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 OPPOSITION TO v. ) DEFENDANT'S MOTION FOR ) SUMMARY JUDGMENT ) UNITED STATES DEPARTMENT OF ) Date: September 20, 1996 STATE et al., ) Time: 12:00 ) Judge: Hon. Marilyn Hall Patel Defendants. ) ) ________________________________) TABLE OF CONTENTS INTRODUCTION 1 STATEMENT OF FACTS 1 Defendants Did Extend ITAR Scheme Jurisdiction Over Plaintiff's Scientific Paper and Instructions 1 Defendants' Assertions Concerning the Public Domain Exception Are Demonstrably False 2 Defendants' Do Control Algorithms 3 I. ITAR SCHEME IS A PRIOR RESTRAINT 4 A. Defendants Applied The ITAR As A Prior Restraint on the Publication of Plaintiff's Source Code, Paper and Technical Data 5 1. Defendants Cannot Rewrite History: They Did Apply the ITAR Licensing Scheme to Plaintiff's Academic Paper and Instructions 6 B. The Technical Data Provisions of the ITAR, On Their Face, Regulate Scientific and Academic Publication and Exchange 8 C. The Technical Data Provisions of the ITAR, As Applied, Regulate Scientific and Academic Publication and Exchange 9 1. The Public Domain Exception Does Not Function To Prevent Academic and Scientific Censorship 10 2. Defendants' Assertion That The ITAR Has No Impact On Scientific and Academic Development of Cryptography Is Demonstrably False 11 D. ITAR Restrictions on Cryptographic Source Code Regulate Scientific Publication and Academic Exchanges 12 E. Neither Edler Nor Defendants' Post Edler Amendments to the ITAR Scheme Eliminates Plaintiff's First Amendments Challenge To The Technical Data Provisions of the Current ITAR 12 F. The Prior Restraint Created by the ITAR Scheme Does Not Pass Constitutional Muster and Must Be Invalidated 13 II. THE ITAR SCHEME IS PRESUMPTIVELY INVALID AS A CONTENT-BASED RESTRAINT ON PROTECTED SPEECH 14 A. The Export Controls on Cryptographic Software Are Content-Discriminatory on Their Face 15 B. Regulation of Cryptographic Software Because It Functions To Encrypt Is Content-Based Regulation 17 C. Because The ITAR's Export Controls Implicate Speech Alone, The First Amendment Demands Strict Scrutiny 19 D. The ITAR's Export Controls Are Facially Invalid Even If Viewed as Content-Neutral 20 1. The ITAR Scheme Cannot Survive The Intermediate Level Of Scrutiny Required By United States v. O'Brien 20 2. The ITAR Scheme Is An Unconstitutional Ban On A Particular Mode of Expression 21 III. THE ITAR SCHEME IS OVERBROAD ON ITS FACE AND AS APPLIED 22 A. The Court May Properly Consider and Grant Plaintiff's Facial Challenge 22 B. The Challenged Provisions Are Substantially Overbroad 25 1. The ITAR Scheme Restricts Speech That Poses No Harm To Government Interests 26 2. The Exemptions Do Not Eliminate Overbreadth 27 IV. THE ITAR SCHEME IS UNCONSTITUTIONALLY VAGUE 27 A. The ITAR's Labels Are Vague 28 B. Defendants' Focus On The Clarity of Their Exemptions Is Misplaced 29 1. The "Public Domain" Exemption Is Vague 30 2. The Academic Exemption Is Vague 31 C. The ITAR Scheme Lacks The Necessary Scienter For Civil and Criminal Liability 33 V. WITHOUT FURTHER SAVING CONSTRUCTION, THE ITAR SCHEME IS EITHER ULTRA VIRES OR AN UNCONSTITUTIONAL DELEGATION OF LEGISLATIVE POWER 35 A. The ITAR Scheme Is Ultra Vires 36 1. Commercial Transactions: The Core Meaning of the AECA 36 2. AECA Does Not Apply to Domestic Speech 38 3. AECA Does Not Apply to General Publication 39 B. AS INTERPRETED BY DEFENDANTS, THE SCHEME VIOLATES THE NON-DELEGATION DOCTRINE 40 CONCLUSION 41 TABLE OF AUTHORITIES CASES A.L.A. Schechter Poultry Corp. v. United States, 295 U.S. 495 (1935) 40 ACLU v. Reno, 1996 WL 311865 (E.D. Pa., June 11, 1996) 22 Baggett v. Bullitt, 377 U.S. 360 (1964) 30,33 Bantam Books v. Sullivan, 372 U.S. 58 (1963) 11 Bd. of Airport Comm'rs of Los Angeles v. Jews for Jesus, Inc., 482 U.S. 569 (1987) 23 Bd. of Trustees of State University of New York v. Fox, 492 U.S. 469 (1989) 24 Bernstein v. Department of State, 922 Fed.Supp. 1426 (N.D. Cal., 1996). 1,17,18 Blount v. Rizzi, 400 U.S. 410 (1971) 6 Boos v. Berry, 485 U.S. 312 (1988) 19 Bowen v. Georgetown Univ. Hosp., 488 U.S. 204 (1988) 3 Brandenberg v. Ohio, 395 U.S. 444 (1969) 14 Brockett v. Spokane Arcades, Inc., 472 U.S. 491 (1985) 23 Buckley v. Valeo, 424 U.S. 1 (1976) 27 Bullfrog Films, Inc. v. Wick, 847 F.2d 502 (9th Cir.1988) 16,27 Canton R.R. Co. v. Rogan, 340 U.S. 511(1951) 38 Cantwell v. Connecticut, 310 U.S. 296 (1940) 16 Chevron U.S.A. v. Natural Resources Defense Council, Inc., 467 U.S. 837 (1984) 3,39 City of Cincinnati v. Discovery Network, Inc., 507 U.S. 410 (1993) 15 City of Houston v. Hill, 482 U.S. 451 (1987) 23,25 City of Mesquite v. Aladdin's Castle, Inc., 455 U.S. 283 (1982) 6,7 City of Lakewood v. Plain Dealer Publishing Co., 486 U.S.750 (1988) 4,10,12,23,39 City of Ladue v. Gillieo, 512 U.S. 43 (1994) 21,22 Cohen v. California, 403 U.S. 15 (1971). 12 Community For Creative Non-Violence v. Turner, 893 F.2d 1387 (D.C. Cir. 1990) 20,21 Edward J. DeBartolo Corp. v. Florida Gulf Coast Building & Construction Trades Council, 485 U.S. 568 (1988) 39 First National Bank of Boston v. Bellotti, 425 U.S. 765 (1978) 16 Forsyth County v. Nationalist Movement, 505 U.S. 123 (1992). 23,27 Freedman v. Maryland, 301 U.S. 51 (1965) 5,6,13,42 FW/PBS, Inc. v. City of Dallas, 493 U.S. 215 (1990) 6 Gaudiya Vaishnava Society v. City of San Francisco, 952 F.2d 1059 (9th Cir. 1990) 21 Gorin v. United States, 312 U.S. 19 (1941) 34,35 Grayned v. City of Rockford, 408 U.S. 104 (1972) 28,30 Greene v. McElroy, 360 U.S. 474 (1959) 39 Grossman v. City of Portland, 33 F.2d 1200 (9th Cir. 1994) 20,26 Gustafson v. Alloyd Co., Inc., 115 S. Ct. 1061 (1995) 36 Hurley v. Irish-American Gay, Lesbian & Bisexual Group of Boston, 115 S.Ct. 2347 (1995) 12 Interstate Circuit v. City of Dallas, 390 U.S. 676 (1968) 28 Jamison v. Texas, 318 U.S. 413 (1943) 21 Joseph Burstyn, Inc. v. Wilson, 343 U.S. 495 (1952) 22 K-Mart v. Cartier, Inc., 486 U.S. 281 (1988) 36 Karn v. U.S. Department of State, 925 F.Supp. 1 (D.D.C. 1996) 29 Keyishian v. Bd. of Regents, 385 U.S. 589 (1987) 33 Kolender v. Lawson, 461 U.S. 352 (1983) 27 Lind v. Grimmer, 30 F.3d 1115 (9th Cir. 1994) 23 Local 74, United Brotherhood of Carpenters v. NLRB, 341 U.S. 707 (1951). 7 Lovell v. City of Griffin, 303 U.S. 444(1938). 21,39 Luckie v. E.P.A., 752 F.2d 454 (9th Cir. 1985). 6,7,8 Members of the City Council of Los Angeles v. Taxpayers for Vincent, 466 U.S. 789(1984). 22,25 Mishkin v. New York, 383 U.S. 502 (1966) 33 Mutual Film Corp. v. Industrial Comm'n of Ohio, 236 U.S. 230 (1915) 22 NAACP v. Button,371 U.S. 415 (1963) 19,27 Near v. Minnesota ex rel Olson, 283 U.S. 697 (1931) 13,15 New York Times Co. v. Sullivan, 376 U.S. 255 (1964) 33 New York Times Co. v. United States , 403 U.S. 713 (1971) 4,13,14 Panama Ref. Co. v. Ryan, 293 U.S. 388 (1935). 40 Police Department of Chicago v. Mosley, 408 U.S. 92(1972). 15 Ratzlaf v. United States, 510 U.S. 135 (1994) 35 Renne v. Geary, 501 U.S. 312 (1991) 23 Rosen v. Port of Portland, 641, F.2d 1243 (9th Cir. 1981) 26 Rust v. Sullivan, 500 U.S. 173 (1991) 40 Schenk v. United States, 249 U.S. 47 (1919) 13 Schneider v. Smith, 390 U.S. 17 (1968) 39 Schnieder v. State, 308 U.S. 147 (1939). 21 Sec'y of State of Maryland v. Joseph H. Munson Co., 467 U.S. 947 (1984) 22 Securities Indus. Ass'n v. Bd. of Governors of the Fed. Reserve Sys. 468 U.S. 137 (1984) . . 3 Simon & Schuster, Inc. v. N.Y. Crime Victims Bd., 502 U.S. 105 (1991). 15 Southeastern Promotions Ltd., v. Conrad, 420 U.S. 546 (1975) 5 Speiser v. Randall, 357 U.S. 513 (1958) 30 Staub v. City of Baxley, 355 U.S. 313 (1958) 4 Talley v. California , 362 U.S. 60 (1960) 26 Thornhill v. Alabama, 310 U.S. 88 (1940). 12 Thomas v. Collins, 323 U.S. 516 (1944) 26 Touby v. United States, 500 U.S. 160 (1991) 40 United States v. Adames, 878 F.2d 1374 (11th Cir. 1989) 30 United States v. Albertini, 472 U.S. 675 (1985). 20 United States v. Beck, 615 F.2d 441(7th Cir. 1980) 34,38 United States v. Durrani, 835 F.2d 410(2nd Cir. 1987) 38 United States v. Edler Industries, Inc., 579 F.2d 516 (9th Cir. 1978). 10,12,13,18,31,34,38,39 United States v. Evans, 667 F.Supp. 974 (S.D.N.Y. 1987) 38 United States v. Gurrola-Garcia, 547 F.2d 1075 (9th Cir. 1976) 38 United States v. Heine, 151 F.2d 813 (2nd Cir. 1945) 34 United States v. Mendoza-Lopez, 107 S.Ct. 2146 (1987) 41 United States v. Murphy, 852 F.2d 1 (1st Cir. 1988) 34 United States v. Nordic Village, Inc., 503 U.S. 30 (1992) . 36 United States v. National Treasury Employees Union, 115 S.Ct. 1003 26, 27 United States v. O'Brien, 391 U.S. 367 (1968). 14,20,21 United States v. Posey, 864 F.2d 1487 (9th Cir. 1989) 31,38 United States v. Progressive, Inc. 467 F.Supp. 990 (W.D. Wis., 1979) 13 United States v. Robel, 389 U.S. 258 (1967) 40 United States v. Thirty-Seven (37) Photographs, 402 U.S. 363 (1971) 6 United States v. Trans-Missouri Freight Ass'n, 166 U.S. 290 (1897) 7 United States v. Tsai, 954 F.2d 155 (3rd Cir. 1992) 34 United States v. W.T. Grant Co., 345 U.S. 629 (1953). 6 United States v. X-citement Video, 115 S.Ct. 464 (1994). 34,35 Village of Hoffman Estates v. Flipside, 455 U.S. 489(1982) 27,28 Virginia v. American Booksellers Ass'n, Inc. 484 U.S. 383 (1988) 23,28 Ward v. Rock Against Racism, 491 U.S. 781 (1989). 20 Walters v. Churchill, 114 S.Ct. 1887 (1994) 19 Yakus v. United States, 321 U.S. 414 (1944). 40 Yniguez v. Arizonans for Official English, 69 F.3d 920 (9th Cir. 1995) 19,23 STATUTES 15 C.F.R. 799.3(a)(1) 30 22 C.F.R. 120.4(a) 28 22 C.F.R. 120.4(d)(2) 29 22 C.F.R. 120.10(a)(1) 29 22 C.F.R. 120.10(a)(4) 15 22 C.F.R. 120.10(a)(5) 8 22 C.F.R. 120.11 30 22 C.F.R. 120.11(a)(1),(2),(3),(4) 8 22 C.F.R. 120.11(a)(6),(8) 8 22 C.F.R. 120.17(a)(1). 18 22 C.F.R. 120.17(a)(4) 16,18,24,39 22 C.F.R. 120.17(a)(5) 16 22 C.F.R. 121.8(f) 3,5,15,24,29 22 C.F.R. 122.1(c) 26 22 C.F.R. 122.5(a) 26 22 C.F.R. 125.1(a) 29,30,31 22 C.F.R. 126.7(a) 7 22 C.F.R. 126.9 7,28 22 U.S.C. 1934(a) 35 22 U.S.C. 2778 33,35,36 22 U.S.C. 2778(a)(1) 35 22 U.S.C. 2778(a)(1)(C)(5) 24 22 U.S.C. 2778(b)(1)(A) 36 22 U.S.C. 2778(c) 33 22 U.S.C. 2778(e) 33 22 U.S.C. 2778(g)(1)(C)(5). 