensing "disclosure" by U.S. persons
   and forbidding "disclosure" by foreign persons. Defendants themselves
   have stated that the ITAR Scheme applies to "professional and academic
   presentation and informal discussions . . . constituting disclosure of
   cryptologic technical data to foreign nationals." Munitions Control
   Newsletter, attached at Tien Decl., Exhibit E, N.R.C. Report, Box
   4.11, (discussed at 4-30 to 4-31). Even the OLC agrees on this point:
   [E]ven a cursory reading of the technical data provisions reveals that
   those portions of the ITAR are directed at communication. A more
   stringent analysis than the O'Brien test is therefore mandated.
   
   Opinion, at p. 31, n. 17 (citing Tien Decl. Exh. A, at 60084, n. 16)
   (emphasis added). In short, because the ITAR Scheme is not content
   neutral and is directed at speech and communication and not conduct,
   O'Brien is not the appropriate test under which to evaluate
   Plaintiff's claims.
   B. THE ITAR SCHEME MUST PASS STRICT SCRUTINY
   
   As a content-based restriction, the ITAR must be subjected to the
   "compelling state
   interest" and "narrowly tailored" tests which comprise strict
   scrutiny.
   1. Defendants Must Show That The Application Of The ITAR Scheme To
   Speech About Cryptography Is Necessary To Further A Compelling State
   Interest.
   To justify a restriction against speech under strict scrutiny, it is
   not enough that the government identify some interest falling within
   the sphere of legitimate government concern. Rather, the asserted
   interest must be a compelling one. "Mere legislative preferences or
   beliefs respecting matters of public convenience may well support
   regulation directed at other personal activities, but be insufficient
   to justify such as it diminishes the exercise of rights so vital to
   the maintenance of democratic institutions. Schneider v.New Jersey,
   308 U.S. 147, 161 (1939), quoted in Schad v. Borough of Mt. Ephraim,
   452 U.S. 61, 69 (1981). The Government must be prepared to articulate,
   and support, a reasoned and significant basis for its decision. Schad
   at 77 (Blackmun, J. concurring). Furthermore, in light of the "danger
   of censorship" presented by a facially content-based statute, Leathers
   v. Medlock, 499 U.S. 439, 448 (1991), the law may be constitutionally
   employed only when it is "necessary to serve the asserted [compelling]
   interest." Burson v. Freeman, 504 U.S. at 199 (emphasis added).
   In Edler the Ninth Circuit explained the government interest in the
   ITAR Scheme: "controlling the conduct of assisting foreign enterprises
   to obtain military equipment and related technical expertise." Edler
   579 F.2d at 521. There is, as yet, no evidence in the record to
   support or even explain how this interest requires extending the reach
   of the ITAR Scheme to general academic and scientific publication on
   the subject of cryptography. See infra at pages 32-33. As the Supreme
   Court observed only weeks ago: " ...where, as here, the record before
   Congress or before an agency provides no convincing explanation, this
   Court has not been willing to stretch the limits of the plausible, to
   create hypothetical, non-obvious explanations in order to justify laws
   that impose significant restrictions upon speech." Denver Area
   Educational Telecommunications Consortium v. F.C.C. 1996 U.S. LEXIS
   4261 *58-59.
   2. The ITAR Must Be Narrowly Tailored
   Even if Defendants were able to satisfy the compelling interest test,
   they further must show that the ITAR scheme is "narrowly tailored,"
   i.e., that it is the least restrictive means of achieving its
   purposes. Boos v. Berry, 485 U.S. 312, 329 (1988). First, the law must
   actually advance the asserted interest, and the Court must determine
   for itself whether it does. See Eu v. San Francisco County Democratic
   Central Comm, 489 U.S. 214, 226-29 (1989); Bellotti, 435 U.S. at
   789-90. Second, the law must not be overinclusive, it must not
   restrict a significant amount of speech that does not implicate the
   government interest. Simon & Schuster, 112 S.Ct. at 511. See infra at
   37- 41. Third, there must be no less restrictive alternatives that
   would serve the interest approximately as well. Florida Star v. BJ.F.,
   491 U.S. 524, 538-40 (1989)(instead of banning class of statements,
   law might have to provide for case-by-case findings that statements
   indeed implicate government interest). Fourth, the law must not be
   underinclusive; it must not fail to restrict a significant amount of
   speech that harms the government interest about as much as the
   restricted speech does. See Florida Star, 491 U.S. at 540.
   Again, Defendants have presented no evidence to support the assertion
   that the ITAR is the least restrictive means of achieving its stated
   purpose. What is apparent that this stage of the proceedings, however,
   is that the ITAR has adopted the most restrictive approach of
   prohibiting publication and public communication about an entire
   subject area: the science of cryptography. "Broad prophylactic rules
   in the area of free expression are suspect. Precision of regulation
   must be the touchstone in an area so closely touching our most
   precious freedoms." NAACP v. Button, 371 U.S. 415, 438 (1963)
   (citations omitted).
   IV. THE ITAR SCHEME DOES NOT CONTAIN THE PROTECTIONS REQUIRED BY
   FREEDMAN V. MARYLAND. 
   
