Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 1 of 90 No. 16-17165 UNITED STATES COURT OF APPEALS FOR THE NINTH CIRCUIT TAMARA FIELDS, et al., Plaintiffs-Appellants, v. TWITTER, INC., Defendant-Appellee. On Appeal from the United States District Court for the Northern District of California, No. 3:16-cv-00213-WHO (Judge William H. Orrick) BRIEF FOR DEFENDANT-APPELLEE TWITTER, INC. MARK D. FLANAGAN WILMER CUTLER PICKERING HALE AND DORR LLP 950 Page Mill Road Palo Alto, CA 94304 (650) 858-6000 SETH P. WAXMAN PATRICK J. CAROME ARI HOLTZBLATT WILMER CUTLER PICKERING HALE AND DORR LLP 1875 Pennsylvania Avenue NW Washington, DC 20006 (202) 663-6000 May 31, 2017 Attorneys for Defendant-Appellee Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 2 of 90 CORPORATE DISCLOSURE STATEMENT Pursuant to Rule 26.1 of the Federal Rules of Appellate Procedure and Ninth Circuit Rule 28(b), Defendant-Appellee Twitter, Inc. (“Twitter”) states as follows: Twitter is a publicly held corporation and does not have a parent corporation. No publicly traded corporation owns ten percent or more of its stock. -i- Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 3 of 90 TABLE OF CONTENTS Page CORPORATE DISCLOSURE STATEMENT ..........................................................i TABLE OF AUTHORITIES ....................................................................................iv INTRODUCTION .....................................................................................................1 JURISDICTION.........................................................................................................3 ISSUES PRESENTED...............................................................................................3 STATUTORY PROVISIONS ...................................................................................4 COUNTERSTATEMENT OF THE CASE ............................................................... 4 A. The Original And First Amended Complaints ...................................... 5 B. The District Court’s Initial Ruling ........................................................ 8 C. The Second Amended Complaint .......................................................11 D. The District Court’s Second Ruling ....................................................12 SUMMARY OF ARGUMENT ...............................................................................16 STANDARD OF REVIEW .....................................................................................18 ARGUMENT ...........................................................................................................19 I. THE DISTRICT COURT CORRECTLY HELD THAT SECTION 230 REQUIRED DISMISSAL OF PLAINTIFFS’ CLAIMS ..............................................19 A. Section 230 Broadly Immunizes Online Service Providers From Liability For Injuries Arising From Third-Party Content.............................................................................19 B. Plaintiffs Cannot Circumvent Section 230 Based On Their Account-Provision Theory ........................................................21 - ii - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 4 of 90 II. 1. The Second Amended Complaint Purports To Hold Twitter Liable For Publishing Third-Party Content ......................................................................................23 2. Decisions About Whom To Let Speak On A Provider’s Platform Are Publishing Decisions .........................24 3. Plaintiffs Cannot Circumvent Section 230 By Confining Allegations Expressly Mentioning Third-Party Content To The Causation Section Of Their Complaint ........................................................................33 C. The Transmission Of Direct Messages Is Publishing Activity And Subject To Section 230 Immunity ................................38 D. Allowing This Case To Proceed Would Give Rise To Precisely The Harms Section 230 Was Enacted To Prevent .................................................................................................43 THE DISTRICT COURT CORRECTLY DISMISSED PLAINTIFFS’ CLAIMS ON THE ALTERNATIVE GROUND THAT THEIR ALLEGATIONS FAIL TO SATISFY THE ATA’S PROXIMATE CAUSATION ELEMENT.....................................................................................47 A. The ATA Requires Plaintiffs To Plead Facts That Plausibly Establish A Direct Relationship Between An Act Of International Terrorism Committed By Twitter And Plaintiffs’ Injuries ........................................................................48 B. Plaintiffs’ Allegations Fail Even Under Their Preferred Test For Causation...............................................................................54 CONCLUSION ........................................................................................................60 STATEMENT OF RELATED CASES CERTIFICATE OF COMPLIANCE STATUTORY ADDENDUM CERTIFICATE OF SERVICE - iii - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 5 of 90 TABLE OF AUTHORITIES CASES Page(s) Al Haramain Islamic Foundation v. United States Department of Treasury, 686 F.3d 965 (9th Cir. 2011).........................................................45 Anza v. Ideal Steel Supply Corp., 547 U.S. 451 (2006)...........................................................................49, 52, 53 Ashcroft v. Iqbal, 556 U.S. 662 (2009).......................................................................................18 Bank of America Corp. v. City of Miami, 137 S. Ct. 1296 (2017)...................................................................................50 Barnes v. Yahoo!, Inc., 570 F.3d 1096 (9th Cir. 2009) ................................................................passim Barrett v. Rosenthal, 146 P.3d 510 (2006) ................................................................................45, 46 Batzel v. Smith, 333 F.3d 1018 (9th Cir. 2003) .....................................................20, 27, 28, 45 Blumenthal v. Drudge, 992 F. Supp. 44 (D.D.C. 1998)................................................................28, 29 Boim v. Holy Land Foundation for Relief & Development, 549 F.3d 685 (7th Cir. 2008) .........................................................................51 Boim v. Quranic Literacy Institute & Holy Land Foundation for Relief & Development, 291 F.3d 1000 (7th Cir. 2002) ............................................50 Burrage v. United States, 134 S. Ct. 881 (2014).....................................................................................54 Carafano v. Metrosplash.com, Inc., 339 F.3d 1119 (9th Cir. 2003) .....................................................20, 31, 36, 43 Cohen v. Facebook, Inc., 2017 WL 2192621 (E.D.N.Y. May 18, 2017) ...................................21, 26, 29 - iv - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 6 of 90 Couch v. Cate, 379 F. App’x 560 (9th Cir. 2010) ............................................................49, 50 Delfino v. Agilent Technologies, Inc., 52 Cal. Rptr. 3d 376 (Ct. App. 2006) ............................................................41 Directory Assistants, Inc. v. Supermedia, LLC, 884 F. Supp. 2d 446 (E.D. Va. 2012) ............................................................42 Doe v. MySpace, Inc., 528 F.3d 413 (5th Cir. 2008) .................................................16, 25, 31, 34, 35 Doe II v. MySpace, Inc., 96 Cal. Rptr. 3d 148 (Ct. App. 2009) ............................................................25 Doe No. 1 v. Backpage.com, LLC, 817 F.3d 12 (1st Cir. 2016).....................................................................passim Doe No. 14 v. Internet Brands, Inc., 824 F.3d 846 (9th Cir. 2016) ...................................................................36, 37 FAA v. Cooper, 566 U.S. 284 (2012).......................................................................................40 Fair Housing Council of San Fernando Valley v. Roommates.com, LLC, 521 F.3d 1157 (9th Cir. 2008) .......................................................passim Goddard v. Google, Inc., 2008 WL 5245490 (N.D. Cal. Dec. 17, 2008) ..............................................21 Hemi Group, LLC v. City of New York, 559 U.S. 1 (2010)...........................................................................................49 Holder v. Humanitarian Law Project, 561 U.S. 1 (2010)...........................................................................................44 Holmes v. Securities Investor Protection Corp., 503 U.S. 258 (1992)...........................................................................48, 49, 52 In re Terrorist Attacks on September 11, 2001, 349 F. Supp. 2d 765 (S.D.N.Y. 2005) ...........................................................19 -v- Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 7 of 90 In re Terrorist Attacks on September 11, 2001, 714 F.3d 118 (2d Cir. 2013) ................................................................2, 48, 57 Jane Doe One v. Oliver, 755 A.2d 1000 (Conn. Sup. Ct. 2000) ...........................................................42 Jones v. Dirty World Entertainment Recordings LLC, 755 F.3d 398 (6th Cir. 2014) .........................................................................21 Kimzey v. Yelp! Inc., 836 F.3d 1263 (9th Cir. 2016) .......................................................2, 18, 33, 35 Klayman v. Zuckerberg, 753 F.3d 1354 (D.C. Cir. 2014).........................................................18, 21, 24 Linde v. Arab Bank, PLC, 97 F. Supp. 3d 287 (E.D.N.Y. 2015), ............................................................59 Mattel v. MGA Entertainment, Inc., 782 F. Supp. 2d 911 (C.D. Cal. 2011) ...........................................................50 Nemet Chevrolet, Ltd. v. Consumeraffairs.com, Inc., 591 F.3d 250 (4th Cir. 2009) .........................................................................18 Peters v. LifeLock Inc., 2014 WL 12544495 (D. Ariz. Sept. 19, 2014) ..............................................42 Phan v. Pham, 105 Cal. Rptr. 3d 791 (Ct. App. 2010) ..........................................................41 Pride v. Correa, 719 F.3d 1130 (9th Cir. 2013) .......................................................................18 Reed v. Town of Gilbert, 135 S. Ct. 2218 (2015)...................................................................................28 Rothstein v. UBS AG, 708 F.3d 82 (2d Cir. 2013) ....................................................48, 52, 53, 54, 57 Stansell v. BGP, Inc., 2011 WL 1296881 (M.D. Fla. Mar. 31, 2011) ..............................................55 - vi - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 8 of 90 Stratton Oakmont, Inc. v. Prodigy Services Co., 1995 WL 323710 (N.Y. Sup. Ct. May 24, 1995) ..........................................39 Tanisha Systems, Inc. v. Chandra, 2015 WL 10550967 (N.D. Ga. Dec. 4, 2015) ...............................................42 Turner Broadcasting System, Inc. v. FCC, 512 U.S. 622 (1994).......................................................................................60 Universal Communications Systems, Inc. v. Lycos, Inc., 478 F.3d 413 (1st Cir. 2007)........................................................16, 24, 25, 45 Witkoff v. Topix, LLC, 2015 WL 5297912 (Cal. Ct. App. Sept. 10, 2015) ........................................21 Zeran v. America Online, Inc., 129 F.3d 327 (4th Cir. 1997) .......................................................20, 36, 39, 45 STATUTES AND RULES 18 U.S.C. § 2331 ............................................................................................................55 § 2333 ..............................................................................................1, 2, 48, 51 § 2339A..........................................................................................................55 § 2339B ....................................................................................................45, 55 47 U.S.C. § 230 ................................................................................................ passim Clayton Act, 15 U.S.C. § 15 ....................................................................................48 Justice Against Sponsors of Terrorism Act, Pub. L. No. 114-222 (enacted Sept. 28, 2016) ................................................................................51 RICO, 18 U.S.C. § 1964 ..........................................................................................48 Sherman Act, 26 Stat. 209 (1890)............................................................................48 Fed. R. Civ. P. 12 .....................................................................................................18 OTHER AUTHORITIES Prosser & Keaton on Torts (5th ed. 1984) ...................................................17, 40, 41 Restatement (Second) of Torts ...........................................................................17, 40 - vii - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 9 of 90 INTRODUCTION Plaintiffs are surviving family members of Lloyd Fields, Jr., and James Damon Creach—two U.S. citizens killed in a terrorist attack in Jordan by former Jordanian police captain Anwar Abu Zaid. Rather than sue any of the individuals or groups that may have authorized, planned, or committed that heinous crime, Plaintiffs have instead sued Twitter, Inc., a provider of free online services used by hundreds of millions of people around the world. Plaintiffs do not allege that Abu Zaid ever used Twitter’s services. Nor do they allege that Twitter’s services were used in any way in connection with Abu Zaid’s attack. Instead, Plaintiffs attempt to hold Twitter liable on the theory that Twitter’s ubiquitous online platform was allegedly used by others—members and sympathizers of the Islamic State of Iraq and Syria (“ISIS”)—to transmit messages promoting their views and agenda. According to Plaintiffs, Twitter’s alleged failure to prevent those individuals from using its services renders it liable for the deaths of Mr. Fields and Mr. Creach under the civil remedy provision of the federal Anti-Terrorism Act (“ATA”), 18 U.S.C. § 2333(a). Twitter deeply sympathizes with Plaintiffs for their losses and is committed to combatting the spread of terrorist content online. But Twitter is not liable for Abu Zaid’s appalling attack. As the district court correctly concluded, Plaintiffs’ claims fail as a matter of law for two fundamental reasons. -1- Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 10 of 90 First, Plaintiffs’ claims are barred by 47 U.S.C. § 230 because they attempt to hold Twitter liable for allegedly failing to prevent third parties from posting content that allegedly caused Plaintiffs’ injuries. This Court repeatedly has held that Section 230 immunizes online service providers from liability for performing or not performing “traditional editorial functions” relating to third-party content, such as disseminating or failing to block or remove objectionable content created by their users. E.g., Fair Hous. Council of San Fernando Valley v. Roommates.com, LLC, 521 F.3d 1157, 1184 (9th Cir. 2008) (en banc); Kimzey v. Yelp! Inc., 836 F.3d 1263, 1266 (9th Cir. 2016). Here, Plaintiffs seek to do exactly that—hold Twitter liable on the theory that Twitter failed to prevent members and supporters of ISIS from opening Twitter accounts that transmitted content that somehow may have indirectly caused Abu Zaid’s attack. Accepting such a theory would impose liability on Twitter for publishing other people’s information— precisely the result Section 230 forbids. Second, Plaintiffs fail to allege facts that plausibly establish injury “by reason of”—or proximately caused by—an act of international terrorism committed by Twitter, as the ATA’s civil remedy provision requires. 