16 22 U.S.C. 2778(g)(2) 36 22 U.S.C. 2778(g)(9)(D) 36 22 U.S.C. 2778(h) 3,8,13,27,28,35,40 22 U.S.C. 2794(7) 35,36,40 50 U.S.C. 2415(5)(C) 33 INTRODUCTION In seeking summary judgment, Defendants again unabashedly seek to rewrite history, ignore facts, and unilaterally eliminate issues in this case. In the end, however, they fail to meet their burden to prove that there is no genuine issue of material fact that the Arms Export Control Act and International Traffic in Arms Regulations (hereinafter "ITAR Scheme"), as applied to protected expression, is constitutional. Instead, it becomes even more obvious that the ITAR Scheme improperly grants wide, unfettered discretion to Defendants. Because they have not justified this prior restraint -- which bears a heavy presumption against validity -- Defendants' Motion must be denied and Plaintiff's Motion for Partial Summary Judgment should be granted. STATEMENT OF FACTS Defendants' arguments rest on several false or disputed factual assertions. Each of these assertions, repeated throughout their Brief and the attached Declarations, is intended to draw attention from Defendants' regulations and administrative actions and focus instead on their post-litigation, attorney-drafted actions and assertions of their policies. As each is disproved or disputed, undue discretion is demonstrated, and the foundation for Defendants' arguments crumbles. With it their entire case must fall. Defendants Did Extend ITAR Scheme Jurisdiction Over Plaintiff's Scientific Paper and Instructions. Defendants falsely assert that they never extended the ITAR Scheme to PROF. BERNSTEIN's Scientific Paper and Instructions. Memorandum of Points and Authorities In Support of Defendants' Motion for Summary Judgment at 4:6-16, 5:24-29, 6:8-9; (hereinafter "Defs' Mem"). These assertions are false and should be disregarded. This Court found: "On October 5, 1993 the OTDC notified Bernstein that all of the referenced items were defense articles under Category XIII(b)(1)." Bernstein v. Department of State, 922 F.Supp. 1426, 1430 (N.D.. Cal. 1996) (emphasis in original). The Court specifically discussed Plaintiff's Paper, noting: "[P]laintiff had every reason to believe his paper had been determined to be a defense article until defendants' clarifying letter of June 29, 1995." Bernstein, 922 F. Supp. at 1437, fn. 19. (emphasis added). That Defendants designated the Paper as a defense article for over three years and the Instructions as a defense article for nearly two years, and only withdrew these designations after they were sued, only demonstrates the breadth of Defendants' discretion. Defendants Assertions Concerning the Public Domain Exception Are Demonstrably False Defendants' also falsely assert that they do not ever "substitute their judgment" about what may be in the public domain or "seek to control the various means by which information is placed in the public domain." Defs' Mem at 7:3-9:3, 8:4-5; 11:20-21; see also Defendants' July 25, 1996 letter, attached as Exhibit 24 to Declaration of William P. Crowell ("Defendants' July 25, 1996 letter"). In his CJ Requests, Plaintiff clearly stated: "I would like to publish these items in a widely read international electronic conference known as 'sci.crypt' for discussion by the worldwide academic community." Bernstein Decl., Exhibits A and G. Defendants did presume to control Plaintiff's placement of these materials into public domain: they informed him that a license was required. More importantly, the pre-litigation statements of Mr. Charles Ray directly refute Defendants' assertions here. Mr. Ray told PROF. BERNSTEIN that Defendants did view their role as including the censorship of information and that someone who published information or placed it into a library where "anyone could get it" risked prosecution. Further, when asked, Mr. Ray could not even name a single piece of information which was in the public domain. Bernstein Decl. paras. 27-31. Plaintiff has presented Declarations from himself, Professor Junger (paras. 18-28), Dr. Blaze (paras. 12-15), Mr. Schneier (para. 27), Mr. Johnson (paras. 22-33), Mr. Behlendorf (paras. 4-11) and Mr. Miller (paras. 27-45) attesting to the actual impact of the ITAR Scheme on information which scientists and academics seek to place into the public domain. He has also presented evidence of this impact from the Justice Department's Office of Legal Counsel ("OLC"), from Congress, and from the National Research Council as recently as a few months ago. It cannot credibly be disputed that the ITAR Scheme, on its face and as applied, controls publication through its administration of these exemptions. Defendants present only conclusory assertions, not facts. When agency pre-litigation actions and statements to the affected public are directly at odds with their post-litigation, attorney-drafted, nonbinding assertions, both common sense and settled law require that the actual actions and statements of the agency, rather than their belated assertions, serve as the basis for judicial evaluation. Thus, Defendants' protests that they do not presume to determine what may be placed into the public domain or how it may be placed there must be disregarded. Defendants' Do Control Algorithms. Defendants do not deny that the ITAR Scheme plainly includes algorithms in its definition of software and so makes them defense articles requiring pre-publication licensing. 22 C.F.R. 121.8(f). Instead, Defendants assert that they never actually apply this portion of the regulation. Defs. Mem. 31:8-25. This assertion is obviously specious; it is directly refuted by Defendants' letters to PROF. BERNSTEIN. In response to Plaintiff's first CJ Request, he was told that the reason a license was required for his items was: "this commodity is a stand-alone cryptographic algorithm which is not incorporated into a finished software product." Bernstein Decl, Exhibit B. Defendants' statements that they do not apply their regulations to algorithms should be disregarded. I. ITAR SCHEME IS A PRIOR RESTRAINT. A prior restraint exists when a licensing scheme has a "close enough nexus to expression, or to conduct commonly associated with expression, to pose a real and substantial threat of the identified censorship risks," City of Lakewood v. Plain Dealer Publishing Co., 486 U.S. 750, 759 (1988) (scheme required an annual license for placement of newsracks on public property). Here, even the OLC agrees that ITAR Scheme licensing "as a prerequisite to 'exports' of cryptographic information clearly raises First Amendment questions of prior restraint." Tien Decl., Exh. A, Bates stamp #60078. The ITAR Scheme is unequivocally a prior restraint on speech about cryptography, including cryptographic source code, since it requires a license prior to publication of these forms of expression. See Staub v. City of Baxley, 355 U.S. 313, 322 (1958); see also Plaintiff's Memorandum of Points and Authorities in Support of his Motion for Partial Summary Judgment and/or Summary Adjudication of Issues at 15:8-19:3. ("Plaintiff's Motion") Defendants, therefore, must overcome the "heavy presumption against Constitutional validity" of such restraints. New York Times Co. v. United States, 403 U.S. 713, 714 (1971) (quotation and citations omitted)("Pentagon Papers"). As in Pentagon Papers, Defendants have here failed to demonstrate that publication of speech about cryptography "will surely result in direct, immediate and irreparable damage to our Nation or its people." Id. at 730 (Stewart, J. concurring). See Plaintiff's Motion at 15:8-21:20. Since Defendants have failed entirely in their burden here, the ITAR Scheme must be invalidated. A. Defendants Applied The ITAR as a Prior Restraint on the Publication of Plaintiff's Source Code, Paper and Technical Data. Defendants virtually ignore the key prior restraint in the ITAR Scheme created by requiring an arms export license for "software." These provisions restrain two types of speech: computer source code and the non-code speech items listed as part of the definition of software, including "system functional design, logic flow, algorithms, . . ." 22 C.F.R. 121.8(f). First, as to source code, Defendants admit that "[t]here is no dispute that the State Department requires a license to export Snuffle (and other cryptographic software)." Defs' Mem at 14. Likewise, they acknowledge "that source code can be speech for First Amendment purposes." Ibid. One of the most important issues in this case is the question of whether Defendants may legitimately require a license prior to the publication of protected expression in the form of source code. Instead of directly addressing the question, however, Defendants attempt to avoid it, not even mentioning it in their section devoted to prior restraints. Defendants' entire discussion of the subject is one sentence in their separate section devoted to content neutrality: "[I]f export controls on source code do not impermissibly restrict the content of protected speech, they are not a prior restraint or otherwise in violation of the First Amendment." Defs' Mem at 14:10-13. This assertion is flatly wrong. Even a speech licensing scheme which is content neutral is still a prior restraint if it "presents a danger of unduly suppressing protected expression." Freedman v. Maryland, 380 US 51, 54, 56-57 (1965) (discretion with respect to timing is excessive). Where procedural safeguards are lacking, the Court need not even consider the substantive standard used by officials. Southeastern Promotions, Ltd. v. Conrad, 420 U.S. 546, 562 (1975). The existence of such discretion and the lack of procedural safeguards, without more, violates the First Amendment. See FW/PBS, Inc. v. City of Dallas, 493 U.S. 215, 223-227 (1990) (plurality); id., at 238 (Brennan, J., concurring in the judgment); see id., at 241 ("danger posed . . . is not derived from the basis on which the license was purportedly denied . . . . [but from] the unlawful stifling of speech that results."). As to the non-code items defined as "software," as noted above at pages 3-4, Defendants assertion that they never actually apply the portion of the definition including algorithms is directly refuted by their own actions here. It also demonstrates that there is a genuine issue of material fact requiring discovery as to Defendants' practice with others. 1. Defendants Cannot Rewrite History: They Did Apply the ITAR Licensing Scheme to Plaintiff's Academic Paper and Instructions. As noted above, Defendants' assertion that "The ITAR Was Not Applied to Restrain Publication of Plaintiff's Paper or the Instructions," (Defs' Mem 4:6-14, 5:24-9) is simply wrong. The question raised by the Court was whether Defendants belated, post-litigation retraction of their decision designating both the Paper and Instructions as defense articles mooted Plaintiff's challenges to the constitutionality of that determination. The answer to that question is no. Voluntary cessation of an allegedly unlawful practice without clear evidence that the censorship cannot be reapplied, does not moot a facial challenge. United States v. W.T. Grant Co., 345 U.S. 629, 632 (1953); See e.g. City of Mesquite v. Aladdin's Castle, Inc., 455 U.S. 283 (1982); Luckie v. E.P.A. 752 F.2d 454, 458-59 (9th Cir. 1985). ("If Plaintiff shows a mere probability of recurrence, coupled with a certainty that the impact will fall on the same objecting litigants, the government has a heavy burden to show there is no reasonable expectation that the alleged violation will recur.") At issue in City of Mesquite, was the vagueness of a local ordinance which allowed the City Manager to refuse an arcade license to those with "connections with criminal elements." After an injunction against the provision was entered, the city amended the licensing scheme to remove the allegedly vague language. This unilateral act by the government did not moot the case, however: It is well settled that a defendant's voluntary cessation of a challenged practice does not deprive a federal court of its power to determine the legality of the practice. . . In this case the city's repeal of the objectionable language would not preclude it from reenacting precisely the same provision if the District Court's judgment [granting the injunction] were vacated. . . . There is no certainty that a similar course would not be pursued if its most recent amendment were effective to defeat federal jurisdiction. City of Mesquite, 455 U.S. at 289. Plaintiff has even less protection than the City of Mesquite Plaintiffs. Neither Defendants' "clarification" letters nor their C.J. determinations are binding. 22 C.F.R. 126.9. Even if a license is granted it can be revoked if Defendants deem it "advisable." 22 C.F.R. 126.7(a). Moreover, PROF. BERNSTEIN plans to continue to work and publish in this field. Bernstein Decl. paras. 