   Regardless of the substantive scrutiny applied, in order to withstand
   Constitutional challenge the ITAR's regulation of speech must be
   procedurally sound. In Freedman, 380 U.S. at 51, the Supreme Court
   held that a system of prior restraints for screening obscene films was
   constitutional only if accompanied by procedural safeguards necessary
   "to obviate the dangers of a censorship system." Id., at 58. Thus, a
   scheme of prior restraints such as the ITAR's application to speech
   about cryptography is invalid unless "accomplished with procedural
   safeguards that reduce the danger of suppressing constitutionally
   protected speech." Southeastern Promotions v. Conrad, 420 U.S. 546,
   559 (1975); Forsyth County, 505 U.S. at 131.
   A. The AECA Statute Does Not Control Administrative Discretion. 
   Where a law authorizes a system of prior restraints, the statutory
   delegation must provide "narrowly drawn, reasonable and definite
   standards for the [administering] officials to follow." Niemotko v.
   Maryland, 340 U.S. 268, 271 (1951) (striking conviction for absence of
   permit where permit administration based only on custom). In Lakewood
   v. Plain Dealer Publishing, 486 U.S. 750 (1988), the Supreme Court
   struck down a law granting standardless discretion in the granting of
   annual permits for newsracks on public property. "[I]n the area of
   free expression a licensing statute placing unbridled discretion in
   the hands of a government official or agency constitutes a prior
   restraint and may result in censorship." Lakewood, 486 U.S. at 757;
   American Jewish Congress v. City of Beverly Hills, No. 93-55085 (9th
   Cir. July 19, 1996 (en banc), 96 Daily Journal D.A.R. 8705, 8708-9.
   The AECA has no narrow, definite statutory standards. It states little
   more than a general policy to further "world peace and the security
   and foreign policy of the United States." 22 U.S.C. Sec. 2778(a)(1).
   No standards control what may be placed on the USML, because the USML
   simply consists of "'those items designated by the President" as being
   on the USML, 22 U.S.C. § 2794(7), and such designations are not
   "subject to judicial review." 22 U.S.C. § 2778(h). As a result,
   anything can be placed on the USML, and as previously explained, the
   licensing decision as to each applicant is also standardless. Thus the
   entire process is discretionary and presents a risk of censorship. The
   Government's treatment of PROF. BERNSTEIN and others demonstrates that
   this risk is not illusory.
   B. The ITAR Scheme Relies On Content Evaluation By The Licensor.
   The Lakewood court expressly relied on the problem of content
   evaluation in holding that an ordinance requiring annual permits for
   newsracks on public property required proper procedural protections.
   The annual renewal requirement permitted the licensor to measure the
   probable content or viewpoint of future expression by speech already
   uttered. Lakewood, 486 U.S. at 759- 760. Here, as described in page 23
   supra, Defendants actually review speech about cryptography and
   appraise its possible uses, strength and effectiveness. Thus the ITAR
   Scheme presents a much more egregious content evaluation and should be
   struck down.
   C. The ITAR Scheme Lacks The Freedman Procedural Safeguards. 
   In light of this Court's holding that source code is protected
   expression under the First Amendment, there can be no doubt that the
   standards of Freedman and Lakewood must be applied to the ITAR Scheme
   to the extent it controls speech. The prerequisites have been met:
   Defendants do not dispute that Plaintiff is subject to the
   regulations. Their letters to him require that he register as an arms
   dealer and apply for a license prior to the publication of Snuffle 5.0
   and technical data related to it.
   Finally, as explained below, even cursory review of the regulations
   reveals that they lack the procedural protections required by Freedman
   and its progeny. To be constitutional, the ITAR Scheme must "conform
   to procedures that will ensure against the curtailment of
   constitutionally protected expression," because "freedoms of
   expression must be ringed about with adequate bulwarks." Bantam Books,
   372 U.S. at 66 (1963). It must operate under judicial superintendence
   and assure an almost immediate judicial determination of the validity
   of the restraint. Id., at 70 (footnote omitted). However, the major
   flaw in the ITAR Scheme remains its lack of enforceable standards.
   "Even if judicial review were relatively speedy, such review cannot
   substitute for concrete standards to guide the decisionmaker's
   discretion." Lakewood, 486 U.S. at 771.
   1. There Is No Provision for Expeditious Judicial Review. 
   Essential to the scheme of procedural safeguards is expeditious
   judicial review; there must be a "prompt final judicial decision"
   reviewing any "interim and possibly erroneous denial of a license."
   Freedman, 380 U.S. at 59. "Because the censor's business is to censor
   . . . he may well be less responsive than a court . . . to the
   constitutionally protected interests in free expression." Id., at
   57-58 (footnote omitted).
   Neither AECA nor ITAR establishes any mechanism for judicial review of
   an interim restraint; in fact each purports to prohibit it. Once
   Defendants determined that Snuffle 5.0 was controlled, Plaintiff was
   no longer free to publish it. There must be judicial review of such a
   determination. Bantam Books, 372 U.S. at 71 (noting absence of "notice
   and hearing"). The Supreme Court has also added prompt appellate
   review to the Freedman list of procedural safeguards. Nat'l Socialist
   Party of America v. Village of Skokie, 432 U.S. 43, 44 (1977)
   (reversing state supreme court denial of stay of injunction against
   Nazi march). The ITAR Scheme plainly flunks this requirement as well.
   2. There Is No Requirement for an Expedited Determination. 
   The failure to confine the time within which the censor must make a
   decision "contains the same vice as a statute delegating excessive
   administrative discretion." Freedman, 380 U.S. at 56-57. The ITAR
   scheme fails this requirement, because neither the CJ nor the
   licensing procedure contains any meaningful time limits. First, the CJ
   process provides at most that the applicant for a CJ determination may
   seek expedited processing if no decision has been rendered within 45
   days.
   Second, the CJ determination only establishes the need to apply for a
   license. The ITAR contains no time limit for license decisions. 22
   C.F.R. § 126.7. One's application to publish under the ITAR, as with
   the application for a newspaper rack in Lakewood, "could languish
   indefinitely." Lakewood, 486 U.S. at 771. Indeed, even if a license is
   granted under the ITAR Scheme, it may be "revoked, suspended, or
   amended without prior notice." 22 C.F.R. § 126.7(a). 
   