18 U.S.C. § 2333(a); see also In re Terrorist Attacks on Sept. 11, 2001, 714 F.3d 118, 123125 (2d Cir. 2013) (“Al Rajhi”). Countless intervening acts and actors separate Twitter’s supposedly unlawful conduct—allegedly failing to block members and -2- Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 11 of 90 sympathizers of ISIS from using Twitter’s platform—from Abu Zaid’s heinous crimes. Plaintiffs therefore cannot show that Twitter’s conduct “led directly” to their injuries—which is the proximate cause standard that the Supreme Court has applied whenever a federal statute includes the “by reason of” language found in the ATA. Nor can Plaintiffs satisfy the more relaxed standard they erroneously propose—which would require them to allege facts plausibly demonstrating that Twitter’s conduct was a “substantial factor” in causing injuries that were “reasonably foreseeable.” The district court correctly held that Plaintiffs’ claims fail to satisfy either standard of causation, thus providing an independent ground for dismissal. JURISDICTION Twitter agrees with Plaintiffs’ jurisdictional statement (Br. 3). ISSUES PRESENTED 1. Whether the district court correctly held that the immunity for online intermediaries created by Section 230 bars imposing liability on Twitter for its alleged failure to prevent members or sympathizers of a terrorist group from using its globally available online platform and direct messaging services to transmit third-party content. 2. Whether the district court correctly held that Plaintiffs failed to state a claim for relief under the civil remedy provision of the Anti-Terrorism Act because -3- Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 12 of 90 they did not plead facts plausibly establishing that Twitter proximately caused Plaintiffs’ injuries. STATUTORY PROVISIONS An addendum to this brief sets forth all pertinent statutory provisions. COUNTERSTATEMENT OF THE CASE Plaintiffs are the widows of Lloyd “Carl” Fields, Jr. and James Damon Creach and two surviving children of Mr. Creach. AER28 ¶¶ 2-5. On November 9, 2015, Anwar Abu Zaid shot and killed Mr. Fields and Mr. Creach in an attack on the International Police Training Center in Amman, Jordan, where Mr. Fields and Mr. Creach were working as government contractors. AER9. Abu Zaid was a Jordanian police captain who had been studying at the Center. Id. Twitter, the Defendant, operates a global Internet platform for public selfexpression and conversation used by hundreds of millions of people that is free of charge and almost universally available. AER32 ¶ 35. Users with a Twitter account can send “Tweets”—messages of 140 characters or less, sometimes with picture or video—to anyone who has chosen to “follow” that user, and those Tweets typically may also be viewed by anyone with access to the Internet. E.g. AER29, 35-36 ¶¶ 11, 51, 55. Other users may, in turn, “retweet” those messages to their own followers. E.g., AER35, 37-38 ¶¶ 48, 68. Using Direct Messaging, users can also communicate “‘with one person or many’” by sending a message -4- Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 13 of 90 only to specific user accounts, rather than making the message available to all Twitter users. AER33 ¶ 43. Users further express themselves through the user name, photograph, and self-description they select for their public profiles. AER29, 32 ¶¶ 12, 39. Each day, Twitter’s users send and share hundreds of millions of Tweets and messages touching on a broad range of topics—from a special election to a joke told on late-night television to the announcement of a new job. See AER32 ¶ 36. A. The Original And First Amended Complaints Plaintiff Tamara Fields commenced this action on January 13, 2016. AER47. After Twitter moved to dismiss, Plaintiffs filed a First Amended Complaint, which, among other things, added Mr. Creach’s widow and children as plaintiffs. AER49; SER4 ¶¶ 8-10. The First Amended Complaint did not allege that Twitter committed the shooting in Jordan in which Mr. Fields and Mr. Creach were killed. Nor did it allege that Abu Zaid, the shooter, was recruited over Twitter or that he (or anyone else) used Twitter to plan, carry out, or raise funds for the attack. SER22 (First Op.). It did not even allege that Abu Zaid ever had an account on Twitter, ever viewed a Tweet, or ever used Twitter’s platform in any way. Id. Plaintiffs instead attempted to tie Twitter to the shooting through a series of other alleged connections—first between Abu Zaid and ISIS and then between -5- Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 14 of 90 members and sympathizers of ISIS and Twitter. Attempting to link Abu Zaid to ISIS, the First Amended Complaint alleged that Abu Zaid’s brother told reporters that Abu Zaid had been inspired to commit his heinous crime by ISIS’s “brutal execution of Jordanian pilot Maaz al-Kassabeh in February 2015.” SER17 ¶ 84. According to that complaint, ISIS allegedly had publicized that execution by, among other things, posting messages about it on Twitter’s platform. Id. The First Amended Complaint did not allege that Abu Zaid learned about the execution through Twitter’s platform, or otherwise explain how he learned of it. See id.; see also SER22 (First Op.). The First Amended Complaint further alleged that ISIS claimed responsibility for Abu Zaid’s attack in a statement that described him as a “lone wolf.” SER16 ¶ 80. The First Amended Complaint separately attempted to link ISIS to Twitter. It alleged that the hundreds of millions of users of Twitter’s online services included some individuals who were affiliated with or sympathetic to ISIS. SER2-3 ¶¶ 1-5. It further alleged that some of those individuals promoted ISIS’s activities and/or agenda by sending Tweets and Direct Messages to recruit new terrorists, raise funds, and spread propaganda. SER5-10 ¶¶ 19-47. For example, it alleged that sympathizers of ISIS had Tweeted “a notorious promotional training video, ‘Flames of War’” as an illustration of how ISIS used Twitter to recruit new -6- Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 15 of 90 members, SER7 ¶ 26, and it highlighted images of alleged fundraising efforts by ISIS on Twitter, SER7-8 ¶¶ 32, 34. Plaintiffs acknowledged that Twitter has rules regarding the types of content permitted on its platform and that those rules “prohibit ‘threats of violence … including threatening or promoting terrorism.’” SER13-14 ¶¶ 66, 70. But they alleged that Twitter had not always blocked such content and instead “knowingly permitted the terrorist group ISIS to use its social network as a tool for spreading extremist propaganda, raising funds, and attracting new recruits.” SER2 ¶ 1. And they criticized Twitter for (allegedly) enforcing its rules only in response to reports from users. SER13 ¶ 66. According to Plaintiffs, that was not good enough: Twitter instead should have “proactively” “monitor[ed]” hundreds of millions of tweets, posts, videos, and other content transmitted over the Twitter platform, and pre-emptively “censor[ed] user content,” “shut down … ISIS-linked account[s],” and blocked ISIS-related accounts from “springing right back up.” SER13 ¶¶ 66, 69. Based on these allegations, Plaintiffs claimed that Twitter was liable for treble damages for the deaths of Mr. Fields and Mr. Creach under the civil remedy provision of the ATA. In Count I, Plaintiffs sought to hold Twitter directly liable on the theory that it “purposefully, knowingly or with willful blindness” provided “material support” to ISIS in the form of “services and support” in violation of 18 -7- Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 16 of 90 U.S.C. § 2339A, and that that “material support” was “a proximate cause of the injury inflicted on Plaintiffs.” SER18 ¶¶ 88-90. Count II asserted a parallel theory under 18 U.S.C. § 2339B. SER18-19 ¶¶ 92-94. B. The District Court’s Initial Ruling Twitter moved to dismiss the First Amended Complaint and the district court granted that motion, SER21-35, holding that Section 230 barred Plaintiffs’ claims and that Plaintiffs had failed to plead facts sufficient to establish proximate cause. With respect to Section 230, the court explained that the statute “‘immunizes providers of interactive computer services against liability arising from content created by third parties’” so long as the defendant is “(a) a provider or user of an interactive computer service (b) that the plaintiff seeks to treat as publisher or speaker (c) of information provided by another information content provider.” SER25. The court further explained that Plaintiffs disputed only the middle element—“whether they seek to treat Twitter as a publisher or speaker”—and that they did so based on two theories: (1) Plaintiffs contended that “their claims are not based on ‘the content of tweets, the issuing of tweets, or the failure to remove tweets,’” but rather on Twitter’s “‘provision of Twitter accounts to ISIS in the first place.’” SER25-27 (quoting Plaintiffs’ opposition). (2) Plaintiffs “highlight[ed] their allegations regarding Twitter’s Direct Messaging capabilities, and assert[ed] that ‘[b]ecause these private messages are not published …, a lawsuit based on -8- Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 17 of 90 their content is not barred by [Section 230].’” SER27 (quoting Plaintiffs’ opposition). The court rejected both theories. As to Plaintiffs’ so-called account-provision theory, the court identified three fundamental defects. First, the court found the theory did not actually “align with the allegations in the [First Amended Complaint],” which was “riddled with detailed descriptions of ISIS-related messages, images, and videos disseminated through Twitter and the harms allegedly caused by the dissemination of that content.” SER27-28. Second, according to the court, even if Plaintiffs had pled an accountprovision theory, it “would be just as barred by section 230(c)(1)” as a theory directly challenging Twitter’s decisions about “what [particular] third-party content may be posted.” SER29. “Under either theory,” the court explained, “the alleged wrongdoing is the decision to permit third parties to post content—it is just that under plaintiffs’ provision of accounts theory, Twitter would be liable for granting permission to post (through the provision of Twitter accounts) instead of for allowing postings that have already occurred.” SER30. Moreover, the court reasoned, finding Twitter liable for failing to block accounts “would significantly affect Twitter’s monitoring and publication of third-party content.” SER32. Third, moving beyond Section 230, the court held that Plaintiffs’ accountprovision theory failed to satisfy the proximate cause element of their ATA claims. -9- Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 18 of 90 SER32-33. Courts in ATA cases, the court explained, had “rejected alleged causal connections that are too speculative or attenuated to raise a plausible inference of proximate causation.” SER33 n.4. That same defect doomed Plaintiffs’ First Amended Complaint, the court concluded, because the only alleged connection between Twitter and Abu Zaid’s attack—“that Abu Zaid’s brother told reporters that Abu Zaid had been very moved” by an execution that “ISIS publicized through Twitter”—was “tenuous at best.” SER32. Accordingly, the court ruled that “[e]ven under … [the] ‘substantial factor’ test” (Plaintiffs’ preferred but erroneous causation test), Plaintiffs had “not plausibly alleged the causal connection necessary” to support their claims. SER32-33. The court also rejected the First Amended Complaint’s other “attempt to evade section 230(c)(1)”—the theory that Twitter’s Direct Messaging capabilities are categorically unprotected by Section 230. SER33-34. The court explained that Congress had enacted the statute in response to a state court ruling that an online service provider could be liable for defamation based on third-party content posted on its service. In light of that history, the court reasoned, “the statute’s protections extend at least as far as the ‘treat[ment] [of] internet service providers as publishers … for the purposes of defamation.’” SER34. And because “the term ‘publication’” in defamation law includes any communication “‘to one other than the person defamed,’” “the private nature of Direct Messaging does not remove the - 10 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 19 of 90 transmission of such messages from the scope of publishing activity under section 230(c)(1).” Id. (emphasis added). Although dismissing all of Plaintiffs’ claims, the court granted Plaintiffs leave to amend again, SER35, which Plaintiffs took. AER51. C. The Second Amended Complaint The Second Amended Complaint, which is the operative pleading for purposes of this appeal, presented a new theory for trying to connect Abu Zaid to ISIS. It dropped the allegations that he had been inspired by ISIS’s execution of a Jordanian pilot that allegedly had been publicized via Twitter, and replaced them with an allegation that, according to Israeli military intelligence, he had once belonged to an ISIS “terror cell” when he was a university student. AER41 ¶ 81. Relatedly, the Second Amended Complaint repeated the allegation that ISIS had claimed responsibility for Abu Zaid’s attack, although it omitted reference to ISIS’s alleged depiction of Abu Zaid as a “lone wolf.” Compare SER16 (FAC), with AER41 ¶ 80 (SAC). Like the First Amended Complaint, the Second Amended Complaint still alleged “no connection between Abu Zaid and Twitter.” AER9 (Second Op.). Beyond these modified allegations about Abu Zaid and ISIS, the Second Amended Complaint largely repleaded the very same allegations as its predecessor, but in a different order and under different headings. It grouped allegations about - 11 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 20 of 90 accounts under new headings such as “Twitter Provided Accounts To ISIS” and “Twitter Provided Accounts To ISIS Knowingly And Recklessly.” AER29 (allcaps removed).1 And Plaintiffs attempted to collect all paragraphs about content allegedly created by ISIS members or sympathizers under another new heading, “Twitter Proximately Caused Plaintiffs’ Injuries.” See AER33-38 (all-caps removed).2 Plaintiffs fell short, however, in their apparent attempt to group and isolate all allegations regarding third-party content in this designated “proximate[] cause” section. As in their prior pleadings, such allegations also appeared in the part of the Second Amended Complaint that purported to describe measures Twitter allegedly could have employed to censor ISIS content. See AER31-33 ¶¶ 30, 32, 36-37, 40. D. The District Court’s Second Ruling Twitter again moved to dismiss and the court again granted Twitter’s motion, finding that “[d]espite … careful repleading, plaintiffs’ claims [had] not 1 Every paragraph about accounts in the Second Amended Complaint was copied or derived from a corresponding paragraph in the First Amended Complaint. Compare AER29-38 ¶¶ 9-13, 20, 24, 29-30, 32, 38-40, 43, 48, 56, 61, 66, 68-69, with SER1-13 ¶¶ 3-6, 20, 25, 33, 38, 43, 45, 49, 53, 59-60, 62, 68-70, 84. 