43-46. His future work will certainly run afoul of the ITAR Scheme again. Unless the ITAR Scheme is changed to limit Defendants' discretion, the illegal censorship has much more than a mere probability of recurrence. Moreover, when it does recur, as is likely, there is an absolute certainty that the impact such censorship will chill Plaintiff's activities -- either directly or as he sees that others are restrained and/or prosecuted. Defendants have therefore failed to meet their burden under Luckie. B. The Technical Data Provisions of the ITAR, On Their Face, Regulate Scientific and Academic Publication and Exchange. Defendants then address the facial challenge which Plaintiff makes to the technical data provisions. Defendants begin by pointing out the various exceptions which they have made to the ITAR for certain specific speech activities. There are three main problems with this line of argument, the first two of which are addressed more fully elsewhere. First, the exceptions are not practically enforceable since after the Edler decision the AECA was amended to add Sec. 2778(h) which precludes judicial review. Plaintiff's Motion at 40:11-19. Second, as explained in further detail in Section IV, below, the exemptions themselves are impermissibly vague. Third, Defendants' argument, which they repeat several time throughout the brief, (Defs' Mem 7:3 - 9:9; 29:20-30:13; 33:3-34:2) turns First Amendment analysis on its head. In evaluating whether a regulatory scheme meets First Amendment scrutiny, the question to be addressed by a Court is not whether the scheme exempts some speech from regulation; the question is whether, even with the exceptions, the scheme still regulates speech. Defendants' argument is based on the following provisions: information concerning general scientific, mathematical or engineering principles commonly taught in schools, colleges, and universities. 22 C.F.R. 120.10(a)(5). and information that is in the "public domain", which includes: information which is published and generally available and accessible through, inter alia, sales at newsstands and bookstores, subscriptions, second class mail, and libraries open to the public. 22 C.F.R. 120.11(a)(1)(2)(3)(4) and information which is made generally available to the public through unlimited distribution at a conference, meeting, seminar, trade show or exhibition, generally accessible to the public, in the United States or through fundamental research in science and engineering at accredited institution of higher learning in the U.S. where the resulting information is ordinarily published and shared broadly in the scientific community. 22 C.F.R. 120.11(a)(6), (8). Even if the definitions were construed unambiguously and even if these clauses were enforceable, the plain meaning of these terms still leaves significant First Amendment protected activity subject to regulation. First, they do not protect a U.S. engineer who wishes to discuss technical matters with foreign colleagues. This situation is presented by Dr. Matthew Blaze, who develops ideas about cryptography for AT&T Laboratories, as a professor at Columbia University and on his own. Dr. Blaze's discussion of his cryptographic ideas with a foreign colleague still requires a license. Blaze Decl. paras. 6-9, 12-15. Second, they do not protect the software published by Mr. Demberger. Demberger Decl. paras. 2-13. Mr. Demberger developed his cryptographic ideas privately, not at an institution of higher learning, and did not seek to present them at a conference or seminar, but instead to post them to sci.crypt, the same public Internet discussion group where PROF. BERNSTEIN intends to publish Snuffle. Dr. Blaze and Mr. Demberger exemplify the ITAR's effects on scientists and engineers who are not in academia. Mr. Demberger also illustrates how the public domain exemption has not caught up with the Internet, since it does not specifically mention electronic publication. Even academics are unprotected in important circumstances. The exemptions do not protect small-group exchanges or personal communications of new ideas. For example, since his Instructions are about a specific new idea--Snuffle--and so are neither a general principle nor fundamental research, PROF. BERNSTEIN's disclosure of them to a foreign academic or scientist would not be exempt. Much innovation and development takes place before formal publication occurs. In sum, these exemptions simply do not protect the large numbers of scientists and hobbyists who are not in academia, those in academia whose work is not "fundamental research" or about "general principles," or those even in academia who do not wish to publish their technical data but merely to communicate it to colleagues or peers abroad. Yet all of these activities are protected scientific expression. C. The Technical Data Provisions of the ITAR, As Applied, Regulate Scientific and Academic Publication and Exchange. 1. The Public Domain Exception Does Not Function To Prevent Academic and Scientific Censorship. Moreover, the State Department's assertions here also stand in direct contrast to their specific pronouncements of their intent to suppress cryptographic technical data. Indeed, these pronouncements indicate that they still intend to control technical data related to cryptography. In 1980, eleven years after United States v. Edler Industries, Inc., (579 F.2d 516 (9th Cir. 1978)), was decided, the State Department stated: "[T]he public is reminded that professional and academic presentation and informal discussions, as well as demonstrations of equipment, constituting disclosure of cryptologic technical data to foreign nationals, are prohibited without the prior approval of this office" subject to certain exemptions. It continued: The interpretation set forth in this newsletter should exclude ... most basic scientific data and other theoretical research information, except for information intended or reasonably expected to have direct cryptologic application. Because of concerns expressed to this office that licensing procedures for proposed disclosures of cryptologic technical data contained in professional and academic papers and oral representations could cause burdensome delays in exchanges with foreign scientists, this office will expedite consideration as to the application of ITAR to such disclosure. Dept. of State, 80 Munitions Control Newsletter (February 1980), attached to Tien Decl., Exhibit E, NRC Report, page 4-47 (emphasis added). Defendants' now state that they no longer consider such scientific and academic activities subject to their export licensing jurisdiction. But no authoritative administrative or judicial construction of the AECA or ITAR holds them to this position, and it is inconsistent with their established record. Even the National Research Council, which was given full access to Defendants' personnel and records could find no better statement about Defendants' current interpretation of the technical data provisions than the statement above. NRC Report 4-30. The NRC also noted that academics are often unclear about the scope of the technical data provisions and that this confusion extends to "teaching a course on cryptography." Id. If Defendants have changed their policy, they certainly have kept this change well-hidden. In any case, because the regulations on their face permit Defendants to regulate academic and scientific activities, Defendants' good intentions are irrelevant. That Defendants will adhere to standards are absent from a regulation's face is "the very presumption that the doctrine forbidding unbridled discretion disallows." Lakewood, 486 U.S. at 770. If Defendants are committed to exempting academic publication, they must not only concede that Plaintiff, a mathematics professor, may freely publish Snuffle in all its forms as part of the international scientific exchange of ideas about cryptography, but also disclaim any censorship authority by regulation. 2. Defendants' Assertion That The ITAR Has No Impact on Scientific and Academic Development of Cryptography Is Demonstrably False. Defendants next claim that the ITAR has no impact on academic discussion of cryptography. Defs' Mem at 8:24-9:9. This factual assertion is not only in dispute, it is demonstrably false. That Plaintiff has been unable to contribute Snuffle to the scientific debate disproves Defendants' assertions. In addition, the Declarations served with Plaintiff's Motion further attest to the presence of a chilling effect and the negative effect the ITAR Scheme has on the development of the science of cryptography. Further, Mr. Demberger and Mr. Behlendorf both report that the Government has warned them that their publications may violate export controls. These warnings were "obviously intended" to "curtail the circulation" of information. Cf. Bantam Books Inc. v. Sullivan, 372 U.S. 58, 64-65 n.6 (1963) (warnings from quasi-official commission to local booksellers that certain materials were obscene are prior restraint). As the Bantam Books court acknowledged, courts have long looked through form to substance and properly condemned informal censorship as well as formal censorship. Id., at 67. Here, the warnings came not from private citizens, but from those with the power to prosecute. In addition, the mere existence of some debate about cryptography does not establish that self-censorship is not occurring. The great vice of a censorship scheme is the chilling effects, which are difficult, if not impossible, to measure. An administrative licensing scheme "intimidates parties into censoring their own speech, even if the discretion and power are never actually abused." Lakewood, 486 U.S. at 757. "It is not merely the sporadic abuse of power by the censor but the pervasive threat inherent in its very existence that constitutes the danger to freedom of discussion." Thornhill v. Alabama, 310 U.S. 88, 97 (1940). Thus, as explained more below at pages 25-26, the government's attempt to portray the existing cryptographic literature as robust and flourishing is also entirely beside the point, because much self-censorship is invisible. The only appropriate way to prevent such chilling effects is to (1) set clear, narrow standards for the licensor to follow; and (2) provide for proper procedural protections in the event the licensor misapplies those standards. Lakewood, 486 U.S. at 758. ( "Only standards limiting the licensor's discretion will eliminate this danger by adding an element of certainty fatal to self-censorship"). D. ITAR Restrictions on Cryptographic Source Code Regulate Scientific Publication and Academic Exchanges. Defendants conveniently ignore another great impact on scientific and academic speech created by the ITAR Scheme -- that caused by the restrictions on cryptographic source code. Defendants admitted regulation has obvious impact on scientific marketplace of ideas. Source code is fundamental to the development of the science of cryptography, as it is for many other areas of science. See Motion at 3:4-5:2; Schneier Decl. paras. 15-26, 32-34; Abelson Decl. paras. 3-17; Appel Decl. paras. 6-17; Blaze Decl. paras. 10-11; Ginsparg Decl. paras. 8-17; Ellison Decl. paras. 8-30; Stallman Decl. paras. 6-9; Johnson Decl. paras. 