   3. The Burden of Going to Court, And the Burden of Proof in Court
   Proceedings, Is on the Citizen Not the Government. 
   
   "[B]ecause only a judicial determination in an adversary proceeding
   ensures the necessary sensitivity to freedom of expression, only a
   procedure requiring a judicial determination suffices to impose a
   valid final restraint" Freedman, 380 U.S. at 58-59. "[T]he burden of
   instituting judicial proceedings, and of proving that the material is
   unprotected, must rest on the censor." Southeastern Promotions, 420
   U.S. at 560.
   AECA does not require Defendants to seek judicial review of their
   decisions to designate speech as controlled, or to deny a license for
   its publication. Instead, a speaker must, as Plaintiff has here,
   initiate legal action. Furthermore, because Defendants need not
   initiate judicial review, the AECA does not require Defendants to
   prove that the information may not be published. In fact, as shown
   above, the Government need not even explain their decisions until the
   discovery phase of formal litigation. This bureaucratic silence
   exacerbates the First Amendment problems. Lakewood, 486 U.S. at
   769-70.
   V. THE ITAR SCHEME IS UNCONSTITUTIONALLY VAGUE.
   "[S]tandards of permissible statutory vagueness are strict in the area
   of free expression." Button, 371 U.S. at 432. Although usually framed
   in terms of fair notice, the more important aspect of vagueness is the
   requirement that a legislature establish minimal guidelines to govern
   law enforcement. Kolender v. Lawson, 461 U.S. 352, 358 (1983). Where
   expression is licensed, "[t]he vice . . . is particularly pronounced"
   because "available judicial review is in effect rendered inoperative"
   by lack of standards. Interstate Circuit v. Dallas, 390 U.S. 676,
   683-685 (1968) (internal citations and quotations omitted). Thus,
   where First Amendment rights are involved, an even greater degree of
   specificity is required than the normal standard of adequate notice to
   a person of ordinary intelligence. Buckley v. Valeo, 424 U.S. 1, 77
   (1976) (citations omitted). The ITAR Scheme falls well below the
   required standard. As we have seen, not even Defendants seem to know
   what their regulations include and exclude.
   A. The AECA Lacks Standards Sufficient to Guide Administrative
   Discretion, to Aid a Court in Assessing the Exercise of Discretion or
   to Allow a Person To Know What the Regulations Regulate. 
   