2 For example, the Second Amended Complaint cited the same alleged use of Twitter by ISIS to distribute “its notorious promotional training video, ‘Flames of War,’” AER35 ¶ 49. But this time, the allegation came under the new proximate cause section, AER33, instead of under a section titled “ISIS Relies On Twitter To Terrorize,” SER5 (all-caps removed). - 12 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 21 of 90 changed in a meaningful way.” AER12. Because Plaintiffs continued to seek “to hold Twitter liable for allowing ISIS to use its network to spread propaganda and objectionable, destructive content,” the court again held that Section 230 barred their claims. Id. And the court again concluded that Plaintiffs had “not adequately alleged causation” and so had again failed to state a claim under the ATA. AER21. The court began by addressing Plaintiffs’ account-provision theory, once again finding it fraught with defects. First, the court explained, “providing accounts to ISIS” is just as much “publishing activity” and is no more “content-neutral” than “monitoring, reviewing, and editing content.” AER15-16. “Twitter could not possibly identify ISIS members,” the court observed, “without analyzing some speech, idea or content expressed by the would-be account holder.” AER16. And “specifically prohibit[ing] ISIS members and affiliates from acquiring accounts” would be “a policy that necessarily targets the content, ideas, and affiliations of particular account holders.” Id. Moreover, “even if a user never posts a tweet, ‘a user who opens an account necessarily puts content online,’” by “display[ing] a public user name such as @ISIS_Media_Hub, and a user photograph, such as a bearded man’s face” that “expresses a number of ideas.” Id. A theory aimed at blocking Twitter accounts, the court concluded, is “still based on Twitter’s alleged violation of a - 13 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 22 of 90 ‘duty … derive[d] from its status or conduct as a publisher’” and therefore barred by Section 230. AER17. Second, the court rejected Plaintiffs’ attempt to circumvent Section 230 simply by “restructur[ing]” their complaint to segregate content-based allegations in a section supposedly addressing only proximate cause. AER17-18. The court refused to “ignore that the majority of the [Second Amended Complaint] still focuses on ISIS’s objectionable use of Twitter and Twitter’s failure to prevent ISIS from using the site, not its failure to prevent ISIS from obtaining accounts.” AER18. And the court observed that, beyond the content-based allegations expressly tied to proximate cause, the Second Amended Complaint also “includes a number of allegations specifically faulting Twitter for failing to detect and prevent the dissemination of ISIS-related content through the Twitter platform.” AER18 (citing ¶¶ 30, 36). Because Plaintiffs’ claims “are inherently tied up with ISIS’s objectionable use of Twitter, not its mere acquisition of accounts,” the court held that they run afoul of Section 230. AER19. Third, the court held that Plaintiffs’ account-provision allegations still failed to satisfy the ATA’s proximate cause requirement. The court observed that “[t]here are no facts indicating that Abu Zaid’s attack was in any way impacted, helped by, or the result of ISIS’s presence on the social network.” AER21. To accept Plaintiffs’ “expansive proximate cause theory” would thus mean that - 14 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 23 of 90 “Twitter’s liability would theoretically persist indefinitely and attach to any and all future ISIS attacks.” Id. Because “[s]uch a standard cannot be and is not the law,” the court ruled that this failure to adequately allege proximate cause was an additional ground for dismissal. Id. Turning to Plaintiffs’ Direct Messaging theory, the court reaffirmed its earlier conclusion that Section 230 applies just as much to a service provider’s transmission of nonpublic messages, as to wholly public messages. AER22-24. The court rejected Plaintiffs’ contention that the defamation definition of “publisher” should apply only to defamation claims, explaining that this Court had already held that Section 230 “precludes courts from treating internet service providers as publishers not just for purposes of defamation law … but in general.” AER23. And “if the goal of [Section 230] is to promote unfettered and unregulated free speech on the internet, there is no reason that section 230(c)(1) should shield providers of private messaging services from defamation liability, but no other civil liability.” Id. Having “confirmed with [P]laintiffs’ counsel at the hearing [on Twitter’s second motion to dismiss] that [Plaintiffs] did not seek further amendment,” and having also “concluded that further amendment would be futile,” the court dismissed the action with prejudice. AER25-26. - 15 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 24 of 90 SUMMARY OF ARGUMENT Section 230 bars all Plaintiffs’ claims because they seek to impose liability on Twitter for publishing and failing to block or remove third-party content that allegedly caused Plaintiffs’ injuries. Plaintiffs dispute only one element of Twitter’s Section 230 defense—whether their claims treat Twitter as the “publisher” of third-party content. But both their attempts to negate that element fail. First, contrary to what Plaintiffs contend, providing accounts is publishing conduct and so Section 230 precludes liability based on Twitter’s accountprovision decisions. The First Circuit in Universal Commc’ns Sys., Inc. v. Lycos, Inc., 478 F.3d 413 (1st Cir. 2007), and the Fifth Circuit in Doe v. MySpace, Inc., 528 F.3d 413 (5th Cir. 2008), reached exactly that conclusion when confronting indistinguishable account-provision arguments. Those courts were correct. As the district court explained, a decision about whom to let speak on Twitter’s platform is fundamentally a decision about what third-party content to publish. AER15-17; SER28-32. Indeed, Plaintiffs’ Second Amended Complaint makes clear that Plaintiffs seek to impose liability not merely because ISIS and its sympathizers allegedly were able to open Twitter accounts but because of how they allegedly used those accounts to disseminate objectionable content. No amount of artful pleading can avoid the conclusion that Section 230 bars Plaintiffs’ claims. - 16 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 25 of 90 Second, Plaintiffs also are wrong that Twitter’s Direct Messaging tool is categorically excluded from Section 230. Section 230’s text does not distinguish between third-party content sent to a limited number of individuals and content made generally available on a provider’s platform. To read such a distinction into the statutory term “publisher” would disregard that Section 230 was enacted against a background of defamation law—which defines the “publishing” of defamatory content to include communications made to only one person. See Restatement (Second) of Torts § 577(1); Prosser & Keaton on Torts § 113, at 797 (5th ed. 1984). Plaintiffs’ claims independently fail because they have not pled facts sufficient to establish that they were injured “by reason of”—i.e., that their injuries were proximately caused by—any “act of international terrorism” committed by Twitter. As the Supreme Court has made clear in interpreting the same phrase in other statutes, the ATA’s “by reason of” language requires Plaintiffs to plead and prove that Twitter’s actions “led directly” to their injuries. Plaintiffs do not dispute that they cannot satisfy that standard. Nor can Plaintiffs satisfy even the more relaxed “substantial factor” causation test that they prefer. Layers of speculation and innumerable intervening actors and events separate Abu Zaid’s attack from Twitter’s supposed “act of international terrorism”—i.e., Twitter’s alleged failure to block ISIS and its - 17 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 26 of 90 sympathizers from opening and using Twitter accounts and sending Direct Messages. Accepting Plaintiffs’ unbounded theory of liability would expose every online platform to possible liability for terrorist violence anywhere in the world simply because the terrorist who committed the attack may have been loosely affiliated with some of the platform’s hundreds of millions of users. The ATA does not permit such expansive liability. STANDARD OF REVIEW This Court reviews de novo the district court’s dismissal for failure to state a claim under Fed. R. Civ. P. 12(b)(6). See Pride v. Correa, 719 F.3d 1130, 1133 (9th Cir. 2013). To survive a motion to dismiss, a plaintiff must plead “factual content that allows the court to draw the reasonable inference that the defendant is liable for the misconduct alleged.” Ashcroft v. Iqbal, 556 U.S. 662, 678 (2009). Section 230 requires dismissal where its application “is evident from the face of the complaint.” Klayman v. Zuckerberg, 753 F.3d 1354, 1357 (D.C. Cir. 2014); see also Kimzey, 836 F.3d at 1266 (affirming grant of motion to dismiss). Because Section 230 protects against not only “ultimate liability,” but also from “having to fight costly and protracted legal battles,” Roommates.com, 521 F.3d at 1175, the immunity attaches and should be resolved “at the earliest possible stage of the case,” Nemet Chevrolet, Ltd. v. Consumeraffairs.com, Inc., 591 F.3d 250, 255 (4th Cir. 2009). - 18 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 27 of 90 Given “‘the extreme nature of the charge of terrorism,’” “‘fairness requires” the court in an action under the ATA to give “‘extra-careful scrutiny’” to the plaintiff’s allegations. In re Terrorist Attacks on Sept. 11, 2001, 349 F. Supp. 2d 765, 831 (S.D.N.Y. 2005), aff’d, 714 F.3d 118 (2d Cir. 2013). ARGUMENT I. THE DISTRICT COURT CORRECTLY HELD THAT SECTION 230 REQUIRED DISMISSAL OF PLAINTIFFS’ CLAIMS A. Section 230 Broadly Immunizes Online Service Providers From Liability For Injuries Arising From Third-Party Content The broad scope of Section 230 is apparent on its face: “No provider or user of an interactive computer service shall be treated as the publisher or speaker of any information provided by another information content provider.” 47 U.S.C. § 230(c)(1). As this Court has held, this provision “immunizes providers of interactive computer services against liability arising from content created by third parties.” Roommates.com, 521 F.3d at 1162. This immunity extends to any lawsuit that “treat[s]” a provider as the “publisher” of third-party content by seeking to impose liability for performing or failing to perform a “‘traditional editorial function’” such as “deciding whether to publish or to withdraw from publication” any content created by users of its service. Barnes v. Yahoo!, Inc., 570 F.3d 1096, 1102 (9th Cir. 2009). - 19 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 28 of 90 Section 230’s broad grant of immunity serves two important purposes. First, because imposing civil liability on service providers for disseminating harmful third-party content would dramatically chill online expression, Congress enacted Section 230 “to encourage the unfettered and unregulated development of free speech on the Internet.” Batzel v. Smith, 333 F.3d 1018, 1027 (9th Cir. 2003); accord Barnes, 570 F.3d at 1099. Second, Congress sought to eliminate disincentives for service providers to self-police their platforms for unlawful and offensive material, out of fear that such action might itself impart them with knowledge that could subject them to liability. See Roommates.com, 521 F.3d at 1163 (Congress “spare[d] interactive computer services th[e] grim choice” between “taking responsibility for all messages and deleting no messages at all”); see also Barnes, 570 F.3d at 1099-1101; Batzel, 333 F.3d at 1028-1029; Zeran v. America Online, Inc., 129 F.3d 327, 331, 333 (4th Cir. 1997). Courts across the country have properly “treated § 230(c) immunity as quite robust.” Carafano v. Metrosplash.com, Inc., 339 F.3d 1119, 1123 (9th Cir. 2003). Section 230 has been applied to a broad swath of content, including material alleged to be discriminatory, Roommates.com, 521 F.3d at 1173-1174, false and defamatory speech, Zeran, 129 F.3d at 327, a vile and lawless advertisement posted by sex traffickers, Doe No. 1 v. Backpage.com, LLC, 817 F.3d 12, 21 (1st Cir. 2016), and an abhorrent and explicit incitement to violence by a terrorist - 20 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 29 of 90 group, Klayman, 753 F.3d at 1356-1357. And it has been held to bar a wide array of different claims and legal theories, including claims for intentional assault, Klayman, 753 F.3d at 1357-1360, negligent undertaking, Barnes, 570 F.3d at 11031106, wrongful death, Witkoff v. Topix, LLC, 2015 WL 5297912, at *7-8 (Cal. Ct. App. Sept. 10, 2015), aiding and abetting, Goddard v. Google, Inc., 2008 WL 5245490 (N.D. Cal. Dec. 17, 2008), intentional infliction of emotional distress, Jones v. Dirty World Entm’t Recordings, 755 F.3d 398, 406-407 (6th Cir. 2014), violation of anti-sex-trafficking laws, Backpage.com, 817 F.3d at 18-24, and, like here, provision of material support to terrorists, Cohen v. Facebook, Inc., 2017 WL 2192621 (E.D.N.Y. May 18, 2017). Just as in those cases, Section 230 bars Plaintiffs’ claims here. B. Plaintiffs Cannot Circumvent Section 230 Based On Their Account-Provision Theory Section 230 mandates dismissal when (1) the defendant is a “provider … of an interactive computer service” (2) that the plaintiff seeks to treat as the “publisher or speaker” of (3) content “provided by another information content provider,” and not the defendant. 47 U.S.C. § 230(c)(1); see also Barnes, 570 F.3d at 1100-1101. Plaintiffs do not dispute the first and last elements but instead limit - 21 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 30 of 90 their arguments to the middle element—whether their claims treat Twitter as the “publisher or speaker” of third-party content. AER13 (Second Op.). 3 The first theory Plaintiffs advance is their so-called account-provision theory. According to Plaintiffs, imposing liability on a service provider for providing accounts to particular users does not treat the provider as a publisher of third-party content. Br. 2, 15-18. And that, they assert, is all their claims seek to do—hold Twitter liable for permitting persons affiliated with ISIS to sign up for accounts, not for allowing such persons to use those accounts to create and transmit harmful content. Id. 2, 14, 19-20. Plaintiffs contend that the myriad references to third-party content in the Second Amended Complaint serve only to demonstrate causation—which, they argue, is not enough to make their claims “content-based.” Id. 19. Plaintiffs’ account-provision theory fails for three interrelated reasons. First, contrary to how Plaintiffs (inaccurately) describe their Second Amended Complaint, it still premises liability on third-party content, and not only in attempting to show causation. Second, Twitter’s decisions about whom to let open 3 Plaintiffs do not dispute the other two elements for good reason. Twitter’s platform is a paradigmatic “interactive computer service.” See 47 U.S.C. § 230(f)(2). And Plaintiffs repeatedly state that the objectionable material that allegedly caused their injuries was created and posted entirely by third parties— ISIS and its sympathizers (e.g., AER30-31, 36 ¶¶ 22, 26, 57)—which means that it is content “provided by another information content provider,” 47 U.S.C. § 230(c)(1) (emphasis added). - 22 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 31 of 90 an account are just as much protected publishing decisions as decisions to block or remove objectionable Tweets. Third, Plaintiffs’ strategy of attempting to segregate all content-based allegations in the causation section of the Second Amended Complaint constitutes exactly the sort of artful pleading that Section 230 forbids. 1. The Second Amended Complaint Purports To Hold Twitter Liable For Publishing Third-Party Content The threshold problem with Plaintiffs’ account-provision theory is that it does not accurately describe the Second Amended Complaint. Contrary to how Plaintiffs describe that complaint, they in fact deploy allegations expressly invoking third-party content in two separate parts of the complaint to support two points: in a section supposedly devoted to causation, see AER33-38, and in another section that describes steps Twitter allegedly failed to take to censor or block ISIS-related content, AER31-33 ¶¶ 30, 32, 35-37, 40; see also AER18. That latter section, for example, rebukes Twitter for allegedly “not proactively monitor[ing] [the] content” of “500 million tweets” sent by its users each day. AER32 ¶ 36 (emphasis added). It contends that Twitter ought to have searched for “terror content” by adapting a (content-based) system for tracking child pornography. AER32 ¶ 37 (emphasis added). It denounces Twitter for (allegedly) only recently suspending hundreds of thousands accounts “for promoting terrorism”—an obviously content-based criteria. AER33 ¶ 40. And it criticizes Twitter for allegedly allowing ISIS “to publicize its acts of terrorism” and - 23 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 32 of 90 for allegedly failing to “censor user content,” “shut down clear incitements to violence by terrorists,” and “take a more proactive approach to countering terrorist messages and recruitment online.” AER31-32 ¶¶ 30, 32, 35, 36 (emphases added). These alleged acts and omissions—which amount to allowing third-party content to be published or failing to block or censor it—are “traditional editorial functions” and therefore are “precisely the kind of activity for which Congress intended to grant absolution with the passage of [S]ection 230.” Roommates.com, 521 F.3d at 1171-1172, 1184; see also id. at 1171-1172 (provider cannot be held liable for “failing to detect and remove” unlawful content); Klayman, 753 F.3d at 1355, 1358-1359 (rejecting claims against Facebook for allegedly “refus[ing]” to take down “Intifada” pages that “called for Muslims to rise up and kill the Jewish people”). “[R]emoving content is something publishers do, and to impose liability on the basis of such conduct necessarily involves treating the liable party as a publisher of the content it failed to remove.” Barnes, 570 F.3d at 1103. For this reason alone, Section 230 bars Plaintiffs’ claims. 