13-21. In restricting the language by which Plaintiff and others may communicate, Defendants unconstitutionally prescribe the "form and content" of their expression. Cohen v. California, 403 U.S. 15, 24 (1971); Hurley v. Irish-American Gay, Lesbian and Bisexual Group of Boston, 115 S.Ct. 2347, 2358 (1995). E. Neither Edler Nor Defendants' Post-Edler Amendments to the ITAR Scheme Eliminates Plaintiff's First Amendment Challenge to the Technical Data Provisions of the Current ITAR. In his Motion, Plaintiff has already provided an extended explanation of why the decision in Edler does not eliminate his challenge to the technical data provision of the ITAR. See Plaintiff's Motion at pages 39-41 and notes 10 and 39. Defendants also assert that their post-Edler amendments to the ITAR have eliminated any constitutional infirmities. Plaintiff has explained in numerous places throughout both this Opposition and his own Motion that these changes, most of which are the exemptions, simply fail to bring the ITAR Scheme into constitutional conformity. See pages 27, 29-33 Plaintiff's Motion at 40:11-19. F. The Prior Restraint Created by the ITAR Scheme Does Not Pass Constitutional Muster and Must Be Invalidated. Defendants here have failed entirely in their burden to show that publication of speech about cryptography, including source code, meets the Pentagon Papers test. In that case, Justice Brennan summarized the types of evidence which the Government must submit to meet this standard: there is a single, extremely narrow class of cases in which the First Amendment's ban on prior judicial restraint may be overridden. Our cases have thus far indicated that such cases may arise only when the Nation 'is at war' Schenk v. United States, 249 U.S. 47, 52 (1919), during which times '(n)o one would question but that a government might prevent actual obstruction to its recruiting service or the publication of the sailing dates of transports or the number and location of troops' Near v. Minnesota ex rel Olson, 283 U.S. 697, 716 (1931). Pentagon Papers, 403 U.S. at 726 (Brennan, concurring). See also United States v. Progressive, Inc, 467 F.Supp. 990 W.D. Wis. 1979), rehear den. at 486 F.Supp.5 (W.D. Wis. 1979)(publication of technical information on the making of a hydrogen bomb). Here, our nation is not at war. PROF. BERNSTEIN does not seek to publish any information about the Government's placement of troops or ships or to interfere with its ability to recruit soldiers. Defendants here do not, and cannot, claim that the publication of cryptographic source code and related information itself can cause any damage, much less create the imminent risk of thermonuclear destruction. The information he seeks to publish is not even classified. Rather, piling speculation on speculation, Defendants allege that this information, if misused by others, could limit Defendants' ability to conduct foreign surveillance. Defs' Mem at 19:22-24. The interest in foreign intelligence gathering is only an intermediate interest. Others must act before harm will occur. There is an huge gap between the occurrence of harm -- some foreign entity using Plaintiff's source code to encrypt communications planning a bad act which communications our government wishes to intercept -- and Plaintiff's publication. As the OLC noted, "[S]peech does not lose its protected character solely because the circumstances of the case give rise to a reasonable fear that persons other than the speaker may be moved or enabled by the speech to do dangerous things at remote times and places." OLC Memo, Exh. B to Tien Decl. at Bates # 60057, citing Brandenberg v. Ohio, 395 U.S. 444 (1969). Here Defendants rely on just such a remote harm and so the ITAR Scheme fails to meet the strict standard for prior restraints. II. THE ITAR SCHEME IS PRESUMPTIVELY INVALID AS A CONTENT-BASED RESTRAINT ON PROTECTED SPEECH. Defendants next attempt to justify their regulations as aimed at the functionality, not the content, of cryptographic software, hoping to invoke the weak scrutiny of United States v. O'Brien, 391 U.S. 367 (1968). That effort is unavailing. A. The Export Controls On Crytographic Software Are Content- Discriminatory On Their Face. At the most basic level, the ITAR Scheme burdens public, academic, and scientific discussion of the entire subject of cryptography. Defendants' primary argument is that the ITAR Scheme's licensing of cryptographic software and its underlying academic and scientific foundations is a content-neutral regulation of something other than speech. This point is belied by Defendants' own words. On its face, the ITAR treats cryptographic expression differently . Much of Defendants' argument emphasizes that cryptographic software is a defense article and not technical data based on 22 C.F.R. 121.8(f)(specifically mentioning Category XIII(b)) -- but software related to other USML categories is technical data. 22 C.F.R. 120.10(a)(4). It is common sense that a regulation which affects speech differently depending on what it is about is "content based." See City of Cincinnati v. Discovery Network, Inc., 507 U.S. 410, 429 (1993); Police Dep't of Chicago v. Mosley, 408 U.S. 92, 95 (1972). In addition, as Defendants again concede, the ITAR on its face presupposes that Defendants actually review software and academic and scientific information and appraise its possible uses, strengths, and effectiveness, to determine whether or not it is "capable of maintaining secrecy." A scheme which involves a licensor in the appraisal of facts, the exercise of judgment and the formation of an opinion is a content-driven scheme. See e.g. Cantwell v. Connecticut, 310 U.S. 296, 305 (1940). Defendants permit relatively "weak" cryptographic speech while prohibiting similar speech if expressed in stronger, more effective key lengths. See Defs' Mem at p.25, n.26; e.g., Johnson Decl., at para. 30. Such inquiries are inherently content-driven since they are meant to prohibit only speech about powerful cryptographic ideas. Third, the ITAR is content-based because it discriminates against foreign persons. 22 U.S.C. (g)(1)(C)(5); 22 C.F.R. 120.17(a)(4),(5). The legislature is constitutionally disqualified from dictating the subjects about which persons may speak and the speakers who may address a public issue. First National Bank of Boston v. Bellotti, 435 U.S. 765, 784-85 (1978). The ITAR Scheme does both. Finally, Defendants' public pronouncements about their interest in maintaining the ITAR Scheme restrictions on cryptography reveal a content-based interest: to influence cryptographic methods and algorithms used domestically. The Government has clearly stated that its current application of the ITAR Scheme ties export control decisions to whether a product supports its preferred encryption method called key-escrow or Government Access to Keys ("GAK"). The Government has also clearly stated that it favors the GAK method in large part because the method allows the Government to have access to domestic private messages. In implementing this interest, Defendants have offered exemptions from ITAR licensing for GAK methods, making it easier for them to be distributed in the U.S. as well as abroad. Id. The misuse of an arms export restriction to attempt to control the availability of domestic, nonmilitary cryptographic software demonstrates not only that the scheme is content based, but also the unfettered breadth of discretion, the overbreadth and the ultra vires nature of the ITAR Scheme. B. Regulation Of Cryptographic Software Because It Functions To Encrypt Is Content-Based Regulation. Defendants argue that the ITAR Scheme restricts cryptographic software not because of its content, but because of its "functionality." See Def's Mem at 21-23. This time, Defendants appear to argue that because software can be used to do something, regulation of software because of what it can be used to do is not content based. This argument, which has already been rejected in a slightly different form by this Court once (Bernstein, 922 F.Supp. at 1434-6), is once again unsupported by any legal authority. This argument is far to sweeping. Software, by itself, does nothing. Defendants argue that a recipe for a cake is not a cake, but that software is different. But just as a recipe for a cake is not a cake without ingredients, an oven and a baker, cryptographic software cannot function without a message, a computer and a person to use it In saying that software "can be made to function," Defendants implicitly concede that cryptographic software needs a person (and a computer) to do anything. Defs' Mem at 22. What a listener does with information can be conduct outside of the First Amendment, but the listeners conduct is not the speaker's speech. More generally, "[t]he music inscribed in code on the roll of a player piano is no less protected for being wholly functional." Bernstein, 922 F.Supp. at 1435. "[It] 'communicates' to and directs the instrument itself, rather than the musician, to produce the music. That does not mean it is not speech." Ibid. The same is true if the music is digitally inscribed on a compact disk (or "CD"). A CD directs the computer in the CD player to turn the digital computer code into a sound that may be heard and understood by the listener. Like cryptographic software, the CD is necessarily imbued with "functionality." That does not mean that Defendants can regulate the content of CDs, even if they may be played too loudly. To the extent that software "functions," that function is inseparable from content. Computer Science and all of its related fields, like cryptography, are devoted to the designing and testing of software. Here, Defendants target cryptographic software; but the functionality rationale extends to all software because all software can be used to do something. Regulation of software necessarily is regulation of computer science. Moreover, Defendants do not regulate functionality per se, but rather the specific functionality of cryptographic software--its enablement of private speech. Viewed in this way, it becomes clear that the real target of the regulation is private speech. Thus the Government is regulating cryptographic "functionality" on the grounds that it might facilitate harmful speech. The Ninth Circuit in Edler rejected a similar proposition in the context of technical data, that "[a]ny information that would in any way help in a process that led to the final product [comes] under the definition of technical data." Edler, 579 F.2d at 522 (permitting Edler on remand to show that its assistance "would not, of itself, suffice for the manufacture of rocket nozzle components"). Similarly, this argument should be rejected here. Indeed, this "functionality" target is doubly problematic because the First Amendment protects the freedom to communicate only with one's chosen audience. See Plaintiff's Motion at 14:16-15:7. Nor can Defendants save the ITAR as a content-neutral regulation, as they attempt to do, by redefining what is at issue as the "conduct" of "exporting" of software, source code and related cryptographic speech. See Defs' Mem., at p.23. A major constitutional problem with the ITAR Scheme is its definitions of "export" for both technical data and defense articles prohibit general publication, public discussion and scientific and academic exchange. See 22 CFR 120.17(a)(1), (a)(4) and Plaintiff's Motion at 41:10-42:3. Obviously, Defendants cannot censor these protected activities merely by relabelling them as the "conduct of export." The "export" controls are thus manifestly a direct regulation of expression, not conduct. C. Because The ITAR'S Export Controls Implicate Speech Alone, The First Amendment Demands Strict Scrutiny. Having attempted to justify the ITAR as a content-neutral regulation aimed at functional "conduct" rather than pure speech, Defendants gloss over the constitutional standard required of content-driven restrictions: the government must affirmatively demonstrates that the "regulation is necessary to serve a compelling state interest and that it is narrowly drawn to achieve that end." Boos v. Berry, 485 U.S. 312 (1988). Defendants' asserted interest in preventing harm to