   To the extent it licenses speech about cryptography, the AECA's lack
   of statutory standards, along with the unreviewability of USML
   designations, constitutes excessive discretion. The law requires "a
   precise statute evincing a legislative judgment that certain specific
   conduct be . . . proscribed . . . assur[ing] us that the legislature
   has focused on the First Amendment interests and determined that other
   governmental policies compel regulation." Grayned v. City of Rockford,
   408 U.S. 104, 109 n.5 (1972) (citation and quotation omitted).
   Congress has "impermissibly delegate[d] basic policy matters . . . for
   resolution on an ad hoc and subjective basis, with the attendant
   dangers of arbitrary and discriminatory application." Grayned, 408
   U.S. at 108-09.
   As noted above, the AECA provides no substantive standards for
   Defendants to follow or for a court to use in examining Defendants'
   exercise of that discretion to censor speech. Standards allow courts
   quickly and easily to determine whether the licensor is discriminating
   against disfavored speech. Without these guideposts, post hoc
   rationalizations by the licensing official and the use of shifting or
   illegitimate criteria are far too easy. Lakewood, 486 U.S. at 758.
   Here the lack of standards has allowed Defendants to misuse a statute
   aimed at commercial, military arms sales to limit academic and
   scientific publication. Even the OLC has expressed doubt as to whether
   Congress intended to control any academic speech:
   [i]t is by no means clear . . . that Congress intended that the
   President regulate noncommercial dissemination of information, or
   considered the problems such regulation would engender. We therefore
   have some doubt whether § 38 of the Arms Export Control Act provides
   adequate statutory authorization for the broad controls over public
   cryptography which the ITAR imposes.
   
   Tien Decl., Exhibit A, Harmon Memo, at 4 n.7 (Bates stamp 60077). In a
   later memorandum, the OLC argued that "[t]he technical data provision
   received scant attention," and noted: "it does not appear that
   Congress addressed itself to the scope of §414 of the Mutual Security
   Act." Tien Decl., Exhibit C, Memorandum from Kathryn Fuller to Wayne
   Kay, at 2, 3 ("Fuller Memo"). The Fuller Memo continues:
   [T]he legislative history underlying the authorization of regulation
   of 'technical data' provides little guidance as to the intended scope
   of that term. It seems clear that Congress contemplated the regulation
   of commercial exports of physical items such as blueprints and
   specifications. But it is not clear that Congress understood or
   intended its action to authorize controls over academic publication or
   oral communication of cryptographic ideas .... Although it might be
   argued that ... Congress was aware of and approved the existing
   provisions and interpretations of the ITAR, the sketchy and
   inconclusive legislative record pertaining to that section makes such
   an argument unpersuasive.
   
   Tien Decl., Exhibit C, Fuller Memo, at 3-4. Thus the lack of standards
   for its application to protected expression, and indeed the
   uncertainty about whether Congress ever intended the AECA to be
   applied to protected expression, create constitutional vagueness in
   the AECA.
   B. The ITAR Regulations Only Deepen The Vagueness. 
   The dangers of arbitrary and discriminatory enforcement against speech
   about cryptography created by the lack of standards in the AECA are
   manifest in the Government's implementation of the scheme. Defendants
   have defined "technical data" to include much scientific or technical
   information, including information about cryptography, and "export" to
   include all manner of communication and general publication. In
   reaching so broadly vagueness is enlarged by the ITAR, not confined.
   1. The USML Is Vague. 
   Even what is considered "cryptographic" is not clearly defined in the
   ITAR Scheme. The USML prohibits items with "the capability of
   maintaining secrecy," 22 C.F.R. §121.1 XIII(b)(1). This clause gives
   little notice to the person of common intelligence, and lacks the
   "greater degree of specificity" required where speech is concerned.
   The Government itself recognizes and admits that the USML is vague by
   the very creation of the CJ Process. The CJ Process is to be used "if
   doubt exists as to whether an article or service is covered" by the
   USML. 22 C.F.R. § 120.4(a). Such a "doubt" is a constitutional flaw
   when speech is regulated.
   2. The Categories of "Defense Articles," "Defense Services," and
   "Technical Data" Are Vague. 
   