2. Decisions About Whom To Let Speak On A Provider’s Platform Are Publishing Decisions More fundamentally, Plaintiffs are simply wrong that they can evade Section 230 by casting their claims as concerning only the provision of accounts— as courts have concluded when confronting exactly this theory. In Universal Communications Systems, Inc. v. Lycos, Inc., for example, the First Circuit held - 24 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 33 of 90 that Section 230 barred claims allegedly based only on Lycos’s decision “not [to] prevent a single individual from registering under multiple screen names,” which allegedly “ma[de] it possible for individuals to spread misinformation more credibly.” 478 F.3d 413, 420 (1st Cir. 2007). The court explained that “Lycos’s decision not to reduce misinformation by changing its web site policies [regarding multiple accounts] was as much an editorial decision with respect to that misinformation as a decision not to delete a particular posting.” Id. at 422. And in Doe v. MySpace, Inc., the Fifth Circuit likewise rejected a plaintiff’s attempt to evade Section 230 based on an account-provision theory. 528 F.3d 413 (5th Cir. 2008). The plaintiff in MySpace, as here, asserted that she was “not complaining about any of the content that was exchanged” on the provider’s platform, but was instead complaining only about the platform’s failure to screen users when opening an account—in that case, by using “age verification software” to block minors from opening an account or to steer them into accounts with settings designed to protect children. Id. at 421. The Fifth Circuit rejected the plaintiff’s theory, finding that her “allegations [were] merely another way of claiming that MySpace was liable for publishing the communications and [the allegations] speak to MySpace’s role as a publisher of online third-party-generated content.” Id. at 419-420; see also Doe II v. MySpace, Inc., 96 Cal. Rptr. 3d 148, 156 (Ct. App. 2009) (dismissing analogous claims against MySpace, explaining - 25 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 34 of 90 that “[a]t its core, appellants want MySpace to regulate what appears on its Web site”); Cohen, 2017 WL 2192621, at *12 (explaining that any “attempt to draw a narrow distinction between policing accounts and policing content must ultimately be rejected”). These courts—like the court below—were correct to reject an accountprovision theory because imposing liability for permitting particular individuals to open or maintain an account would penalize a protected publishing decision: “the decision to permit third parties to post content.” SER30 (First Op.). “The broad construction accorded to [S]ection 230 as a whole has resulted in a capacious conception of what it means to treat a website operator as the publisher … of information provided by a third party.” Backpage.com, 817 F.3d at 19. “[W]hat matters is not the name of the cause the action” or the “label[]” that a plaintiff puts on her “theory” but “whether the cause of action inherently requires the court to treat the defendant as the publisher or speaker of content provided by another.” Barnes, 570 F.3d at 1101-1103. If “the duty that the plaintiff alleges the defendant violated derives from the defendant’s status or conduct as a publisher or speaker,” that is the end of the analysis, and Section 230 “precludes liability.” Id. at 1102. No “intellectual gymnastics” are required to see that providing accounts is publishing conduct. Barnes, 570 F.3d at 1102. This Court has recognized repeatedly that publishing conduct includes “any activity that can be boiled down - 26 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 35 of 90 to deciding whether to exclude material that third parties seek to post online.” Roommates.com, 521 F.3d at 1170-1171; see also Barnes, 570 F.3d at 1102. A decision about who may have an account is simply one of many ways a service provider controls what third-party content appears on its platform. Twitter can restrict users from posting Tweets by denying them an account at the outset, by removing particular Tweets, or by shutting down accounts because of the Tweets they have posted. When it comes to social media accounts, the difference between “handing someone a tool” and “supervising the use of that tool” (Appellant Br. 16) is merely a matter of “timing” and “scope,” AER16-17 (Second Op.). As the district court concluded, whether blocking an account or removing a Tweet, the provider is deciding whether “to permit third parties to post content,” SER30, and is therefore engaging in “publishing activity,” AER16-17. The district court could hardly have concluded otherwise given this Court’s oft-repeated admonition that “[a] distinction between removing an item once it has appeared on the Internet and screening before publication cannot fly.” Batzel, 333 F.3d at 1032; see also Barnes, 570 F.3d at 1102 n.8 (“[i]t is immaterial” whether the “decision comes in the form of deciding what to publish in the first place or what to remove among the published material”). “The scope of the immunity cannot turn on whether the publisher approaches the selection process as one of inclusion or removal, as the difference is one of method or degree, not substance.” - 27 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 36 of 90 Batzel, 333 F.3d at 1032. At any stage, decisions about what may be posted are publishing decisions, and therefore within the scope of Section 230 immunity. Plaintiffs are simply wrong, moreover, that barring users from signing up for accounts based on “status” has “nothing whatsoever” to do with the content of Tweets that appears on Twitter’s platform. Br. 18. As the district court held, there is nothing “content-neutral” about “specifically prohibit[ing] ISIS members and affiliates from acquiring accounts—a policy that necessarily targets the content, ideas, and affiliations of particular accountholders.” AER16. A “status-based” decision about whom to publish, in other words, is ultimately a decision about what to publish. Consider, as a hypothetical, a left-leaning newspaper’s decision to hire an author known for holding conservative views to write a weekly column in its opinion pages. Such a decision would be a publishing decision that may have been motivated by, for example, the newspaper’s goal to diversify the ideological viewpoints it publishes. See, e.g., Blumenthal v. Drudge, 992 F. Supp. 44, 47, 50 (D.D.C. 1998) (holding that Section 230 immunized AOL for making Matt Drudge’s column available on its platform). Conversely, “‘restrictions based on the identity of the speaker are all too often simply a means to control content.’” Reed v. Town of Gilbert, 135 S. Ct. 2218, 2230 (2015). A service provider’s “choices as to who may use its platform are inherently bound up in its decisions as - 28 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 37 of 90 to what may be said on its platform.” Cohen, 2017 WL 2192621, at *12 (dismissing claims against Facebook for allegedly providing “material support” to Hamas by allegedly allowing Hamas to use Facebook services) (emphases added). Plaintiffs’ own pleadings illustrate the tight interplay between whom Twitter allows to open an account and what third-party content is published. As the district court explained, the Second Amended Complaint is “riddled with detailed descriptions of ISIS-related messages, images, and videos disseminated through Twitter and the harms allegedly caused by the dissemination of that content.” AER18. These voluminous allegations about third-party content “cannot [be] ignore[d]” and leave no doubt that Plaintiffs’ “claims are inherently tied up with ISIS’s objectionable use of Twitter, not its mere acquisition of accounts.” AER1819 (Second Op.). If ISIS had opened accounts but never used them to recruit followers, raise funds, and spread propaganda (or used them only to share knitting lessons or the like), then Plaintiffs could not possibly allege, as they do, that “Twitter enabled ISIS to acquire the resources needed to carry out numerous terrorist attacks,” including the attack that killed Mr. Fields and Mr. Creach. AER28 ¶ 1. Plainly, the “rationale behind” Plaintiffs’ claims is “that Twitter should not have provided accounts to ISIS because ISIS would and has used those accounts to post objectionable content.” AER19 (Second Op.). - 29 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 38 of 90 Like those of other social media platforms, Twitter’s policies regarding who may open and use an account are closely tied to its “choices about what content can appear on the website,” Backpage.com, 817 F.3d at 21, as the district court recognized, AER17. Twitter has chosen to create a platform that “allows for the freedom of expression for hundreds of millions of people around the world,” AER32 ¶ 35, rather than a heavily curated site that features only a handful of select voices. Undertaking the onerous task of pre-screening every user would severely limit Twitter’s ability to permit such a large number of voices with a wide array of perspectives to speak on Twitter’s platform. That is why Twitter presumptively allows everyone who has access to the Internet and who agrees to its Terms of Service (including its Rules, which ban content promoting terrorism) to open an account and typically only “shut[s] down” accounts when Twitter learns that its Rules have been violated. AER32 ¶ 39. Thus, Twitter’s alleged “hands-off policy” regarding account-provision itself “‘reflect[s] choices about what [third-party] content can appear on [Twitter] and in what form.’” AER17 (Second Op.) (quoting Backpage.com, 817 F.3d at 21). While Plaintiffs assert (at 19) that they “do not claim that Twitter should have built it[s] website differently,” it is difficult to see how Twitter could have prevented every potential ISIS member or sympathizer from even opening an account without - 30 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 39 of 90 adopting policies and strategies that would transform the basic nature of the platform itself. Further, as the district court recognized, third-party users of Twitter’s services put content online even before publishing a single Tweet, simply by opening an account. See Barnes, 570 F.3d at 1103 (equating removing profiles with removing content). An online profile has content the moment a “user actively creates it.” Carafano, 339 F.3d at 1124; see also id. at 1121 (a profile “evoke[s] the subject’s personality”); MySpace, 528 F.3d at 415 (“[A]n online profile … serves as a medium for personal expression.”). On some platforms, user profiles are the only third-party content displayed publicly. See Carafano, 339 F.3d at 1121 (Matchmaker members can view profiles and then contact other members “via electronic mail sent through the Matchmaker server”). In Twitter’s case, as soon as an account “spring[s] … up,” AER32-33 ¶ 39, it displays a user name—such as “@TurMedia334,” id.—and a photograph—such as a “photograph of a bearded man’s face,” id. As the district court correctly observed, a user name like “ISIS_Media_Hub,” AER29 ¶ 12, “on its own expresses a number of ideas: ‘I am affiliated with ISIS’; ‘I am a media source’; and ‘Follow me for information and publicity about ISIS’s activities.’” AER16 (Second Op.). Thus, even focusing only on what happens when an individual opens an account, any decision to allow or block the account necessarily “involves - 31 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 40 of 90 . . . deciding whether to publish or to withdraw from publication third-party content,” Barnes, 570 F.3d at 1102, and so is a publishing decision protected by Section 230. Moreover, as explained above (at 23-24), the monitoring, tracking and suspension strategies Plaintiffs criticize Twitter for allegedly not deploying would necessarily have relied on user-provided content to determine whether or not any particular user is affiliated with ISIS. For the first time on appeal, Plaintiffs characterize (at 17-18) their position as being that Twitter had a duty to deny accounts only to persons it knew from external sources (such as media reports or government communications) were named on a government watch list. That characterization cannot be reconciled with Plaintiffs’ allegations faulting Twitter for allegedly not “proactively monitor[ing] content,” deploying a tracking system for “terror content,” or suspending accounts “for promoting terrorism.” AER32-33 ¶¶ 36-37, 40; see also supra pp. 23-24. And as the district court explained, Plaintiffs’ theory of harm depends on the claim that Twitter should have achieved “a blanket ban on pro-ISIS content.” AER16 (emphasis added); see also AER33 ¶ 40 (portraying removal of 235,000 accounts over a few months as inadequate). “Twitter could not possibly [have] identif[ied] [enough] ISIS members” to achieve such a “blanket ban” “without analyzing some speech, idea or content expressed by - 32 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 41 of 90 the would-be account holder: i.e. ‘I am associated with ISIS.’” AER16 (Second Op.). As in Lycos and MySpace, Plaintiffs’ claims “at their core,” AER17 (Second Op.), seek to hold Twitter liable for publishing—and failing to prevent—thirdparty communications. Section 230 therefore bars their claims. 3. Plaintiffs Cannot Circumvent Section 230 By Confining Allegations Expressly Mentioning Third-Party Content To The Causation Section Of Their Complaint While Plaintiffs concede (at 19-22) that their claims depend on harmful third-party content having been published on Twitter’s platform, they nonetheless argue (at 2, 19-20), based on their reading of this Court’s decisions in Internet Brands and Barnes, that their claims are not barred by Section 230 because the Second Amended Complaint supposedly references content only to demonstrate causation. As noted earlier, that is not an accurate description of the Second Amended Complaint, which devotes an entire section to describing the steps Twitter allegedly failed to take to eliminate ISIS content from its platform. See supra pp. 23-24. Section 230, moreover, precludes exactly this sort of gamesmanship. This Court has consistently rejected efforts “to plead around [Section 230] to advance the same basic argument that the statute plainly bars.” Kimzey, 836 F.3d at 1266. As this Court explained in Barnes, “a plaintiff cannot sue someone for - 33 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 42 of 90 publishing third-party content simply by changing the name of the theory” or by placing a different label on “an action that is quintessentially that of a publisher.” 570 F.2d at 1102-1103. Likewise, the Fifth Circuit in MySpace condemned such tactics as “disingenuous” and “artful pleading.” 528 F.3d at 419. Yet that is exactly what Plaintiffs have done in attempting to artificially divorce the opening of an account from how that account is used. Plaintiffs’ original Complaint and First Amended Complaint both forthrightly announced their intent to hold Twitter liable for publishing third-party content, alleging in the very first paragraph that Twitter “has been instrumental to the rise of ISIS” by “knowingly permit[ting]” the terrorist group to use Twitter’s platform “for spreading extremist propaganda, raising funds and attracting new recruits.” Dkt. 1 at 1; SER2 ¶ 1 (emphasis added). After the district court dismissed the First Amended Complaint, SER24-35, Plaintiffs then relabeled and reorganized the same allegations in the now-operative Second Amended Complaint, grouping allegations regarding Twitter’s alleged provision of accounts to ISIS under sections titled “Twitter Provided Accounts To ISIS” and “Twitter Provided Accounts To ISIS Knowingly And Recklessly,” and segregating many of the allegations that expressly describe the content created using those accounts under the header: “Twitter Proximately Caused Plaintiffs’ Injuries.” AER29-38 (all-caps removed). Section 230 would be a flimsy shield if it could be so easily pierced by such - 34 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 43 of 90 transparent artful pleading. As the district court made clear: “[N]o amount of careful pleading can change the fact that, in substance, plaintiffs aim to hold Twitter liable as a publisher or speaker of ISIS’s hateful rhetoric.” See AER8-9. Moreover, Plaintiffs are wrong that Internet Brands and Barnes somehow authorize them to avoid Section 230 by quarantining allegations referencing content under a proximate cause heading. There are two basic problems with Plaintiffs’ strained reading of those decisions. First, this and other Courts have repeatedly held, both before and after Internet Brands and Barnes, that the “basic argument that [Section 230] plainly bars” is that an online provider, like Twitter, is liable for “publish[ing] usergenerated speech that was harmful to [the plaintiff],” Kimzey, 836 F.3d at 1266— an argument, in other words, that the provider caused harm to befall the plaintiff by publishing user-generated content. That was the recent holding of the First Circuit in Backpage.com: Section 230 bars claims where “there would be no harm to [plaintiffs] but for the content of the postings.” 