foreign intelligence gathering is not compelling, because it is only an interest in efficiency. The real harm flows from whatever planned harm might not be prevented due to the use of encryption, since U.S. decisionmakers did not receive the planning messages. In context, the asserted harm is that the government will be less efficient in gathering foreign intelligence in general if cryptographic software is freely publishable. "The governments' interest in achieving its goals as effectively and efficiently as possible" is "a relatively subordinate interest . . . The government cannot restrict the speech of the public at large just in the name of efficiency." Yniguez v. Arizona for Official English, 69 F.3d 920, 938 (9th Cir. 1995)(en banc), cert granted 64 U.S.L.W. 3639 (Mar. 25, 1996)(No. 95-974), quoting Walters v. Churchill, 114 S.Ct. 1887, 1888 (1994)(plurality opinion). Accordingly, the ITAR Scheme fails the first prong of the strict scrutiny analysis. Moreover, even if after discovery Defendants were able to satisfy the compelling interest test, they must also show the ITAR is "narrowly drawn," i.e., that it is the least restrictive means of achieving its stated purpose. Boos 485 U.S. at 312 (1988). On its face, the ITAR adopts the most restrictive method of prohibiting publication, public communication, and cryptographic expression -- not by restricting military and national security applications -- but by banning discussion 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, 428 (1963). As amply demonstrated on pages 25-27 below and on pages 38:1-43:9 of Plaintiff's Motion, the ITAR Scheme lacks any such precision. It is not narrowly drawn and, thus, fails the second prong of the strict scrutiny analysis. D. The ITAR's Export Controls Are Facially Invalid Even If Viewed As Content-Neutral. 1. The ITAR Scheme Cannot Survive The Intermediate Level Of Scrutiny Required By United States v. O'Brien. Defendants argue that the less exacting standard specified in O'Brien should be applied here. For at least three reasons, however, Defendants' reliance on O'Brien is unavailing here. First, 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) (emphasis added). As shown above, the ITAR is not facially content neutral. Second, a condition of applying O'Brien is that the regulation "incidentally," not directly, burdens speech. The ITAR controls are plainly not an "incidental burden" on protected expression. They are instead concerned with the direct, primary aspects of the expressive activities sought to be controlled, the "disclosure" of technical data and the general publication and public discussion of source code. Third, as a related matter, the ITAR Scheme cannot survive O'Brien's final prong, i.e., that the "incidental" restriction on expression be narrowly tailored. As explained above, the ITAR is not narrowly drawn. Even under O'Brien's more relaxed narrow tailoring requirement, a restriction on speech may not "burden substantially more speech than is necessary to further the government's legitimate interest." Ward v. Rock Against Racism, 491 U.S. 781 (1989); Grossman v. City of Portland, 33 F.3d 1200, 1205-08 (9th Cir. 1994). As explained in further detail in Plaintiff's Motion at 41:10-43:9 the ITAR Scheme burdens U.S. citizens who would receive public communications, and place burdens on their ability to receive software containing strong cryptography. The Scheme further prohibits all foreign persons from either giving or receiving any information about cryptography. Id. The ITAR Scheme casts this wide, vague net despite the fact that its stated goal is only to prevent foreign terrorists and enemy governments from receiving these ideas. That is the most draconian way of addressing Defendants' stated concerns. In short, the ITAR Scheme fails even the intermediate level of scrutiny required by United States v. O'Brien. 2. The ITAR Scheme Is An Unconstitutional Ban On A Particular Mode Of Expression. Moreover, even assuming arguendo that the ITAR Scheme may be viewed as content- neutral, it is nonetheless unconstitutional because it imposes a complete prohibition against Internet communication as a means of expressing ideas in the field of cryptographic science. American jurisprudence has long been concerned with laws that foreclose an entire medium of expression. For example, the Supreme Court has held invalid laws that banned the distribution of pamphlets within a municipality (Lovell v. Griffin, 303 U.S. 444, 451-452, (1938)), handbills on public streets (Jamison v. Texas, 318 U.S. 413, 416 (1943)), and door-to-door circulation of literature. Schnieder v. State, supra, 308 U.S. at 164-165. Similarly, in City of Ladue v. Gillieo, 512 U.S. 43 (1994) the Court struck down an ordinance that foreclosed a common, inexpensive, and convenient medium of political, religious, or personal expression -- signs placed on one's property -- concluding: [a]lthough prohibitions foreclosing entire media may be completely free of content or viewpoint discrimination, the danger they pose to the freedom of speech is readily apparent -- by eliminating a common means of speaking, such measures can suppress too much speech. The same is true here. As a three judge panel in the Eastern District of Philadelphia recently observed, the Internet, through its on-line communication and information services -- like the front-door signage of old -- offers citizens around the world a convenient, inexpensive, and common means of political, religious, personal, professional, scientific, and academic expression. See ACLU v. Reno, 1996 WL 311865 (E.D. Pa.,) June 11, 1996) Findings of Fact para. 4, para. 7, para. 81. Under the ITAR, however, this important mode of communication is entirely cut off to those who wish to use it to express their scientific and academic ideas regarding cryptography. See, e.g., Bernstein Decl., at para. 9; Demberger Decl., para. 2; Appel Decl., para. 11; Ellison Decl. paras. 28-30; Ginsparg Decl., paras. 3-8; and Stallman Decl., paras. 6-9. As such, the Scheme fails strict scrutiny under City of Ladue. III. THE ITAR SCHEME IS OVERBROAD ON ITS FACE AND AS APPLIED. In First Amendment cases, the rules of standing are relaxed. Sec'y of State of Maryland v. Joseph H. Munson Co., 467 U.S. 947, 957 n. 7 (1984); id. at 958 (finding that Plaintiff could bring facial challenge to restrictions on charitable solicitations "whether or not its own First Amendment rights are at stake"). Where a statute unquestionably attaches sanctions to protected conduct, the likelihood that the statute will deter that conduct is ordinarily sufficiently great to justify an overbreadth attack. Members of the City Council of Los Angeles v. Taxpayers for Vincent, 466 U.S. 789, 800, n. 19 (1984). A. The Court May Properly Consider and Grant Plaintiff's Facial Challenge. Defendants argue that under Brockett v. Spokane Arcades, Inc., 472 U.S. 491 (1985), a party who claims his or her own conduct is protected speech may not invoke facial overbreadth. This authority is dubious at best. Even after Brockett, the Supreme Court has heard facial overbreadth claims without ruling on the as-applied claim. See City of Houston v. Hill, 482 U.S. 451, 458 n. 6 (1987); Bd. of Airport Com'rs of Los Angeles v. Jews for Jesus, Inc., 482 U.S. 569, 573-74 (1987). The Ninth Circuit has rejected the argument that facial overbreadth is improper whenever the challenger's conduct is protected. Lind v. Grimmer, 30 F.3d 1115, 1122 (9th Cir., 1994). cert den sub nom. Want v. Lind, 115 S.Ct. 902 (1995) (distinguishing Brockett and Renne v. Geary, 501 U.S. 312, 324 (1991)); accord, Yniguez, 69 F.3d at 931. Lind identifies the proper judicial concern: "[W]here a statute's overbreadth is substantial, its chilling effect is likely to be significant, and consequently the entire statute may be invalidated to protect First Amendment interests." Lind 30 F.3d at 1122 (citation omitted). Here, as in Lind, even if the Court strikes those portions of the ITAR Scheme which are unconstitutional as applied to Plaintiff, the remainder of the ITAR Scheme will have numerous potential unconstitutional applications. More importantly, facial challenges to highly discretionary licensing schemes like the ITAR Scheme are favored. "[W]hen a licensing statute allegedly vests unbridled discretion in a government official over whether to permit or deny expressive activity, one who is subject to the law may challenge it facially." Lakewood, 486 U.S. at 755; accord, Forsyth County v. Nationalist Movement, 505 U.S. 123, 129-130 (1992). This is because an "as applied" challenge cannot remedy the harms of self-censorship inherent in a broad licensing scheme. See Virginia v. American Booksellers Ass'n, Inc., 484 U.S. 383, 393 (1988) (Plaintiff booksellers alleged that bookbuyers' First Amendment rights were infringed). Plaintiff's facial challenge is therefore independently justified because the ITAR Scheme is a standardless licensing scheme. Even if Plaintiff can only facially challenge as overbroad those portions of the ITAR Scheme which do not affect him directly, Plaintiff's facial challenge stands. First, Plaintiff challenges the ITAR Scheme's discrimination against foreign persons whose rights to speak are infringed because AECA renders foreign persons ineligible for a license. 22 U.S.C. 2778(a)(1)(C)(5). Their rights to receive information are infringed because the ITAR defines an export, inter alia, as disclosure of technical data to a foreign person. 22 C.F.R. 120.17(a)(4); see Plaintiff's Motion at 14:1-15. Clearly, there is a "significant difference" between the claims of Plaintiff and those of foreign persons not before the Court. Second, Plaintiff challenges the ITAR Scheme's restrictions on all technical data and all software which may be treated as a defense article by Defendants' application of 22 C.F.R. 121.8(f). Plaintiff's as-applied attack addresses restrictions on cryptographic speech only. But the ITAR Scheme has affected many other areas of science as well. Thus Plaintiff's overbreadth challenge seeks to protect the rights of many kinds of scientists, not just cryptographers. In short, where, as here, Plaintiff challenges a regulation as applied to his interests, and other aspects of the scheme solely pertain to third party interests that the Plaintiff does not personally share, it is the settled law that both claims may proceed. See Board of Trustees of State University of New York v. Fox, 492 U.S. 469 (1989) (students may seek to vindicate their own rights, as well as third party commercial interests, in challenging school regulation barring use of dormitories to demonstrate commercial products). B. The Challenged Provisions Are Substantially Overbroad. Overbreadth requires a realistic danger that the statute itself will significantly compromise recognized First Amendment protections of parties not before the Court. Vincent 466 U.S. 789 (1984) at 801. Here, the burden on speech is quantitatively substantial and the chilling effects are plainly seen. The ITAR Scheme on its face requires a license for any disclosure of technical data to a foreign person, any disclosure of cryptographic software, including algorithms and source code, which leaves the United States, and any service which aids a foreign person in using cryptographic software. As a practical matter, the licensing requirements reach general publication because the U.S. public includes foreign persons and because, as a rule, such publication leaves the United States. Since every "disclosure" of technical data to a foreign person is burdened, OLC identified two common situations where technical data might be innocently disclosed: the teacher in a university whose class includes foreign students; the U.S. engineer who wishes to discuss subjects with foreign colleagues. Tien Decl., Exh. A, Bates #60007, 60018, 60045. The OLC did not mention the publication of books or journal articles, but these raise the same issues. In the modern scientific world, electronic communications increase the number of instances where scientists and engineers can engage in discussion with foreign persons, either by speaking to a worldwide audience on the Internet, or to colleagues outside the United States via e-mail. PROF. BERNSTEIN might, for instance, seek to transmit a copy of Snuffle to a U.S. colleague in another country. The ITAR Scheme requires a license for these types of protected electronic communications as well. A substantial chilling effect is also seen in requirement of advance notice inherent in the ITAR Scheme's licensing requirements. This requirement "drastically burden[s] free speech" because the procedural and temporal hurdles of submitting an application and waiting for a license may discourage potential speakers. Grossman, 33 F.2d at 1206 (quotation and citation omitted). Moreover, by discouraging speakers, the Scheme significantly burdens the public's right to read and hear what would otherwise have been written or said. United States v. National Treasury Employees Union, 115 S.Ct. 1003, 1015 (1995) ("NTEU") ("no way to measure the true cost" of burden imposed by honoraria ban, "but we cannot ignore the risk that it might deprive us of the work of a future Melville or Hawthorne") (footnote omitted). 