   As previously noted at pages 5-8, the application and licensing
   requirements of the ITAR Scheme differ according to whether a defense
   article, defense service, or technical data is being exported. Yet
   each definition includes or is included by the other. See 22 C.F.R. §
   123, 120.17(a)(1) (defense articles); 124, 120.17(a)(5) (defense
   services); 125, 120.17(a)(4) (technical data). When terms are
   cross-referenced this way vagueness persists until each term is
   narrowed. As also noted above, for this reason alone, Defendants'
   reliance on Edler is misplaced; Edler only narrowed the definition of
   "technical data," leaving "defense articles" and "defense services"
   untouched.
   The ITAR's treatment of software is vague for another reason.
   "Software" normally refers to expression in a computer language.
   Opinion at 29 nn. 3 & 4. But under ITAR, software "includes but is not
   limited to" logic flow and algorithms. 22 C.F.R. § 121.8(f). This
   definition has no limit and easily includes mathematical equations or
   diagrams. Cf. Gottschalk v. Benson, 409 U.S. 63, 65 (1972)
   ("algorithm" is "[a] procedure for solving a given type of
   mathematical problem."). Thus, contrary to normal definitions of
   "software" the ITAR definition could include virtually any discussion
   of how to solve a mathematical problem which Defendants might construe
   as "cryptographic."
   3. The Exemptions are Vague. 
   The definition of technical data does exclude certain information
   concerning general scientific, mathematical or engineering principles
   commonly taught in schools, colleges and universities. 22 C.F.R.
   6§120.10(a)(5). This exemption only muddies the waters, however. What
   is a "general" principle; what does "commonly taught" mean? Suppose
   UC-Berkeley teaches a principle, but Stanford does not?
   Similarly, "information in the public domain" is not technical data
   and is "not subject to the controls of this subchapter." 22 C.F.R. §
   120.10(a)(5), 120.11, 125.1(a). Here lurks a subtle but critical
   Catch-22. To be "public domain," information must be "published" and
   "generally accessible or available" to the public in specified ways.
   22 C.F.R. § 120.11(a). But publishing for the first time is "export".
   How can technical data become generally available "information" when
   publishing it for the first time is a prohibited export?
   Furthermore, the Defendants' tortured interpretations of the ITAR
   create further uncertainties. For example, an ordinary person would
   believe that cryptographic source code could be "public domain," since
   it could be "information which is generally accessible or available to
   the public" under 22 C.F.R. §120.11 Yet Defendants apparently take the
   position that the public domain exception applies only to "technical
   data," despite the fact that it says it applies to all "information."
   Then they state that cryptographic software like Snuffle 5.0 is not
   technical data, and so cryptographic software can never enter the
   public domain. Demberger Decl. ¶¶ 3-4, 6-8.
   This position is bizarre, because the definition of "public domain"
   only refers to "information." It does not mention "technical data."
   The definition of "technical data" uses the phrase "[i]nformation,
   other than software," implying that software is information. 22 C.F.R.
   § 120.10(a)(1). So Defendants apparently contend that software is
   information, but cryptographic software is not.
   Whether or not Defendants' positions make sense, it is clear that the
   regulations fail to give fair notice of what is exempt. Equally
   important, they fail to confine Defendants' discretion as to what may
   be controlled and how. Such interpretive gerrymandering epitomizes the
   way a vague statute makes the use of "post hoc rationalizations" and
   "shifting or illegitimate criteria [] far too easy." Lakewood, 486
   U.S. at 758.