817 F.3d at 19-20 (emphasis added). It also was the holding of the Fifth Circuit in MySpace: Parties “harmed by a Web site’s publication of user-generated content” cannot sue “the interactive computer service that enabled [the third-party user] to publish the content online.” 528 F.3d at 419-420. And it was the holding of the Fourth Circuit’s seminal decision in Zeran: In enacting Section 230, “Congress made a policy choice … not - 35 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 44 of 90 to deter harmful online speech through the separate route of imposing tort liability on companies that serve as intermediaries for other parties’ potentially injurious messages.” 129 F.3d at 330-331; see also Carafano, 339 F.3d at 1123 (quoting Zeran). Nothing in Barnes or Internet Brands suggests that this Court meant to contradict other Ninth Circuit precedent and deviate from this settled judicial consensus by permitting plaintiffs to rely heavily on third-party content so long as they do so only to try to satisfy a causation element of their legal claim. Second, as the district court correctly recognized, Plaintiffs’ argument misapprehends how Barnes and Internet Brands describe the Section 230 inquiry. SER31. According to Barnes, what matters is “whether the duty that the plaintiff alleges the defendant violated derives from the defendant’s status or conduct as a ‘publisher or speaker’” of third-party content. See Barnes, 570 F.3d at 1102. The dividing line, then, is between duties derived from publishing third-party content and duties derived from other types of conduct, such as a duty derived from agreeing to a contract, see Barnes, 570 F.3d at 1107, or a provider’s duty to publish its own content, see Doe No. 14 v. Internet Brands, Inc., 824 F.3d 846, 852 (9th Cir. 2016) (provider failed to “generate its own warning” (emphasis added)). See also AER20-21 (Second Op.) (distinguishing Internet Brands and Barnes on the grounds that, unlike in those cases, Plaintiffs in this case “seek to hold [an online service provider] liable for allowing [third parties] to post . . . objectionable - 36 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 45 of 90 content on its site”). As explained above, acts and omissions that determine who may speak on a platform are “publishing conduct.” See Barnes, 570 F.3d at 1103. And so, a duty derived from such conduct—such as a duty to block certain categories of speakers—necessarily falls on the immunity side of the line. Moreover, there are numerous other differences between this case and the two cases on which Plaintiffs rely. The Court in Internet Brands repeatedly emphasized that the information allegedly triggering the duty to warn “was obtained by Internet Brands from an outside source, not from monitoring postings on the Model Mayhem Website,” and that the persons who lured and raped Doe “did not post on the website.” 824 F.3d at 849. And the Court stressed that imposing a duty to warn on Internet Brands would not “require [it] to remove any user content or otherwise affect how it publishes or monitors such content.” Id. at 851 (emphasis added). The Court discussed “but-for” causation solely to make the unexceptional point that Section 230 “does not provide a general immunity” to platforms simply because they are in the “business” of “publishing user content.” Id. at 853. Here, by contrast, Plaintiffs’ entire case—including their theory of causation—depends on harmful user-generated content and Twitter’s alleged publishing conduct with respect to that content. And quite unlike the claims in Internet Brands, Plaintiffs’ claims here, if successful, “would significantly affect - 37 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 46 of 90 Twitter’s monitoring and publication of third-party content.” AER20 (Second Op.). Other distinctions abound with respect to Barnes. As already explained, the promissory estoppel claim at issue in that case—unlike the claims here—“did not ‘seek to hold Yahoo liable as a publisher or speaker of third-party content, but rather as the counterparty to a contract, as a promisor who has breached.’” AER19 (Second Op.) (quoting Barnes, 570 F.3d at 1107). But even beyond that distinction, the promise Yahoo made in Barnes “also signifie[d] the waiver of [any Section 230] defense[].” Barnes, 570 F.3d at 1108. Twitter, by contrast, has done nothing to waive its Section 230 defenses. For all these reasons, neither Internet Brands nor Barnes can rescue Plaintiffs’ claims. C. The Transmission Of Direct Messages Is Publishing Activity And Subject To Section 230 Immunity The district court likewise was correct to reject Plaintiffs’ other theory for “evad[ing] section 230(c)(1)”—the theory that Direct Messaging is categorically unprotected by Section 230. AER22. Abandoning all pretense of not seeking to hold Twitter liable for disseminating third-party content, Plaintiffs assert a theory of liability based on Twitter’s “[g]iving ISIS the capability to send and receive Direct Messages.” Br. 7; see also AER15-16 (Second Op.).4 Section 230 plainly 4 In the cosmetic reorganization of their allegations from their prior pleadings into the Second Amended Complaint, all allegations regarding Direct - 38 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 47 of 90 forecloses this attempt to hold Twitter liable for publishing—or failing to block— Direct Messages created and sent by its users. The text of Section 230(c)(1) makes no distinction between third-party content sent to a limited number of individuals and content made generally available on a provider’s platform. And to read such a distinction into the statute’s use of the term “publisher,” as Plaintiffs urge (at 27-28), would wrench the statute from the context in which it was enacted. As Plaintiffs acknowledge, Congress established Section 230’s protections “in reaction to the decision in Stratton Oakmont, Inc. v. Prodigy Services Co., 1995 WL 323710 (N.Y. Sup. Ct. May 24, 1995),” which held that “an internet service provider could be liable for defamation.” Br. 29. In light of this background, both the Fourth Circuit and this Court have recognized that “[t]he term[] ‘publisher’ [in Section 230]. . . derive[s] [its] legal significance from the context of defamation law.” Zeran, 129 F.3d at 332; accord Barnes, 570 F.3d at 1103-1104 (relying on the “reach” of defamation law to “confirm” the Court’s construction of Section 230). “[T]he statute’s protections,” as the district court explained, must therefore “extend at least as far Messages wound up in the causation section of the Second Amended Complaint. See AER33-34. Yet plaintiffs conceded both below and in their brief in this Court that their Direct Messaging theory depends on third-party content for more than merely attempting to meet the causation element of their claims. See Br. 28, 31; Dkt. 52 at 7; AER22-23 (“In support of their Direct Messaging theory, plaintiffs abandon all pretense of a content-less basis for liability.”). - 39 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 48 of 90 as the ‘treat[ment] [of] internet service providers as publishers … for purposes of defamation.’” SER34 (quoting Barnes, 570 F.3d at 1104 (emphasis added)). Contrary to what Plaintiffs contend (at 26-29), reading “publisher” in light of the term’s meaning in the defamation context—rather than selectively quoting from dictionaries—is the only way to give effect to the plain meaning of the statute. “It is a ‘cardinal rule of statutory construction’ that, when Congress employs a term of art, ‘it presumably knows and adopts the cluster of ideas that were attached to each borrowed word in the body of learning from which it was taken.’” FAA v. Cooper, 566 U.S. 284, 291-292 (2012). Congress has done exactly that here—and to import a limit on the term “publisher” from a source other than defamation law would in no way give effect to the statute’s plain meaning, as Congress understood it. Incorporating the term’s meaning in defamation law, “publisher” in Section 230 includes both one-to-many and one-to-one (or one-to-a-few) communications. AER23-24; SER34-35. The term “publication” in defamation law “simply means ‘communication intentionally or by a negligent act to one other than the person defamed.’” Barnes, 570 F.3d at 1104 (quoting Restatement (Second) of Torts § 577(1) (emphasis added)). “There may,” in other words, “be publication to any third person.” Prosser & Keaton on Torts § 113, at 798 (5th ed. 1984). As the district court therefore concluded, “the private nature of Direct Messaging does not - 40 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 49 of 90 remove the transmission of such messages from the scope of publishing activity under section 230(c)(1).” SER34. To be sure, Congress did not limit Section 230 to the defamation context, see Roommates.com, 521 F.3d 1157; Backpage.com, 817 F.3d 12, and courts accordingly have applied the provision to bar all manner of other types of common law and statutory claims, see supra pp. 20-21. Plainly, however, the term “publisher” also retains its original common law meaning: Section 230 “precludes courts from treating internet service providers as publishers” as that term is understood “for purposes of defamation,” Barnes 570 F.3d at 1104—that is, as one who communicates to “any third person,” Prosser & Keaton on Torts § 113, at 798 (emphasis added). As the district court correctly noted, where courts have considered claims against online service providers premised on one-to-one or one-to-a-few e-mail communications, they have not hesitated to reject them under Section 230. AER24. For example, in Delfino v. Agilent Technologies, a California Court of Appeal held that Section 230 barred liability for an interactive computer service’s transmission of threatening emails to a single individual. 52 Cal. Rptr. 3d 376, 381, 390-391 (Ct. App. 2006). And in Phan v. Pham, another California appellate court likewise held that Section 230 bars liability for forwarding a “defamatory email” to “at least one” recipient absent “material contribution.” 105 Cal. Rptr. 3d - 41 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 50 of 90 791, 792, 795 (Ct. App. 2010). Numerous other courts have applied Section 230 in a similar context.5 Although Plaintiffs object (at 31) that none of these cases explicitly addressed whether Section 230 applies to private communications, Plaintiffs cite not a single decision—and Twitter is aware of none—that has excluded private messages from Section 230’s protection. The entirely one-sided weight of this authority speaks volumes. Lest there be any doubt, the logical consequences of Plaintiffs’ theory evidence its implausibility. If Plaintiffs were right that Section 230 does not bar lawsuits premised on the contents of private online communications, then any networking website with a messaging tool—indeed, any email provider—could potentially be held liable for any harmful message sent through that component of its platform. Google could be sued for a defamatory statement sent via Gmail, Facebook for a threatening note transmitted through WhatsApp, or LinkedIn for a discriminatory job description delivered through its messaging tool. In today’s vast realm of online communication, interactive service providers potentially would be subject to endless lawsuits and staggering liability. Plaintiffs’ 5 Directory Assistants, Inc. v. Supermedia, LLC, 884 F. Supp. 2d 446, 452 (E.D. Va. 2012) (Section 230 bars liability for sending allegedly unlawful e-mails); Tanisha Sys. v. Chandra, 2015 WL 10550967, at *7 n.6, *8-9 (N.D. Ga. Dec. 4, 2015) (same); Peters v. LifeLock Inc., 2014 WL 12544495, at *3 (D. Ariz. Sept. 19, 2014) (same); Jane Doe One v. Oliver, 755 A.2d 1000, 1003-1004 (Conn. Sup. Ct. 2000) (AOL immune under Section 230 for emails sent through its service). - 42 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 51 of 90 construction of Section 230, in other words, would not merely chill online expression in contravention of Congress’s clear intent. See 47 U.S.C. § 230(b)(2); see also supra Part I.A; infra Part I.D. It would threaten providers of email service and other online-messaging tools with crippling liability. D. Allowing This Case To Proceed Would Give Rise To Precisely The Harms Section 230 Was Enacted To Prevent Plaintiffs’ contention (at 23) that reversing the district court’s decision would further the policy goals of Section 230 is completely backwards. Plaintiffs’ theories of liability, if permitted to proceed, would eviscerate the law’s protections and engender precisely the problems Congress sought to avoid. Under Plaintiffs’ account-provision theory, immunity would evaporate the moment a service provider permitted a terrorist, or defamer, or fraudster to create an account, leaving service providers little choice but to exhaustively evaluate every would-be user before allowing him to sign up for service. Indeed, many claims this and other courts have previously held to be barred by Section 230 could be refashioned as “account-provision” claims and, under Plaintiffs’ reasoning, subject providers to burdensome litigation and possible liability. For example, rather than suing based on the fabricated and defamatory content of an online dating profile, the plaintiff in Carafano could just as easily have brought a negligence claim based on Matchmaker’s alleged failure to prohibit persons from signing up for accounts using identities other than their own. 339 F.3d 1119. And, - 43 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 52 of 90 as discussed above, the First Circuit’s decision in Lycos and the Fifth Circuit’s decision in MySpace would have come out differently had those courts thought that a plaintiff could avoid Section 230 by purporting to recast his or her claims in terms of negligent provision of an account. Plaintiffs’ alternative theory—that service providers may be held liable for harms arising from any third-party content transmitted “privately” (Br. 6)—fares no better. Because such speech is virtually impossible for a platform-provider to police, service providers might well choose to protect themselves from liability by altogether ceasing to offer private messaging applications. Plaintiffs’ theory could do more than just chill private online speech, it could eliminate it altogether. Contrary to Plaintiffs’ assertion (at 23), courts have repeatedly recognized that application of the material support statutes very much can “implicate [the] free speech concerns” that animate Section 230. For example, although the Supreme Court in Holder v. Humanitarian Law Project, 561 U.S. 1 (2010) (“HLP”) upheld 18 U.S.C. § 2339B under the First Amendment “as applied to the particular activities” the plaintiffs wished to pursue, id. at 8, it did so only after applying strict scrutiny in light of the important speech interests at stake, id. at 27-28 (rejecting government’s request for intermediate scrutiny because “§ 2339B regulates speech on the basis of its content”). And this Court, noting these interests and the manner in which the Supreme Court “carefully circumscribed its - 44 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 53 of 90 analysis in HLP,” has “hesit[ated] to apply that decision to facts far beyond those at issue in that case.” Al Haramain Islamic Found. v. United States Dep’t of Treasury, 686 F.3d 965, 1001 (9th Cir. 2011). Indeed, Congress issued a similar warning in Section 2339B itself, directing that it be carefully construed to safeguard “the exercise of rights guaranteed under the First Amendment to the Constitution of the United States.” 