1. The ITAR Scheme Restricts Speech That Poses No Harm To Government Interests. While some communications of technical data or software might well be prohibited by a properly drawn statute, such as where the speaker intends to assist a hostile foreign government, this ITAR Scheme does not distinguish such situations. This results in further overbreadth. According to Defendants, the ITAR Scheme's restrictions on cryptographic information target harm to "critical foreign intelligence-gathering functions." But even if we look only at cryptography, the ITAR Scheme clearly reaches protected expression that is in no way linked to that harm. The Government cannot require that all such disclosures and exchanges be licensed just because some of them may be harmful. "Deferring to the Government's speculation about the pernicious effects of thousands of articles and speeches yet to be written or delivered would encroach unacceptably on the First Amendment's protections." NTEU at 1017-18 n. 21. At a minimum, such a claim requires discovery. 2. The Exemptions Do Not Eliminate Overbreadth. Defendants then claim that the various exemptions in the ITAR limit its facial overbreadth. But these exemptions are legally meaningless because, as noted above, a court cannot review Defendants' decision that something is controlled by them. 22 U.S.C. 2778(h). Even if we ignore this problem, the exemptions fail because they are ambiguous, as explained below. Such vagueness amplifies the Scheme's overbreadth; "a court should evaluate the ambiguous as well as the unambiguous scope of the enactment. To this extent, the vagueness of a law affects overbreadth analysis." Village of Hoffman Estates v. Flipside, 455 U.S. 489, 494 n. 6 (1982). IV. THE ITAR SCHEME IS UNCONSTITUTIONALLY VAGUE. Defendants' response to Plaintiff's vagueness challenge is seriously misguided. Where speech is at issue, laws must be narrowly specific. Buckley v. Valeo, 424 U.S. 1, 77 (1976); Button, 371 U.S. at 432; Bullfrog Films, 847 F.2d at 513. Defendants, for the most part, defend the ITAR Scheme as "reasonably clear," making only passing reference in a footnote to the stricter First Amendment standard. Laws, however, are not only rules of public conduct which must give fair notice of what is prohibited; laws are also rules of official conduct which must provide legislative standards for enforcement officials and for the courts. Kolender v. Lawson, 461 U.S. 352, 358 (1983). This aspect of vagueness - on which Defendants are, essentially, silent - is especially important where speech is concerned. The absence of standards is a First Amendment vice in three main ways. It permits arbitrary enforcement. Grayned v. City of Rockford, 408 U.S. 104, 108-09 (1972) (abdication on policy matters creates risk of "arbitrary and discriminatory application"). It renders "inoperative" judicial review of enforcement. Interstate Circuit v. City of Dallas, 390 U.S. 676, 684-85 (1968). Most important, vagueness engenders self-censorship "even without an actual prosecution." American Booksellers, 484 U.S. at 393; see Hoffman Estates, 455 U.S. at 499 ("Perhaps the most important factor . . . is whether [a law] threatens to inhibit the exercise of constitutionally protected rights."). Judged by these well-settled tests, the ITAR Scheme fails entirely. A. The ITAR's Labels Are Vague. Defendants argue that, taken individually, the technical data and cryptographic software provisions are reasonably clear. But a root of the ITAR's vagueness is the ambiguity of the labels "technical data," "defense article," and "defense service." By taking these provisions one at a time, Defendants obscure the more fundamental problem that on the face of the ITAR, these labels at best overlap and at worst cannot be differentiated. Plaintiff's Motion at 6:3-7:13; 34:13-35:1. If anything in ITAR must be especially clear, it is these initial categories -- yet they are not. The most glaring example of this vagueness in the definition of defense articles and technical data is presented by the facts in Karn v. U.S. Department of State, 925 F.Supp. 1 (D.D.C. 1996) (appeal docketed). When Mr. Karn submitted a CJ Request for printed cryptographic source code from Mr. Schneier's book, Applied Cryptography, Defendants determined that it was in the public domain. When he submitted virtually the same information to Defendants on a floppy disk, they determined that it was a defense article and so could never be placed into the public domain (an assertion unsupported by the text of the exemption). It is difficult to see how these two determinations can be reconciled, or indeed what clear labelling and realistic distinction between technical data and defense articles compels this conclusion. This categorical distinction, based solely upon medium on which the information written, is not only not narrowly specified in the regulations, it is not even "reasonably clear" from them. B. Defendants' Focus on the Clarity of Their Exemptions Is Misplaced. Instead of defending their vague labels, Defendants insist that the ITAR Scheme survives Plaintiff's vagueness challenge because it contains "reasonably clear" exemptions. This approach virtually concedes the vagueness of the ITAR Scheme's prohibitions. No authority is cited for the proposition that a "reasonably clear" exemption can save a vague law. On the contrary, vagueness doctrine has always focused on defining what is prohibited. Individuals will "'steer far wider of the unlawful zone'" unless "the boundaries of the forbidden areas [are] clearly marked." Baggett v. Bullitt, 377 U.S. 360, 372 (1964), quoting Speiser v. Randall, 357 U.S. 513, 526 (1958); Grayned, 408 U.S. at 108 (law is vague if "its prohibitions are not clearly defined" because person "must know what is prohibited"). 1. The "Public Domain" Exemption is Vague. This fallacy need not be belabored further, however, for even the exemptions are vague. The best example is again the public domain exemption. As demonstrated above, Defendants' application of this exemption depends on the vague distinction between technical data and defense articles, and their false assertions that they leave the determination of what the term includes and how it is applied to those affected. The public domain exception also creates a classic "Catch-22" overbreadth problem. Information is in the "public domain" if it "is published" and "is generally . . . available" to the public in any of several ways. 22 C.F.R. 120.11. What does "is published" mean? Does it include information that is being published at this very moment? Or is it limited to information that is already published? One Congressional committee clearly identified the ambiguity here: The status of technical data under ITAR is summarized this way in a National Telecommunications and Information Administration contractor's document: 'Therefore, if the information is unclassified, once published domestically it is no longer under the control of ITAR. The difficulty is that the act of publication domestically could be interpreted as a violation of the regulation.' House Committee on Government Operations, The Government's Classification of Private Ideas, H. Rept. No. 96-1540, 96th Cong., 2d Sess. 107 (1981) (citation omitted). If "is published" includes "will be published," PROF. BERNSTEIN would be free to publish all his items - the Paper, the Snuffle source code, and the Instructions - without a license. Information which will be published and so made available would be in the "public domain" and not subject to ITAR controls at all. 22 C.F.R. 125.1(a) ("information which is in the public domain. . . is not subject to the controls of this subchapter"). Unfortunately, the above interpretation is not consistent with Defendants' actions. When PROF. BERNSTEIN asked for permission to publish Snuffle, the State Department said no. The most telling example is Plaintiff's conversation with Mr. Charles Ray, who told him that placing information into a library where "anyone could get it" could trigger prosecution under the ITAR. Bernstein Decl., paras. 24-25; see Junger Decl. paras. 9-16. Defendants obviously do not give fair warning of what the ITAR Scheme controls and what it does not, which discourages people from publishing. The only reasonable conclusion is that Defendants equate "is published" with "has been published." So construed, the "public domain" provisions merely allow republication of old material and do not provide protection for PROF. BERNSTEIN and many others who intend to publish new material. Defendants' extended discussion of how something might be "generally available" therefore, misses the point. 2. The Academic Exemption Is Vague. Plaintiff has already noted that the exclusion of "general scientific . . . principles commonly taught in schools, colleges, and universities" from the definition of "technical data" is vague. Plaintiff's Motion at 35:10-36:16. Declarant Lowell characterizes this language as "an assurance to the academic community as to the general non-applicability of the ITAR to a university setting." Defendants' July 25, 1996 letter at 3. It should be obvious that Defendants' chosen phrase "general non-applicability," still permits "specific" application, and so gives no real assurance. Miller Decl., paras. 18-22, 40; Junger Decl., paras. 23-28. More importantly, every time the ITAR Scheme is applied to the academic setting, the State Department has, in fact, "substituted its judgment for that of university or scientific scholar." It did so for Prof. Bernstein, for Prof. Junger, and for Mr. Miller. Even when Defendants do not apply ITAR, its existence produces self-censorship, as illustrated by Dr. Blaze's experiences. Declaration of Matt Blaze, paras. 