   In short, Defendants have defined "technical data" confusingly and
   defined exemptions from "technical data" vaguely. Even if each
   position were individually defensible, this system of regulation is as
   a whole too imprecise to pass First Amendment muster. See id., at 757,
   769-70.
   4. "Export" is Vague. 
   As noted above, AECA does not define "export"; the ITAR then creates a
   constitutional problem in this case by defining "export" to include
   communication and publication of speech about cryptography. No
   reasonable person would think that the statutory term "export" reached
   purely domestic communication of information. See United States v.
   Posey, 864 F.2d 1487, 1496-97 (9th Cir. 1989) (contrasting "export
   abroad" with "domestic disclosure"). Nor would a reasonable person
   think that "export" includes publishing information, even if that
   publication - like that of a newspaper or book or Internet publication
   - crossed the U.S. border. Indeed, if Congress had intended to reach
   publication, it could have done so explicitly in the AECA, as it has
   done in the espionage laws. See, e.g., 18 U.S.C. § 794(b) (publishing
   defense information in wartime). Despite this, Defendants state that
   "posting cryptographic software and related source code to
   internationally distributed Usenet newsgroups on the Internet
   constitutes an export." Demberger ¶¶ 4, 10.
   The circularity of the definitions of technical data, defense articles
   and defense services is crucial here, because whether a given action
   is an "export" depends on this categorization. See supra at pages 7-8.
   If you don't know which category your work is in, you don't know what
   counts as export and are left to guess. Speech cannot be punished when
   the rules are so unclear. Cf. Herndon v. Lowry, 301 U.S. 242, 261-63
   (1937). As in the scheme requiring teachers to sign loyalty oaths in
   Keyishian, the ITAR Scheme is a "regulatory maze" wholly lacking in
   terms susceptible of objective measurement in which vagueness of
   wording is aggravated by prolixity and profusion of statutes,
   regulations, and administrative machinery, and by manifold
   cross-references to interrelated enactments and rules. Keyishian, 385
   U.S. at 604. As in Keyishian, the Scheme must be struck down.
   VI. THE ITAR SCHEME IS OVERBROAD.
   A licensing scheme is facially overbroad when it "does not aim
   specifically at evils within the allowable area of [government]
   control, but . . . sweeps within its ambit other activities that ...
   constitute an exercise" of First Amendment rights. Thornhill, 310 U.S.
   at 97. The danger is "not merely the sporadic abuse of power by the
   censor" but the "continuous and pervasive restraint on all freedom of
   discussion that might reasonably be regarded as within its purview."
   Id., at 97-98. Almost 20 years ago, OLC concluded that the ITAR
   scheme's licensing standards "are not sufficiently precise to guard
   against arbitrary and inconsistent administrative action." Tien Decl.,
   Exh. A., Harmon Memo at 10 (Bates stamp 60083). In later memos, the
   OLC warned that the scheme's coverage was so broad that it could apply
   to "communication of unclassified information by a technical lecturer
   at a university or to the conversation of a United States engineer who
   meets with foreign friends at home to discuss matters of theoretical
   interest." Tien Decl. Exhibit A, 1984 ITAR Memorandum, page 2, (Bates
   Stamp 60007-8); See id. at pages 12-13 (Bates Stamp 60017-18).
   The OLC was right. The problems they identified have not been
   remedied. The ITAR Scheme's broad definitions confer enormous
   discretion on Defendants, yet its terms provide that a court may not
   review a determination that academic publication of scientific speech
   must be licensed.
   A. The Scheme Reaches A Substantial Amount Of Constitutionally
   Protected Activity. 
   