18 U.S.C. § 2339B(i). Plaintiffs would have this Court ignore these many admonitions. That Plaintiffs’ theory would impose liability only if a service provider knowingly allowed a terrorist to sign up for service (Appellant Br. 25) is no answer. For one thing, because what a service provider knew about a particular account applicant is a question of fact, service providers would inevitably be denied the benefit of prompt dismissals of suits that Section 230 was designed to provide. See supra p. 18. For another, imposing liability based on a service provider’s knowledge—whether with respect to a particular user or user content— would undermine Congress’s second key objective: eliminating disincentives for service providers to self-police their platforms. See Batzel, 333 F.3d at 1028. Notice-based liability would “motivate providers to insulate themselves from receiving complaints” and “discourage active monitoring of Internet postings.” Barrett v. Rosenthal, 146 P.3d 510, 525 (Cal. 2006); accord Zeran, 129 F.3d at 330; Lycos, 478 F.3d at 420. Indeed, a rule that would expose Twitter to potential - 45 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 54 of 90 liability only if it reviewed third-party content (for purposes of making removal or account suspension decisions) would create a serious disincentive for Twitter to take voluntary proactive steps to enforce its prohibition against “threats of violence . . . including threatening or promoting terrorism.” AER33 ¶ 40; see also AER10-11. Section 230’s broad immunity thus has played a critical role in enabling Twitter to suspend 235,000 accounts for promoting terrorism between February and August 2016 alone. AER33 ¶ 40. Plaintiffs further err in suggesting (at 19) that there is a conflict here between the ATA and Section 230. Even assuming that Twitter’s opening of its platform for free speech to virtually all comers could somehow give rise to a cause of action under the ATA’s civil remedy provision—and it cannot, see infra pp. 4760—Congress enacted Section 230 for the very purpose of barring a cause of action where one might otherwise lie. Whether that cause of action is predicated on the common law, supra p. 7 (Congress enacted Section 230 in response to defamation case), a local ordinance, 47 U.S.C. § 230(e)(3) (prohibiting any cause of action under “any State or local law that is inconsistent with this section”), or a federal statute establishing civil causes of action, e.g., Roommates.com, 521 F.3d 1157 (dismissing in part complaint alleging violations of federal Fair Housing Act on Section 230 grounds); Backpage.com, 817 F.3d 12 (TVPRA), Section 230 - 46 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 55 of 90 mandates dismissal where, as here, a lawsuit seeks to hold a service provider liable for harm allegedly arising from third-party content. II. THE DISTRICT COURT CORRECTLY DISMISSED PLAINTIFFS’ CLAIMS ON THE ALTERNATIVE GROUND THAT THEIR ALLEGATIONS FAIL TO SATISFY THE ATA’S PROXIMATE CAUSATION ELEMENT The district court also dismissed Plaintiffs’ Second Amended Complaint on a second ground: it fails to plead facts plausibly establishing that Twitter proximately caused the deaths of Mr. Fields and Mr. Creach, as required under the ATA’s civil remedy provision. AER21-22; see also SER 32-33 (same). That conclusion was correct as well, and should be affirmed for two independent reasons. First, to satisfy the ATA’s proximate causation requirement, a plaintiff must plead and prove that an act of international terrorism committed by the defendant “led directly” to the plaintiff’s injuries. Plaintiffs have never even attempted to argue that their allegations could satisfy that standard, effectively conceding that they do not. Second, while Plaintiffs press an alternative standard for proximate causation (at 32)—under which they need only plead and prove that Twitter’s “acts were a substantial factor in the sequence of responsible causation” and that their injuries were “reasonably foreseeable or anticipated as a natural consequence”— their allegations fail even that test. As the district court explained, “[e]ven under [P]laintiffs’ proposed ‘substantial factor’ test,” Plaintiffs’ causation theory is too - 47 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 56 of 90 attenuated to “support a plausible inference of proximate causation between Twitter’s provision of accounts to ISIS and the deaths of [Mr.] Fields and [Mr.] Creach.” AER21. A. The ATA Requires Plaintiffs To Plead Facts That Plausibly Establish A Direct Relationship Between An Act Of International Terrorism Committed By Twitter And Plaintiffs’ Injuries The plain text of the ATA’s civil remedy provision dictates the causation standard that Plaintiffs must satisfy. That provision creates a private right of action for “[a]ny national of the United States injured … by reason of an act of international terrorism.” 18 U.S.C. § 2333(a) (emphasis added). The “by reason of” language in § 2333(a) restricts liability “to situations where plaintiffs plausibly allege that defendant’s actions proximately caused their injuries.” In re Terrorist Attacks on Sept. 11, 2001 (“Al Rajhi”), 714 F.3d 118, 125 (2d Cir. 2013); see also Rothstein v. UBS AG, 708 F.3d 82, 95 (2d Cir. 2013). Congress borrowed this proximate cause language from the remedies provisions of the Sherman Act, 26 Stat. 209 (1890), the Clayton Act, 15 U.S.C. § 15(a), and RICO, 18 U.S.C. § 1964(c), all of which provide that a person injured “by reason of” specified conduct “may sue therefor.” Because Congress “used the same words” in the ATA as it used in those previously enacted statutes, this Court should “assume it intended them to have the same meaning.” Holmes v. Securities Inv’r Prot. Corp., 503 U.S. 258, 268 (1992). The Supreme Court followed this - 48 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 57 of 90 exact approach when it construed the “by reason of” language in RICO—assigning the phrase the same “meaning that courts had already given” those words in the previously enacted Clayton Act. Id. at 268. This Court must do the same when interpreting “by reason of” in the ATA. The Supreme Court has explained repeatedly that a plaintiff can satisfy this statutory “by reason of” standard only if she can show “some direct relation between the injury asserted and injurious conduct alleged.” Holmes, 503 U.S. at 268 (proximate cause requirement under RICO) (emphasis added); see also id. at 269 (“directness of relationship is [a] … requirement of Clayton Act causation”). The “central question” in evaluating proximate cause under a statute employing the “by reason of” language, in other words, is “whether the alleged violation led directly to the plaintiff’s injuries.” Anza v. Ideal Steel Supply Corp., 547 U.S. 451, 456, 461 (2006) (proximate cause requirement under RICO) (emphasis added). A “theory of causation” that “stretche[s] the causal chain” linking defendant’s allegedly unlawful conduct to plaintiff’s injury “well beyond the first step” in that chain cannot satisfy this “direct relationship requirement.” Hemi Grp., LLC v. City of N.Y., 559 U.S. 1, 10-11 (2010) (same). The Supreme Court, moreover, has “explicitly rejected” Plaintiffs’ preferred formulation for proximate cause in statutes that impose a “by reason of” requirement. Couch v. Cate, 379 F. App’x 560, 565-566 (9th Cir. 2010) - 49 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 58 of 90 (summarizing holding of Hemi Grp., 559 U.S. at 12). In Hemi Group, the Court construed RICO’s “by reason of” requirement to “definitively foreclose[]” liability for “consequences that are only foreseeable without some direct relationship” to the defendant’s allegedly unlawful conduct. See Couch, 379 F. App’x at 565-566; see also Mattel v. MGA Entm’t, Inc., 782 F. Supp. 2d 911, 1024 (C.D. Cal. 2011) (similarly describing Hemi Group’s holding). The Supreme Court likewise rejected a mere “foreseeability” standard for proximate cause under the Fair Housing Act (“FHA”) in its recent decision in Bank of America Corp. v. City of Miami, 137 S. Ct. 1296, 1305-1306 (2017). Although the FHA does not contain “by reason of” language, it creates a cause of action that—just like the ATA—“is akin to a ‘tort action.’” Id. at 1305; see also Boim v. Quranic Literacy Inst. & Holy Land Found. for Relief & Dev., 291 F.3d 1000, 1010 (7th Cir. 2002) (ATA incorporates “general common law tort principles”). The Supreme Court explained that it has “repeatedly applied directness principles to statutes with ‘common-law foundations.’” Bank of Am., 137 S. Ct. at 1306 (emphasis added). The ATA’s common-law roots thus provide yet another reason why the ATA must be interpreted to impose a “direct relationship” standard for proximate cause. Plaintiffs’ interpretation of “by reason of” is also inconsistent with the careful limits Congress placed on secondary liability under subsection 2333(d) of - 50 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 59 of 90 the ATA. To state a claim for aiding and abetting or conspiracy under that provision, it is not enough for a plaintiff to show that the defendant aided or conspired with the designated foreign terrorist organization that “committed, planned, or authorized” the terrorist attack from which the plaintiff’s injury arose. 18 U.S.C. § 2333(d)(2). Rather, a plaintiff must plead and prove that the defendant substantially assisted or conspired with “the person who committed” that attack. Id. Subsection 2333(d) thus limits claims asserted under that provision in exactly the way Plaintiffs resist—by requiring the plaintiff to link the defendant’s alleged assistance to the actual perpetrator of the terrorist attack. While Plaintiffs did not sue under subsection 2333(d)—pressing instead a theory of liability under subsection 2333(a) that “has the character of secondary liability,” Boim v. Holy Land Foundation for Relief & Dev., 549 F.3d 685, 691-692 (7th Cir. 2008) (en banc); see also infra pp. 54-55 (explaining Plaintiffs’ theory)—it is exceedingly unlikely that Congress intended to permit this secondary-like form of primary liability to sweep more broadly than the express secondary liability cause of action Congress created. 6 6 Section 2333(d), which provides limited causes of action for secondary liability (i.e., aiding and abetting and conspiracy), was added to the ATA’s civil remedy provision last year by the Justice Against Sponsors of Terrorism Act, Pub. L. No. 114-222 (enacted Sept. 28, 2016) (“JASTA”). By its terms, JASTA may be invoked in any civil action pending on the date of enactment. See JASTA, § 7(1). The Second Amended Complaint, filed before JASTA’s enactment, did not purport to include any claim under § 2333(d). At the November 9, 2016 motion to - 51 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 60 of 90 Further, failing to require a “direct relationship” between defendant’s conduct and plaintiff’s injury would create a grave risk of proximate cause becoming unmoored from its foundations. The more remote an injury, “the more difficult it becomes to ascertain the amount of a plaintiff’s damages attributable to the violation, as distinct from other, independent factors.” Holmes, 503 U.S. at 269. Allowing recovery for remote causes thus would threaten to impose liability on a defendant even when a plaintiff’s injury resulted from “any number of [other] reasons.” Anza, 547 U.S. at 458-459. Not only would that be unfair to the defendant, but it could also lead to duplicative recoveries. See Holmes, 503 U.S. at 269. These are exactly the problems that the “direct relationship” requirement is designed to avoid. The out-of-circuit, lower-court decisions Plaintiffs cite (at 32-34, 39-40) provide no basis to disregard the Supreme Court’s teachings. Indeed, the Second Circuit’s decision in Rothstein cuts the other way. In Rothstein, the court emphasized that by the time the ATA was enacted, the “‘by reason of’ language had a well-understood meaning”—the same meaning as that identical language in dismiss hearing, the district court asked Plaintiffs whether they needed another amendment to perfect their appeal, including any amendment for purposes of asserting a claim under the new provision. Plaintiffs acknowledged that if Section 230 bars the claims asserted in their Second Amended Complaint, it would also bar any claims they might seek to assert under § 2333(d), and for that reason they would not seek leave to amend if the court dismissed on Section 230 grounds. Dkt. 61 at 3:7-10, 18:14-17. - 52 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 61 of 90 RICO, the Sherman Act, and the Clayton Act. 708 F.3d at 95. And the court relied on that textual correspondence to reject plaintiffs’ argument that the ATA “permit[s] recovery on a showing of less than proximate cause, as the term is ordinarily used.” Id. As Plaintiffs note, Rothstein does quote their preferred “substantial factor” and “foreseeab[ility]” test. Id. at 91-92. But in the very next sentence, the opinion also quotes RICO’s “direct relationship” standard. See id. (“with respect to ‘proximate causation, the central question … is whether the alleged violation led directly to the plaintiff’s injuries’” (quoting Anza, 547 U.S. at 461) (emphasis added)). The reason is obvious: Because the allegations in that case satisfied neither standard, the difference between the two formulations was not at issue. Rothstein thus does not support Plaintiffs’ position. As for the Seventh Circuit’s decision in Boim and the smattering of district court decisions Plaintiffs cite (at 32-36), the standard they applied cannot be reconciled with the established meaning that the phrase “by reason of” had when the ATA was enacted. In fact, Boim did not even quote the text of the statute in its discussion of causation. Finally, Humanitarian Law Project, which Plaintiffs also discuss (at 34-36), construed a criminal material support statute, which, like most criminal laws, requires no proof of causation at all. Plaintiffs’ claims must therefore be dismissed unless the allegations pled in the Second Amended Complaint plausibly establish a “direct relationship” between - 53 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 62 of 90 Plaintiffs’ injuries and the act of international terrorism Twitter allegedly committed. Plaintiffs have never claimed that their allegations come close to satisfying that standard. 7 B. Plaintiffs’ Allegations Fail Even Under Their Preferred Test For Causation The district court correctly concluded that Plaintiffs’ causation theory is too attenuated to pass muster “[e]ven under [P]laintiffs’ proposed ‘substantial factor’ test.” AER21. To satisfy that (erroneously relaxed) test, Plaintiffs would have to plead facts plausibly establishing that an “act of international terrorism” committed by Twitter was “a substantial factor in the sequence of responsible causation” and that Plaintiffs’ injuries were “reasonably foreseeable or anticipated as a natural consequence” of that act. Br. 32. A straightforward reading of the Second Amended Complaint indicates that the “act of international terrorism” that caused the deaths of Mr. Fields and Mr. Creach is the attack carried out in Jordan by Abu Zaid. Twitter, however, did not commit that attack. To overcome this seemingly dispositive problem, Plaintiffs advance a theory that—they argue—establishes that 7 While there is no need to resolve the issue on this appeal, Plaintiffs are wrong (at 32 n.3) that the ATA’s civil remedy provision does not require but-for causation. The Second Circuit in Rothstein reasoned that the provision’s “by reason of” language requires a showing of both but-for and proximate cause. 