13-15. The ITAR Scheme is enforced by uncertainty, not the rule of law. When the question of restraints on scientific publication arose during congressional oversight hearings in 1981, the State Department offered the "clarification" from the Munitions Control Newsletter, quoted above at page 10. NRC Report 4-47. The House report issued as a result of these hearings commented that this clarification . . . seems to clarify little while insisting that algorithms can be dangerous if they purport to have advanced cryptologic application. . . . [and] place[s] the burden on the would-be (or unknowing) scientific exporter to find out whether the technical data and the setting require an export license. The researcher in cryptography stands at special risk of being found in violation of these export regulations, judged less by known enforcement activity than by the vagueness of the regulations and the National Security Agency's admonitions. . . . H. Rept. No. 96-1540, at 68. In fact, even Defendants' latest letter on the subject, sent to Plaintiff with his brief, reveals that a license is still required, or at a minimum that the answer is still unclear. First, Defendants state that "Dr. Bernstein does not have to register with this office merely to publish a scientific paper that might contain this technical data. Also, there is no requirement to register for merely teaching a class on cryptography." Defendants' July 25, 1996 letter, page 3. Yet only two paragraphs later, Defendants state "Dr. Bernstein would have to register [as an arms dealer] only if he intends to . . . furnish direct technical assistance to a foreign person to assist in their obtaining or using a defense article." Defendants' July 25, 1996 letter, page 4. If Plaintiff gives a foreign student his technical data in order to assist them in understanding how Snuffle 5.0 works, is he "assisting them in using a defense article" and so a license is required? The Supreme Court's words in Keyishian v. Bd of Regents, 385 U.S. 589 (1987), might have been said about the ITAR Scheme. "It is no answer to say that the statute would not be applied in such a case. We cannot gainsay the potential effect of this obscure wording on 'those with a conscientious and scrupulous regard for such undertakings.'" Keyishian, 385 U.S. at 599, quoting Baggett, 377 U.S. at 374. Like the "complicated plan" in Keyishian, the "very intricacy" and "the uncertainty as to the scope" of the ITAR Scheme's proscriptions "make it a highly efficient in terrorem mechanism. . . . The uncertainty as to the utterances and acts proscribed increases [] caution in 'those who believe the written law means what it says.'" Keyishian, 385 U.S. at 601, quoting Baggett, 377 U.S. at 374. C. The ITAR Scheme Lacks the Necessary Scienter for Civil and Criminal Liability. Defendants have argued that scienter cures the ITAR Scheme's vagueness. That is wrong for two key reasons. First, 22 U.S.C. 2778 expressly does not require scienter for civil liability. On the contrary, one can unknowingly export technical data because "export" is defined as the mere disclosure -- whether knowingly made or not -- of technical data to a foreign person. Cf 22 U.S.C. 2778(c) and 2778(e). The Scheme thus does not require knowledge that the person is foreign, or that the scientific information is deemed "technical data" by Defendants. Because scienter is necessary "to avoid the hazards of self-censorship," Mishkin v. New York, 383 U.S. 502, 511 (1966), the ITAR Scheme's lack of any scienter for civil liability is fatal. Cf. New York Times Co. v. Sullivan, 376 U.S. 255, 277 (1964). Second, the scienter requirement for criminal liability is also insufficient. Generally, the purpose of criminal intent is to target those "who understand the wrongful nature of their conduct." United States v. X-Citement Video, 115 S.Ct. 464, 469 n. 3 (1994). Where protected speech is concerned, however, the First Amendment demands that scienter carefully and specifically focus solely upon those who have wrongful intent. For example, in reviewing the scienter of a statute that punished dissemination of documents connected with the national defense to an agent of a foreign nation, the Supreme Court observed that because "[n]ational defense . . . is a generic concept of broad connotations," the statute's prohibitions would have been vague but for "[t]he obvious delimiting words . . . requiring intent or reason to believe that the information to be obtained is to be used to the injury of the United States, or to the advantage of any foreign nation." Gorin v. United States, 312 U.S. 19, 26-28 (1941) (internal quotation marks omitted); see United States v. Heine, 151 F.2d 813, 815 (2nd Cir. 1945) cert den., 328 U.S. 833 (1946) (discussing breadth of "national defense"). Thus, where criminal liability is concerned, the ITAR Scheme must at least contain the narrow scienter imposed in Gorin. See Edler, 579 F.2d at 521 (citing Gorin). Courts have, however, interpreted ITAR Scheme scienter so unclearly as to leave the publisher of scientific information at peril. At least two circuits have held that one need not specifically know that an export license is needed, only that one has a duty not to export the articles. United States v. Beck, 615 F.2d 441, 450 (7th Cir. 1980); United States v. Tsai, 954 F.2d 155 (3rd Cir. 1992) cert den. 506 U.S. 830 (1992). Similarly, the First Circuit has held that specific intent does not require "proof that the defendant know the arms are on the [USML]." United States v. Murphy, 852 F.2d 1, 7 n. 6 (1st Cir. 1988), cert den. 489 U.S. 1022 (1989) But, because under the ITAR one's duty not to "export" is the same as one's duty not to speak or publish without a license, speakers are not sufficiently protected under Gorin. PROF. BERNSTEIN knew that the items he submitted in his CJ request could be on the USML. What he did not know with certainty, however, was whether the items were on the list -- or if they were, whether they were defense articles, technical data, or subject to one of the exemptions. Whether information is in fact on the USML, and whether one's communication is in fact an export, are "the crucial element[s] separating legal innocence from wrongful conduct." X-Citement Video, 115 S.Ct. at 469. Without knowledge of them, there is no "purpose to disobey the law." Ratzlaf v. United States, 510 U.S. 135, 141 (1994) (quotations omitted). To be clear, scienter must include knowing that what is done is unlawful or at least so wrong that it is probably unlawful. V. WITHOUT FURTHER SAVING CONSTRUCTION, THE ITAR SCHEME IS EITHER ULTRA VIRES OR AN UNCONSTITUTIONAL DELEGATION OF LEGISLATIVE POWER. Nothing in the statutory text or legislative history of 22 U.S.C. 2778 authorizes licensing of Plaintiff's speech. Properly construed, the statute only authorizes regulation of commercial, military exports. See 22 U.S.C. 2794(7)("defense articles and defense services "with respect to commercial exports subject to 2778). Defendants argue that the authority to control arms exports "[i]n furtherance of world peace and the security and foreign policy of the United States," 22 U.S.C. 2778(a)(1), justifies their sweeping regulations. But this argument proves too much. There are no limits on what may be placed on the USML. 22 U.S.C. 2794(7). There is no judicial review of Defendants' decision that something is on the USML. 22 U.S.C. 2778(h). Thus, with respect to speech, either the regulations are ultra vires or the statute violates the nondelegation doctrine. A. The ITAR Scheme Is Ultra Vires. 1. Commercial Transactions: The Core Meaning of the AECA . That Defendants' interpretation of the AECA extends well beyond the plain meaning of the statute is easily seen in their extension of it to noncommercial "exports, since this extension, if enforced, requires that this Court ignore the very words of the statute. Settled law provides that courts "will avoid a reading which renders some words altogether redundant." Gustafson v. Alloyd Co., Inc., 115 S. Ct. 1061, 1069 (1995). The AECA states that defense articles and defense services "means, with respect to commercial exports subject to the provisions of section 38 of this Act [22 U.S.C. 2778], those items designated by the President" as being on the USML. 22 U.S.C. 2794(7) (emphasis added). The phrase "with respect to commercial exports" is superfluous, however, if the statute is interpreted as reaching non-commercial matters. Yet this is precisely what the ITAR regulations do. They reach all disclosures of technical data and all "exports" of software, commercial or not, as seen in Defendants' application to of the ITAR Scheme to Plaintiff, Mr. Demberger, Mr. Zimmermann, Mr. Miller and Professor Junger. Such interpretation obviously violates "the settled rule that a statute must, if possible, be construed in such fashion that every word has some operative effect." United States v. Nordic Village, Inc., 503 U.S. 30, 36 (1992). Sec. 2778 provides further textual support for this construction, because its terms assume that it applies only to commercial matters. E.g., 22 U.S.C. 2778(b)(1)(A) (registration requirement for those "in the business of manufacturing, exporting or importing"); 2778(g)(2) (applicants must "identify . . . all consignees and freight forwarders involved in the proposed export"); 2778(g)(9)(D) (defining "party to the export" as "senior officers of the license applicant," "freight forwarders or designated export agent," and "any consignee or end user of any item to be exported"). To ascertain plain meaning, the court may also analyze "the language and design of the statute as a whole." K-Mart v. Cartier, Inc., 486 U.S. 281, 291 (1988). This analysis reinforces Plaintiff's construction. 2778 is part of a law with a clear, stated purpose: "Statutory requirements for controls over commercial and government-to-government military exports are revised and combined in a new Arms Export Control Act so that all military export matters are controlled under one statute." S. Rep. No. 94-876 at 4. Congress's main concern was to control government military exports such as "grant military assistance" and "foreign military sales." H.Rep. No. 94-1144, at 1-2, reprinted at 1976 USCCAN 1378, 1385-86. The prior statutory framework for arms export, however, was also "outdated" because it gave "complete discretion to the President" over "[c]ommercial sales." S.Rep. No. 94-876, at 9 (quotation and citation omitted). That the "sale of defense articles through commercial channels is increasing each year" was one of five key conclusions emphasized by the House. H. Rep. No. 94-1144, at 12, 1976 USCCAN at 1388. Thus, Congress sought to regulate "government-to-government military sales," "sale of defense articles through commercial channels," and defense services like "military training." Ibid In short, AECA as a whole was intended to regulate government-to-government and commercial military exports and 2778 is the portion of the statute reaching the commercial military exports. The caselaw supports this conclusion as well. Plaintiff is unaware of any AECA case involving a noncommercial disclosure of information. The Ninth Circuit construed the ITAR as "control[ling] the conduct of assisting foreign enterprises to obtain military equipment and related technical expertise." United States v. Edler Industries, Inc., 579 F.2d 516, 521 (9th Cir. 1978). Edler specifically addresses "transmissions of information in a commercial context." Id., at 519-520. Other cases have assumed that AECA/ITAR governs "commercial export of arms and ammunition." United States v. Adames, 878 F.2d 1374, 1376 (11th Cir. 1989); United States v. Beck, 615 F.2d 441, 444 (7th Cir. 1980) (same); see United States v. Evans, 667 F.Supp. 974, 982 (S.D.N.Y. 1987) ("exports, Government and commercial"); cf. United States v. Durrani, 835 F.2d 410, 420 (2nd Cir. 1987) ("foreign military sales"). Thus, the plain meaning of 2778 and the caselaw interpreting it exclude noncommercial speech such as Plaintiff's here. 2. AECA Does Not Apply to Domestic Speech. The foregoing analysis also shows that AECA does not authorize control of noncommercial speech. In addition, the AECA was also not intended to apply to purely domestic speech, even if it is commercial. Simply put, the AECA is about "export," a term applied to transborder transactions. In 1978, the OLC stated that "further Congressional authorization would obviously be necessary in order to extend governmental controls to domestic as well as foreign disclosures of public cryptographic information." Tien Decl. Ex.A, Bates #60088. In the absence of a specific definition in the AECA, "export" must be construed "in accordance with its ordinary or natural meaning. FDIC v. Meyer, 510 U.S. 471, 476 (1994). "To export means to carry or send abroad; to import means to bring into the country. Those acts begin and end at water's edge." Canton R.R. Co. v. Rogan, 340 U.S. 511, 515 (1951) (constitutionality of state tax under the Import-Export Clause, U.S. Const., Art. I, 10, cl. 2). Canton, while not directly on point, strongly suggests that disclosures within the United States cannot be an export. At the very least, it is not an export for Plaintiff to disclose technical data to a foreign national within the United States; there is no "carry[ing] abroad," no crossing of the "water's edge." See United States v. Posey, 846 F.2d at 1496-97 (contrasting "export abroad" with "domestic disclosure"). 