   As described in detail in Section I above, the ITAR Scheme reaches
   substantial First Amendment activities by both Plaintiff and others by
   licensing scientific and technical communication. See Village of
   Hoffman Estates v. Flipside, 455 U.S. 489, 494 (1982) (Court first
   determines if law reaches a "substantial amount of constitutionally
   protected conduct"). Equally important, Defendants have broad
   discretion to interpret the USML so as to require licensing of speech
   without fear of judicial review.
   The Government has in practice exploited the overbreadth of the ITAR,
   even applying it to computer code which contains no cryptography at
   all. Mr. Brian Behlendorf maintains and develops the Apache server, a
   computer program which does not contain any cryptographic source code.
   Despite this, the Government has stated that it considers the program
   to be in violation of the ITAR. Declaration of Brian Behlendorf
   ("Behlendorf Decl."), ¶ 5. The Government's statement, carrying the
   implicit threat of prosection to the authors and maintainers of the
   program, is apparently based upon the fact that someone else could add
   cryptographic source code to the program and then have the ability to
   encrypt. Id., ¶¶. 5-10. As a result, with no proof whatsoever that the
   Apache computer code is either a defense article or "directly related"
   to an item on the USML as required by Edler, or that Mr. Behlendorf
   and his colleagues had any connection with foreign enterprises, the
   Government intimidated them into removing portions of the computer
   code from the Apache server. Behlendorf Decl., ¶ 8. The Government's
   acts have thereby both chilled and compelled Mr. Behlendorf's speech.
   Id., ¶¶ 8-11.
   Thus, the scope of the ITAR Scheme, as applied by the Government,
   reaches
   beyond encryption source code to a much wider range of computer source
   code, including nearly all computer code which facilitates
   communication. Both on its face and as applied, it allows Defendants
   to require that authors of this computer code modify their ideas to
   fit the Government's will.
   B. Edler Failed To Confine Defendants' Discretion
   The Government has contended that the narrowing construction imposed
   by the Ninth Circuit in United States v. Edler Industries, Inc. cured
   any overbreadth problems in the ITAR Scheme. As described above, this
   contention is false, as even the OLC has acknowledged. Tien Decl.,
   Exhibit D, Edler acknowledged that on its face, the definition of
   "technical data" is so expansive that if taken literally, it would
   require a license for the design principles of the diesel engine.
   Edler, 579 F.2d at 519, Such "expansive language may be construed to
   restrict . . . the interchange of scientific and technical information
   that of itself is without any substantial military application." Id.
   at 520. Thus, the Court sought to narrow the statute by requiring a
   "significant and clear" relation between the technical data and the
   production of an article on the USML. Id., at 520-21. It also imposed
   a scienter requirement for criminal liability for exports of technical
   data with peaceful uses. Id., at 521.
   Unfortunately, Edler's narrowing construction has become meaningless.
   First, as noted above at Note 10 and Page 34, Edler construed only
   technical data and not defense articles or defense services. Here,
   Plaintiff's source code has been designated as a defense article, so
   Edler simply does not apply on its own terms. More generally, given
   the demonstrated vagueness and circularity of the current ITAR
   definitions, which were changed since Edler was decided, narrowly
   construing only one term like "technical data" does nothing to limit
   Defendants' discretion. For example, Defendants now apparently
   maintain that software generally is technical data - but that
   cryptographic software is not. They thereby escape Edler's
   restrictions by relabeling their terms. The government "cannot
   foreclose the exercise of constitutional rights by mere labels."
   Button, 371 U.S. at 429.
   Second, even as to "technical data," Edler's narrowing effect has been
   eliminated as a practical matter by the subsequent elimination of
   judicial review of the regulatory designations of items as being on
   the USML in 22 U.S.C. §2778(h), which was also added after Edler was
   decided. As a result, a person charged with violating the ITAR cannot
   challenge Defendants' determination that an item was "significantly
   related to" an item on the USML. Regulatory language that apparently
   limits Defendants' discretion, such as the ITAR provision limiting
   application to items "specifically designed ... for a military
   application" (ITAR 120.3(a)), is similarly meaningless. As a practical
   matter then, whether information is controlled is completely up to
   Defendants. Edler's narrowing construction is unenforceable.
   Defendants' application of ITAR to Plaintiff demonstrates that Edler
   did not fix the problem. In theory, Edler confined Defendants to
   "controlling the conduct of assisting foreign enterprises to obtain
   military equipment and related technical expertise." Edler, 579 F.2d
   at 521 ("So confined, the statute and regulations are not
   overbroad."). If Defendants do follow Edler, their reading of it mocks
   the Ninth Circuit. Defendants assert that Plaintiff must get a license
   to publish Snuffle, even though he wishes to publish Snuffle as part
   of the scientific exchange of ideas and information extolled in Edler.
   Nor have Defendants disputed that cryptography, like much scientific
   and technical information, has important nonmilitary applications or
   that Plaintiff's intentions are nonmilitary and unconnected with any
   foreign enterprise. If Defendants truly followed Edler, they would not
   claim licensing authority over Plaintiff or any over general
   publication of such scientific information.
   C. The ITAR Scheme Is Overbroad In That Its Broad Definition Of
   "Export" Prohibits General Publication and Public Discussion. 
   Plaintiff has shown that the ITAR Scheme is so broad that it even
   allows Defendants to regulate non-cryptographic computer programs as
   source code. Behlendorf Declaration, ¶¶ 5-10. The ITAR Scheme is also
   overbroad in that it applies to a broad range of communicative acts.
   See supra pages 9-10.
   As Professor Nimmer has reasoned:
   [w]here communicative activities occur with the intent to achieve a
   public disclosure to the American people (as distinguished from a
   private disclosure to an agent of a foreign nation), then it seems
   proper to conclude that such activities may be the subject of criminal
   punishment only if a 'serious injury' to the state can be proven both
   likely and imminent as a result of such public disclosure, as the
   Supreme Court has required in other free speech contexts.
   