708 F.3d at 95. And the Supreme Court has likewise “observed that … ‘the phrase, “by reason of,” requires at least a showing of “but for” causation.’” Burrage v. United States, 134 S. Ct. 881, 889 (2014). - 54 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 63 of 90 Twitter also committed an “act of international terrorism.” Specifically, Plaintiffs contend that by broadly providing accounts and/or direct messaging services to hundreds of millions of people, some of whom allegedly included ISIS and its sympathizers, Twitter violated two federal criminal statutes (18 U.S.C. §§ 2339A and 2339B) that prohibit the provision of “material support” to certain acts of terrorism or to a designated terrorist organization. See Br. 1-2, 10, 23. According to Plaintiffs, those alleged violations automatically meet the ATA’s multi-part definition of “international terrorism,” see id. 1-2, supposedly rendering Twitter’s provision of accounts and direct messaging services an “act of international terrorism.” There are many problems with Plaintiffs’ theory, some of which, although presented in Twitter’s successive motions to dismiss, the district court did not (and did not need to) reach. See Dkt. 27 at 23-25; Dkt. 32 at 13-15; Dkt. 53 at 11-13.8 8 Plaintiffs’ overall liability theory appears to be that providing a broadly available online platform that allegedly was used by some members or supporters of a terrorist organization violated §§ 2339A and 2339B, and that violating those criminal statutes in turn automatically satisfies the definition of “international terrorism” under § 2331(1). But violation of a criminal law is only one of several independent elements in the ATA’s definition of “international terrorism.” 18 U.S.C. § 2331(1)(A). The statutory definition also requires, among other things, that the defendant have committed “violent” or “dangerous” acts that “appear[ed] to be intended” to achieve specifically defined terrorist purposes. Id. § 2331(1)(A), (B). And Plaintiffs alleged no such act by Twitter. In particular, Twitter’s decision to make its services widely available to individuals all over the world was certainly not “violent” or “dangerous.” Nor could that decision possibly “lead an objective observer to conclude” that Twitter intended to accomplish any - 55 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 64 of 90 For present purposes, what matters is how reliance on that theory frames Plaintiffs’ proximate cause burden. Because Plaintiffs identify Twitter’s provision of accounts and Direct Messaging services as the supposed “act of international terrorism” allegedly committed by Twitter that allegedly injured them (e.g., Br. 37), Plaintiffs claims can survive only if this Court agrees that there is a legally sufficient causal link between Abu Zaid’s attack and the alleged provision of those accounts and services to ISIS. Any such link, however, is extraordinarily attenuated. Layer upon layer of speculation and innumerable intervening actors and events—each a more proximate cause than Twitter of Plaintiffs’ injuries—separate Twitter’s supposed “act of international terrorism” from Abu Zaid’s attack: (1) Twitter allegedly committed an “act of international terrorism” by allegedly failing to block ISIS and its sympathizers from opening Twitter accounts or using Direct Messaging. AER28, 30, 32-34 ¶¶ 1, 19, 36, 39, 43-45. (2) ISIS and its sympathizers allegedly used those accounts and Direct Messaging services to send messages, via Twitter’s platform, designed to recruit new members, raise money, and spread propaganda. AER33-38 ¶¶ 42-71. (3) Recipients of those messages allegedly responded by joining ISIS, contributing funds to ISIS, or paying more attention to ISIS. AER35 ¶¶ 47, 50-51. terrorist ends. Stansell v. BGP, Inc., 2011 WL 1296881, at *9 (M.D. Fla. Mar. 31, 2011). - 56 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 65 of 90 (4) Those recruits, funds, and publicity allegedly helped ISIS to grow into a larger terrorist organization. AER35 ¶ 52. (5) All that somehow caused or enabled Abu Zaid to become part of a “clandestine ISIS terror cell” when he was a student at some unspecified time in the past. AER41 ¶ 81. (6) Some amount of time after joining that cell Abu Zaid, acting alone, committed the attack that killed Mr. Fields and Mr. Creech. AER4041 ¶¶ 77-78. As the district court held, this alleged chain of causation is far too remote and tenuous to satisfy even Plaintiffs’ relaxed proximate cause test. AER22-23. On multiple occasions the Second Circuit has rejected similarly attenuated efforts to link businesses offering widely available services to acts of terrorism—under both Plaintiffs’ preferred “substantial factor” and “foreseeability” test and the “direct relationship” test that the ATA actually imposes. See supra p. 53 (explaining that Rothstein referenced both tests). In Al Rajhi, for example, the Second Circuit rejected allegations that “providing routine banking services to organizations and individuals said to be affiliated with al Qaeda … proximately caused the September 11, 2001 attacks or plaintiffs’ injuries.” 714 F.3d at 124. That court reached the same conclusion about allegations that UBS provided cash to a state sponsor of terrorism that would be used to cause and facilitate terrorist acts, Rothstein, 708 F.3d at 97, and about allegations that defendants “provided funding to purported charity organizations known to support terrorism that, in turn, provided funding to al Qaeda and other terrorist organizations,” Al Rajhi, 714 F.3d - 57 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 66 of 90 at 124. Each of those alleged causation theories was deemed “too speculative [and] attenuated to raise a plausible inference of proximate causation.” AER22 n.3. The same is true here. Indeed, the online services that Twitter makes generally available to the public are at least as far removed from terrorism as the “routine banking services” discussed in Al Rajhi. While Plaintiffs claim (at 40) that this case is somehow different because Twitter allegedly knew it was providing accounts to ISIS, Plaintiffs have alleged, at most, that Twitter generally knew that some of its hundreds of millions of users may have been affiliated with or sympathetic to ISIS. Plaintiffs certainly have not identified any particular accounts that Twitter knew about but refused to take down. To the contrary, they allege the opposite, acknowledging that Twitter repeatedly shuts down “ISIS-linked account[s]” when it becomes aware of them. SER13 ¶ 69. Al Rajhi does not permit liability to be imposed on Twitter simply because it may have generally known that a small and unspecified handful of its hundreds of millions of users were somehow connected to or supportive of ISIS. As for district court cases that Plaintiffs cite (at 33-34), all but one of them was decided before Al Rajhi made clear that the mere provision of routine services cannot establish proximate cause. And the facts in the one case decided after Al Rajhi established a far closer causal nexus than is alleged here. In that case - 58 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 67 of 90 (Linde), evidence revealed that the defendant bank had wired “‘martyr’ payments to the immediate relatives of the Hamas suicide attackers who perpetrated four of the terrorist attacks at issue,” and plaintiffs were able “to trace” other “transfers for Hamas leaders to specific attacks.” Linde v. Arab Bank, PLC, 97 F. Supp. 3d 287, 329 (E.D.N.Y. 2015), appeal pending No. 16-2134 (2d Cir.). Plaintiffs have alleged nothing remotely approaching that degree of connection between Abu Zaid and the “material support” that Twitter allegedly provided. Accepting Plaintiffs’ allegations as sufficient would have staggering consequences, exposing every online platform to possible liability for terrorist violence anywhere in the world simply because the terrorist who committed the attack may have been loosely affiliated with—or perhaps even just inspired by— some of the platform’s hundreds of millions of users. Indeed, the supposed “limits” that Plaintiffs identify (at 37-38)—“the importance of the material support” and its “temporal proximity” to an attack—actually reveal the unbelievable breadth of their theory, as such “limits” would do little to protect Twitter or scores of other on- and off-line services providers that make their services generally available to the public from facing potentially crippling liability for countless attacks that have no actual connection to those services. Stretching the doctrine of proximate cause that far would be problematic in any context, but it would be especially troubling here, given the serious First Amendment concerns - 59 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 68 of 90 that would be raised if online service providers were exposed to liability merely for operating a neutral forum on which some third persons had engaged in offensive or hateful speech. See Turner Broad. Sys., Inc. v. FCC, 512 U.S. 622, 636-637 (1994) (First Amendment protects an intermediary’s editorial judgments about what third-party speech to transmit). As the district court correctly held, even Plaintiffs’ preferred “substantial factor” and “foreseeability” test does not permit such expansive liability. CONCLUSION The judgment of the district court should be affirmed. May 31, 2017 Respectfully submitted, /s/ Seth P. Waxman SETH P. WAXMAN PATRICK J. CAROME ARI HOLTZBLATT WILMER CUTLER PICKERING HALE AND DORR LLP 1875 Pennsylvania Avenue NW Washington, DC 20006 (202) 663-6000 MARK D. FLANAGAN WILMER CUTLER PICKERING HALE AND DORR LLP 950 Page Mill Road Palo Alto, CA 94304 (650) 858-6000 - 60 - Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 69 of 90 STATEMENT OF RELATED CASES Counsel for Twitter is unaware of any related cases pending in this Court. Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 70 of 90 CERTIFICATE OF COMPLIANCE Pursuant to Fed. R. App. P. 32(g)(1), the undersigned hereby certifies that this brief complies with the type-volume limitation of Fed. R. App. P. 32(a)(7)(B)(i) and Ninth Circuit Rule 32-1(a). 1. Exclusive of the exempted portions of the brief, as provided in Fed. R. App. P. 32(f), the brief contains 13,997 words. 2. The brief has been prepared in proportionally spaced typeface using Microsoft Word 2016 in 14 point Times New Roman font. As permitted by Fed. R. App. P. 32(a)(7)(B), the undersigned has relied upon the word count feature of this word processing system in preparing this certificate. /s/ Seth P. Waxman SETH P. WAXMAN May 31, 2017 Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 71 of 90 STATUTORY ADDENDUM Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 72 of 90 ADDENDUM TABLE OF CONTENTS Page 47 U.S.C. § 230 ................................................................................................. Add. 1 18 U.S.C. § 2331 ............................................................................................... Add. 5 18 U.S.C. § 2333 ............................................................................................... Add. 7 18 U.S.C. § 2339A ............................................................................................ Add. 9 18 U.S.C. § 2339B .......................................................................................... Add. 10 Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 73 of 90 47 U.S.C. § 230 Protection for private blocking and screening of offensive material (a) Findings The Congress finds the following: (1) The rapidly developing array of Internet and other interactive computer services available to individual Americans represent an extraordinary advance in the availability of educational and informational resources to our citizens. (2) These services offer users a great degree of control over the information that they receive, as well as the potential for even greater control in the future as technology develops. (3) The Internet and other interactive computer services offer a forum for a true diversity of political discourse, unique opportunities for cultural development, and myriad avenues for intellectual activity. (4) The Internet and other interactive computer services have flourished, to the benefit of all Americans, with a minimum of government regulation. (5) Increasingly Americans are relying on interactive media for a variety of political, educational, cultural, and entertainment services. (b) Policy It is the policy of the United States— (1) to promote the continued development of the Internet and other interactive computer services and other interactive media; (2) to preserve the vibrant and competitive free market that presently exists for the Internet and other interactive computer services, unfettered by Federal or State regulation; (3) to encourage the development of technologies which maximize user control over what information is received by individuals, families, and schools who use the Internet and other interactive computer services; Add. 1 Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 74 of 90 (4) to remove disincentives for the development and utilization of blocking and filtering technologies that empower parents to restrict their children's access to objectionable or inappropriate online material; and (5) to ensure vigorous enforcement of Federal criminal laws to deter and punish trafficking in obscenity, stalking, and harassment by means of computer. (c) Protection for “Good Samaritan” blocking and screening of offensive material (1) Treatment of publisher or speaker No provider or user of an interactive computer service shall be treated as the publisher or speaker of any information provided by another information content provider. (2) Civil liability No provider or user of an interactive computer service shall be held liable on account of— (A) any action voluntarily taken in good faith to restrict access to or availability of material that the provider or user considers to be obscene, lewd, lascivious, filthy, excessively violent, harassing, or otherwise objectionable, whether or not such material is constitutionally protected; or (B) any action taken to enable or make available to information content providers or others the technical means to restrict access to material described in paragraph (1).1 (d) Obligations of interactive computer service A provider of interactive computer service shall, at the time of entering an agreement with a customer for the provision of interactive computer service and in a manner deemed appropriate by the provider, notify such customer that parental control protections (such as computer hardware, software, or filtering services) are commercially available that may assist the customer in limiting access to material that is harmful to minors. Such notice shall identify, or provide the customer with access to information identifying, current providers of such protections. Add. 2 Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 75 of 90 (e) Effect on other laws (1) No effect on criminal law Nothing in this section shall be construed to impair the enforcement of section 223 or 231 of this title, chapter 71 (relating to obscenity) or 110 (relating to sexual exploitation of children) of Title 18, or any other Federal criminal statute. (2) No effect on intellectual property law Nothing in this section shall be construed to limit or expand any law pertaining to intellectual property. (3) State law Nothing in this section shall be construed to prevent any State from enforcing any State law that is consistent with this section. No cause of action may be brought and no liability may be imposed under any State or local law that is inconsistent with this section. (4) No effect on communications privacy law Nothing in this section shall be construed to limit the application of the Electronic Communications Privacy Act of 1986 or any of the amendments made by such Act, or any similar State law. (f) Definitions As used in this section: (1) Internet The term “Internet” means the international computer network of both Federal and non-Federal interoperable packet switched data networks. (2) Interactive computer service The term “interactive computer service” means any information service, system, or access software provider that provides or enables computer access by multiple users to a computer server, including specifically a service or system that provides access to the Internet and such systems operated or services offered by libraries or educational institutions. Add. 3 Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 76 of 90 (3) Information content provider The term “information content provider” means any person or entity that is responsible, in whole or in part, for the creation or development of information provided through the Internet or any other interactive computer service. (4) Access software provider