3. AECA Does Not Apply to General Publication. Defendants' interpretation poses massive constitutional difficulties mainly because the ITAR sweeps in as "export" the publication of scientific and technical information. Although the Ninth Circuit sought to limit the ITAR's reach in Edler, that effort failed, in part because the Ninth Circuit did not analyze the issue of overbroad definition of export in the ITAR Scheme. The ITAR Scheme reaches any disclosure of technical data to a foreign person, whether in newspapers, in books, face to face or through the Internet. 22 C.F.R. 120.17(a)(4). This definition on its face restricts freedom of expression. Lovell, 303 U.S. at 452 ("press in its historic connotation comprehends every sort of publication which affords a vehicle of information and opinion.") As Plaintiff has outlined above, neither noncommercial transactions nor purely domestic speech are properly within the AECA's reach. But this attempt at saving the AECA, if not the ITAR, does not protect the author of a book; the worldwide sale of a book like Mr. Schneier's Applied Cryptography could conceivably be considered a "commercial export." A clear statutory statement is a precondition of any authorization of executive power to regulate First Amendment rights. Schneider v. Smith, 390 U.S. 17, 25-27 (1968); Greene v. McElroy, 360 U.S. 474, 507 (1959). As the OLC noted that "it is by no means clear . . . that Congress intended that the President regulate noncommercial dissemination of information, or considered the problems such regulation would engender." Tien Declaration, Ex. A, at Bates #60077. Prudential considerations therefore dictate that the AECA does not apply to general publication, because "where an otherwise acceptable construction of a statute would raise serious constitutional problems, the Court will construe the statute to avoid such problems unless such construction is plainly contrary to the intent of Congress." Edward J. DeBartolo Corp. v. Florida Gulf Coast Bldg & Constr Trades Council, 485 U.S. 568, 575 (1988). B. As Interpreted By Defendants, The Scheme Violates The Non-Delegation Doctrine. Plaintiff here contends that the delegation by Congress in the AECA contains standards to which Defendants have failed to conform. Defendants appear to argue that the delegation of power to further "world peace and the security and foreign policy of the United States" is sufficiently broad to encompass their circular rules and their various actions. If we accept Defendants contention, arguendo, then it is clear that the delegation at issue here is impermissibly standardless and thus unconstitutional. See A.L.A. Schechter Poultry Corp. v. United States, 295 U.S. 495, 529-30 (1935); Panama Ref. Co. v. Ryan, 293 U.S. 388, 421 (1935). World peace and national security are goals, not standards which sufficiently mark the field within which the bureaucrat is to act so that it may be known whether he has kept within it in compliance with the legislative will. Yakus v. United States, 321 U.S. 414, 425 (1944), compare Touby v. United States, 500 U.S. 160 (1991). "While legislatures ordinarily may delegate power under broad standards ..., [the] area of permissible indefiniteness narrows ... when the regulation ... potentially affects fundamental rights like those protected by the First Amendment." United States v. Robel, 389 U.S. 258, 274-75 (1967) (Brennan, J., concurring). Obviously, the ITAR affects domestic speech. To make matters worse, there is no judicial review of Defendants' determinations that Snuffle 5.0 is a defense article or that the Instructions are technical data. 22 U.S.C. 2778(h). "[T]he use of the result of an administrative proceeding to establish an element of a criminal offense is troubling." United States v. Mendoza-Lopez, 481 U.S. 828, 838 n. 15 (1987)(requiring collateral challenge to use of fundamentally unfair deportation proceeding). There must be judicial review "before the administrative order may be used to establish conclusively an elelment of a criminal offense." Id. at 838. Plaintiff has suggested various problems which must be addressed in order to save the ITAR Scheme. But no saving construction can exist if Defendants' view of the statute is acceptable because the definition of "defense articles and defense services" in 22 U.S.C. 2794(7) would then be wholly discretionary and judicial review of such determinations would be precluded by 22 U.S.C. 2778(h). Defendants would retain (in the AECA) the unreviewable discretion to decide what is controlled. For this reason, the Court must either construe the AECA as regulating only commercial, military exports or must invalidate both the ITAR and the AECA on their face as excessive delegations of constitutional rights. CONCLUSION The ITAR Scheme must be judged by its own words and plain meaning and by Defendants' real world actions and statements in applying it, both to Plaintiff and to others. Defendants' would have this Court instead base its judgment on post-litigation, often demonstrably false, assertions about the scope, meaning and application of the ITAR Scheme. In sum, the ITAR Scheme is a a prior restraint on speech about cryptography. It censors that speech as technical data, as computer code and as the very broadly defined non-code areas of the term "software." It is content-based, is aimed at and applied to censor communication and is so overbroad and vague that Defendants' themselves don't even know its contours, much less the many private scientists, academics and other citizens who face severe civil and criminal penalties if they fail to abide by it. The root of most of these problems can be summed up in a single word: discretion. The ITAR Scheme, starting with the AECA, simply grants too much discretion to Defendants, allowing them to extend the reach of the Scheme well beyond its facial and intended purpose of controlling commercial arms export, and into the realm of First Amendment protected expression. Because of this discretion, the statute must be invalidated on its face. Based upon the foregoing, Plaintiff respectfully requests that Defendants' Motion for Summary Judgment be denied and that Plaintiff's Motion for Partial Summary Judgment be granted as described below: 1. Procedural Prior Restraint: Plaintiff contends that, to the extent that it requires a license prior to protected expression, the AECA must be invalidated on its face on the grounds that it fails to provide adequate procedural safeguards required by Freedman v. Maryland and its progeny. 2. Substantive Prior Restraint. Plaintiff contends that the ITAR must be invalidated to the extent it censors "software with the capability of maintaining secrecy, confidentiality or of information or information systems" and related information ("speech about cryptography"), on the grounds that the Government has failed to meet its burden to show that "direct, immediate and irreparable damage" will result from the communication of this protected expression. 3. Subsequent Punishment. Plaintiff contends that the standard of strict scrutiny applies and that Defendants have failed in their burden to show that, to the extent they extend to speech about cryptography, the subsequent punishment provisions of the ITAR and AECA meet that standard. Accordingly, each must be invalidated on its face. 4. Vagueness. Plaintiff contends that the AECA is vague in that it fails to limit agency discretion to regulate protected expression. The ITAR is also vague, as to the categories of "defense article," "defense service" and "technical data", its exemptions, and its definition of "export" among others, creating further confusion and a chilling effect in practice. As a result, both the statute and the regulation must be invalidated on their face and as applied to Plaintiff. 5. Overbreadth. Plaintiff contends that the ITAR is overbroad, both on its face and as applied to Plaintiff, since it reaches a substantial amount of First Amendment protected expression. As a result, the ITAR must be invalidated on its face to the extent that it reaches such protected speech or speech activities and invalidated as applied to Plaintiff since his activities are non-commercial, academic publication. 6. Ultra Vires or Unconstitutional Delegation of Power To the extent that the ITAR Scheme includes regulation of noncommercial speech, domestic speech, publication or scientific exchange, either the ITAR regulations are ultra vires or the AECA has unconstitutionally delegated discretionary authority over speech to Defendants. Plaintiff believes that it is the former, and so contends that the ITAR must be invalidated to the extent that it controls such speech. If the Court determines that it is the latter, the AECA must be invalidated on its face. 7. Immediate Effect The Constitutional violations inherent in the ITAR Scheme continue every day and both the direct actions of Defendants and their chilling effect continues to harm the growth of the science of cryptography. Plaintiff has now been restrained from his publication for over three years and every day that this restraint continues, both he and the science of cryptography are harmed, as the Declarations demonstrate. Others are similarly being restrained. Further, Defendants are actively and improperly using the ITAR Scheme to control the methods of cryptography which are developed and used domestically. For these reasons, Plaintiff respectfully requests that this Court both invalidate the statute and regulations on their face and as applied, and order that, pending the probable appeal of that decision, Defendants are enjoined entirely from applying the ITAR Scheme in any way to speech about cryptography, including source code. Dated:___________________ Respectfully Submitted, McGLASHAN & SARRAIL Professional Corporation CINDY A. COHN Attorneys for Plaintiff Daniel J. Bernstein