   Nimmer, "National Security Secrets v. Free Speech: The Issues Left
   Undecided in the Ellsberg Case," 26 Stan. L. Rev. 311, 331-32 (1974)
   (footnotes omitted).
   The broad definition of "export" harms not only publishers but U.S.
   listeners as well. See infra at pages 14-15. Since any general
   publication of information -- whether in books, journals or electronic
   forums -- may be distributed internationally or will at least,
   possibly disclose that information to foreign persons in the US,
   general publication is an "export" for which a license is needed.
   Since the ITAR Scheme so discourages publication, U.S. listeners are
   deprived of valuable information. When PROF. BERNSTEIN may not speak,
   the rights of all who wish to hear him are also violated. This concern
   is particularly acute where, as here, the speaker is a teacher and the
   intended recipients are his colleagues and students.
   
   
   
   D. The ITAR Scheme Is Overbroad in That it Assumes That All Foreigners
   Are Terrorists. 
   
   As noted in Section I.D above, the AECA prohibits granting export
   licenses to foreign persons. 22 U.S.C. § 2778(g)(5). There are no
   exceptions. Even a Nobel Prize winner in the United States, if a
   foreign person, may not get a license and so cannot teach a class
   containing foreign students or share with foreign colleagues any
   controlled information. Equally, because disclosure of technical data
   to foreign persons constitutes export for which a license is needed,
   every foreign person's right to receive information from a willing
   speaker is affected.
   Plaintiff has already argued that the ITAR Scheme's discrimination
   against foreign persons violates the prohibition against content
   discrimination. Bellotti, 435 U.S. at 777 (1978). ("The inherent worth
   of speech ... does not depend on the identify of its source.") Even if
   it did not, however, blanket prohibitions based on foreign nationality
   are overbroad because not every foreign person poses the risk feared
   by the government. The ITAR Scheme is unconstitutionally overbroad not
   simply because it includes within its sweep some impermissible
   applications, but because in all its applications it operates on a
   fundamentally mistaken premise - that all foreign persons are a risk
   to national security. Munson, 467 U.S. at 965-66 (footnote omitted).
   In Thornhill, an anti-picketing statute was held unconstitutional
   despite Alabama's undoubted interests in protecting against violence
   and breaches of the peace because that risk is not presented by every
   picketer; the law did not target serious encroachments or evidence
   care for the interest of the community and of the individual in
   freedom of discussion. Thornhill, 310 U.S. at 105. The law made "no
   exceptions" based on any number of circumstances relevant to the
   likely risk of picketing. Id., at 99.
   Blanket assumptions about "the substantiality of the relationship
   between an individual and a group" are impermissible. Aptheker v.
   Secretary of State, 378 U.S. 500, 510 (1964) (crime for members of
   Communist groups to use or obtain U.S. passport); United States v.
   Robel, 389 U.S. 258, 265 (1967) (law imposing criminal penalties on
   Communists who got employment in a defense facility was overbroad
   because it "establishe[d] guilt . . . without any need to establish
   that an individual's association poses the threat feared by the
   Government in proscribing it.") (footnote omitted). These laws, like
   that in Thornhill, reached First Amendment activities without making
   exceptions where no genuine risk existed.
   The ITAR Scheme functions the same way. Every disclosure of technical
   data to a foreign person requires a license. No foreign person can get
   a license. "Freedom of expression, and its intersection with the
   guarantee of equal protection, would rest on a soft foundation indeed
   if government could distinguish among [speakers] on such a wholesale
   and categorical basis." Mosley, 408 U.S. at 100.
   CONCLUSION
   There is no genuine issue of material fact that the ITAR Scheme
   creates a prior restraint on speech about cryptography, in requiring a
   license prior to publication or public communication of privately
   developed, unclassified cryptographic software, technical data and
   even such fundamental mathematics tools as algorithms, designs and
   logic flows. There is no genuine issue of material fact that, even
   absent the censorship scheme, the ITAR Scheme punishes U.S. citizens
   for engaging in speech about cryptography. As a result, longstanding
   U.S. Constitutional law mandates that the ITAR Scheme pass both the
   prior restraint test of Pentagon Papers and the strict scrutiny tests.
   Plaintiff submits that these tests cannot be applied without discovery
   as to the damage caused by publication, the Government's interests,
   and the fit between those interests and the regulatory scheme and the
   existence of less restrictive means, among others.
   As to the procedural protections required by the First Amendment
   however, there is no need for discovery. There is no genuine issue of
   material fact that the ITAR Scheme, to the extent it controls speech
   about cryptography, falls far short of the minimum requirements for
   speech licensing schemes. Indeed, the undisputed facts demonstrate
   that the ITAR Scheme is a Kafkaesque procedural labyrinth pervaded by
   unfettered discretion and delay. First Amendment law requires a much
   more precise and protected scheme before scientific speech can be
   properly regulated. On this ground alone, Plaintiff urges this Court
   to invalidate the ITAR Scheme to the extent it reaches speech about
   cryptography, including source code.
   In addition, while the lack of procedural protections in the ITAR
   Scheme is a critical violation of the First Amendment , the ITAR
   Scheme is a prior restraint which cannot be saved by procedural
   safeguards alone. Given the vagueness and facial and applied
   overbreadth of the ITAR Scheme, it is guaranteed to sweep more widely
   than its stated purpose of preventing "the conduct of assisting
   foreign enterprises to obtain military equipment and related technical
   expertise." Edler, 579 F.2d at 521. As the Supreme Court long ago
   observed:
   [T]o the extent that vague standards do not sufficiently guide the
   censor, the problem is not cured merely by affording de novo judicial
   review. Vague standards, unless narrowed by interpretation, encourage
   erratic administration whether the censor be administrative or
   judicial; 'individual impressions become the yardstick of action, and
   result in regulation in accordance with the beliefs of the individual
   censor rather than regulation by law[.]'
   
   Interstate Circuit, 390 U.S. at 685 (quotations omitted)
   The preclusion of judicial review under 22 U.S.C. § 2778(h) magnifies
   every defect in the scheme. As demonstrated above, it defeats judicial
   attempts to narrow the scope or application of the ITAR Scheme, since
   the required findings and showings by the Government are fundamentally
   unenforceable and the agency retains unbounded discretion.
   Accordingly, the statute must be invalidated on its face.
   Dated:_____________________ Respectfully submitted,
   
   McGLASHAN & SARRAIL
   Professional Corporation
   
   
   By:______________________
   CINDY A. COHN