The term “access software provider” means a provider of software (including client or server software), or enabling tools that do any one or more of the following: (A) filter, screen, allow, or disallow content; (B) pick, choose, analyze, or digest content; or (C) transmit, receive, display, forward, cache, search, subset, organize, reorganize, or translate content Add. 4 Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 77 of 90 18 U.S.C. § 2331 Definitions As used in this chapter— (1) the term “international terrorism” means activities that— (A) involve violent acts or acts dangerous to human life that are a violation of the criminal laws of the United States or of any State, or that would be a criminal violation if committed within the jurisdiction of the United States or of any State; (B) appear to be intended— (i) to intimidate or coerce a civilian population; (ii) to influence the policy of a government by intimidation or coercion; or (iii) to affect the conduct of a government by mass destruction, assassination, or kidnapping; and (C) occur primarily outside the territorial jurisdiction of the United States, or transcend national boundaries in terms of the means by which they are accomplished, the persons they appear intended to intimidate or coerce, or the locale in which their perpetrators operate or seek asylum; (2) the term “national of the United States” has the meaning given such term in section 101(a)(22) of the Immigration and Nationality Act; (3) the term “person” means any individual or entity capable of holding a legal or beneficial interest in property; (4) the term “act of war” means any act occurring in the course of— (A) declared war; (B) armed conflict, whether or not war has been declared, between two or more nations; or Add. 5 Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 78 of 90 (C) armed conflict between military forces of any origin; and (5) the term “domestic terrorism” means activities that— (A) involve acts dangerous to human life that are a violation of the criminal laws of the United States or of any State; (B) appear to be intended— (i) to intimidate or coerce a civilian population; (ii) to influence the policy of a government by intimidation or coercion; or (iii) to affect the conduct of a government by mass destruction, assassination, or kidnapping; and (C) occur primarily within the territorial jurisdiction of the United States. Add. 6 Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 79 of 90 18 U.S.C. § 2333 Civil remedies (a) Action and jurisdiction.— Any national of the United States injured in his or her person, property, or business by reason of an act of international terrorism, or his or her estate, survivors, or heirs, may sue therefor in any appropriate district court of the United States and shall recover threefold the damages he or she sustains and the cost of the suit, including attorney's fees. (b) Estoppel under United States law. — A final judgment or decree rendered in favor of the United States in any criminal proceeding under section 1116, 1201, 1203, or 2332 of this title or section 46314, 46502, 46505, or 46506 of title 49 shall estop the defendant from denying the essential allegations of the criminal offense in any subsequent civil proceeding under this section. (c) Estoppel under foreign law. — A final judgment or decree rendered in favor of any foreign state in any criminal proceeding shall, to the extent that such judgment or decree may be accorded full faith and credit under the law of the United States, estop the defendant from denying the essential allegations of the criminal offense in any subsequent civil proceeding under this section. (d) Liability. — (1) Definition. — In this subsection, the term “person” has the meaning given the term in section 1 of title 1. Add. 7 Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 80 of 90 (2) Liability. — In an action under subsection (a) for an injury arising from an act of international terrorism committed, planned, or authorized by an organization that had been designated as a foreign terrorist organization under section 219 of the Immigration and Nationality Act (8 U.S.C. 1189), as of the date on which such act of international terrorism was committed, planned, or authorized, liability may be asserted as to any person who aids and abets, by knowingly providing substantial assistance, or who conspires with the person who committed such an act of international terrorism. Add. 8 Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 81 of 90 18 U.S.C. § 2339A Providing material support to terrorists (a) Offense.— Whoever provides material support or resources or conceals or disguises the nature, location, source, or ownership of material support or resources, knowing or intending that they are to be used in preparation for, or in carrying out, a violation of section 32, 37, 81, 175, 229, 351, 831, 842(m) or (n), 844(f) or (i), 930(c), 956, 1091, 1114, 1116, 1203, 1361, 1362, 1363, 1366, 1751, 1992, 2155, 2156, 2280, 2281, 2332, 2332a, 2332b, 2332f, 2340A, or 2442 of this title, section 236 of the Atomic Energy Act of 1954 (42 U.S.C. 2284), section 46502 or 60123(b) of title 49, or any offense listed in section 2332b(g)(5)(B) (except for sections 2339A and 2339B) or in preparation for, or in carrying out, the concealment of an escape from the commission of any such violation, or attempts or conspires to do such an act, shall be fined under this title, imprisoned not more than 15 years, or both, and, if the death of any person results, shall be imprisoned for any term of years or for life. A violation of this section may be prosecuted in any Federal judicial district in which the underlying offense was committed, or in any other Federal judicial district as provided by law. (b) Definitions.—As used in this section— (1) the term “material support or resources” means any property, tangible or intangible, or service, including currency or monetary instruments or financial securities, financial services, lodging, training, expert advice or assistance, safehouses, false documentation or identification, communications equipment, facilities, weapons, lethal substances, explosives, personnel (1 or more individuals who may be or include oneself), and transportation, except medicine or religious materials; (2) the term “training” means instruction or teaching designed to impart a specific skill, as opposed to general knowledge; and (3) the term “expert advice or assistance” means advice or assistance derived from scientific, technical or other specialized knowledge. Add. 9 Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 82 of 90 18 U.S.C. § 2339B Providing material support or resources to designated foreign terrorist organizations (a) Prohibited activities. — (1) Unlawful conduct. — Whoever knowingly provides material support or resources to a foreign terrorist organization, or attempts or conspires to do so, shall be fined under this title or imprisoned not more than 20 years, or both, and, if the death of any person results, shall be imprisoned for any term of years or for life. To violate this paragraph, a person must have knowledge that the organization is a designated terrorist organization (as defined in subsection (g)(6)), that the organization has engaged or engages in terrorist activity (as defined in section 212(a)(3)(B) of the Immigration and Nationality Act), or that the organization has engaged or engages in terrorism (as defined in section 140(d)(2) of the Foreign Relations Authorization Act, Fiscal Years 1988 and 1989). (2) Financial institutions.—Except as authorized by the Secretary, any financial institution that becomes aware that it has possession of, or control over, any funds in which a foreign terrorist organization, or its agent, has an interest, shall— (A) retain possession of, or maintain control over, such funds; and (B) report to the Secretary the existence of such funds in accordance with regulations issued by the Secretary. (b) Civil penalty.—Any financial institution that knowingly fails to comply with subsection (a)(2) shall be subject to a civil penalty in an amount that is the greater of— (A) $50,000 per violation; or (B) twice the amount of which the financial institution was required under subsection (a)(2) to retain possession or control. Add. 10 Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 83 of 90 (c) Injunction.— Whenever it appears to the Secretary or the Attorney General that any person is engaged in, or is about to engage in, any act that constitutes, or would constitute, a violation of this section, the Attorney General may initiate civil action in a district court of the United States to enjoin such violation. (d) Extraterritorial jurisdiction.— (1) In general.—There is jurisdiction over an offense under subsection (a) if— (A) an offender is a national of the United States (as defined in section 101(a)(22) of the Immigration and Nationality Act (8 U.S.C. 1101(a)(22))) or an alien lawfully admitted for permanent residence in the United States (as defined in section 101(a)(20) of the Immigration and Nationality Act (8 U.S.C. 1101(a)(20))); (B) an offender is a stateless person whose habitual residence is in the United States; (C) after the conduct required for the offense occurs an offender is brought into or found in the United States, even if the conduct required for the offense occurs outside the United States; (D) the offense occurs in whole or in part within the United States; (E) the offense occurs in or affects interstate or foreign commerce; or (F) an offender aids or abets any person over whom jurisdiction exists under this paragraph in committing an offense under subsection (a) or conspires with any person over whom jurisdiction exists under this paragraph to commit an offense under subsection (a). (2) Extraterritorial jurisdiction.— There is extraterritorial Federal jurisdiction over an offense under this section. Add. 11 Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 84 of 90 (e) Investigations.— (1) In general. — The Attorney General shall conduct any investigation of a possible violation of this section, or of any license, order, or regulation issued pursuant to this section. (2) Coordination with the Department of the Treasury.—The Attorney General shall work in coordination with the Secretary in investigations relating to— (A) the compliance or noncompliance by a financial institution with the requirements of subsection (a)(2); and (B) civil penalty proceedings authorized under subsection (b). (3) Referral.— Any evidence of a criminal violation of this section arising in the course of an investigation by the Secretary or any other Federal agency shall be referred immediately to the Attorney General for further investigation. The Attorney General shall timely notify the Secretary of any action taken on referrals from the Secretary, and may refer investigations to the Secretary for remedial licensing or civil penalty action. (f) Classified information in civil proceedings brought by the United States.— (1) Discovery of classified information by defendants.— (A) Request by United States.—In any civil proceeding under this section, upon request made ex parte and in writing by the United States, a court, upon a sufficient showing, may authorize the United States to-(i) redact specified items of classified information from documents to be introduced into evidence or made available to the defendant through discovery under the Federal Rules of Civil Procedure; Add. 12 Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 85 of 90 (ii) substitute a summary of the information for such classified documents; or (iii) substitute a statement admitting relevant facts that the classified information would tend to prove. (B) Order granting request.—If the court enters an order granting a request under this paragraph, the entire text of the documents to which the request relates shall be sealed and preserved in the records of the court to be made available to the appellate court in the event of an appeal. (C) Denial of request.—If the court enters an order denying a request of the United States under this paragraph, the United States may take an immediate, interlocutory appeal in accordance with paragraph (5). For purposes of such an appeal, the entire text of the documents to which the request relates, together with any transcripts of arguments made ex parte to the court in connection therewith, shall be maintained under seal and delivered to the appellate court. (2) Introduction of classified information; precautions by court.— (A) Exhibits.—To prevent unnecessary or inadvertent disclosure of classified information in a civil proceeding brought by the United States under this section, the United States may petition the court ex parte to admit, in lieu of classified writings, recordings, or photographs, one or more of the following: (i) Copies of items from which classified information has been redacted. (ii) Stipulations admitting relevant facts that specific classified information would tend to prove. (iii) A declassified summary of the specific classified information. (B) Determination by court.—The court shall grant a request under this paragraph if the court finds that the redacted item, stipulation, or summary is sufficient to allow the defendant to prepare a defense. Add. 13 Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 86 of 90 (3) Taking of trial testimony.— (A) Objection.— During the examination of a witness in any civil proceeding brought by the United States under this subsection, the United States may object to any question or line of inquiry that may require the witness to disclose classified information not previously found to be admissible. (B) Action by court.—In determining whether a response is admissible, the court shall take precautions to guard against the compromise of any classified information, including— (i) permitting the United States to provide the court, ex parte, with a proffer of the witness's response to the question or line of inquiry; and (ii) requiring the defendant to provide the court with a proffer of the nature of the information that the defendant seeks to elicit. (C) Obligation of defendant.—In any civil proceeding under this section, it shall be the defendant's obligation to establish the relevance and materiality of any classified information sought to be introduced. (4) Appeal. — If the court enters an order denying a request of the United States under this subsection, the United States may take an immediate interlocutory appeal in accordance with paragraph (5). (5) Interlocutory appeal.— (A) Subject of appeal.—An interlocutory appeal by the United States shall lie to a court of appeals from a decision or order of a district court-(i) authorizing the disclosure of classified information; Add. 14 Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 87 of 90 (ii) imposing sanctions for nondisclosure of classified information; or (iii) refusing a protective order sought by the United States to prevent the disclosure of classified information. (B) Expedited consideration.— (i) In general.— An appeal taken pursuant to this paragraph, either before or during trial, shall be expedited by the court of appeals. (ii) Appeals prior to trial.—If an appeal is of an order made prior to trial, an appeal shall be taken not later than 14 days after the decision or order appealed from, and the trial shall not commence until the appeal is resolved. (iii) Appeals during trial.—If an appeal is taken during trial, the trial court shall adjourn the trial until the appeal is resolved, and the court of appeals-(I) shall hear argument on such appeal not later than 4 days after the adjournment of the trial, excluding intermediate weekends and holidays; (II) may dispense with written briefs other than the supporting materials previously submitted to the trial court; (III) shall render its decision not later than 4 days after argument on appeal, excluding intermediate weekends and holidays; and (IV) may dispense with the issuance of a written opinion in rendering its decision. Add. 15 Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 88 of 90 (C) Effect of ruling.— An interlocutory appeal and decision shall not affect the right of the defendant, in a subsequent appeal from a final judgment, to claim as error reversal by the trial court on remand of a ruling appealed from during trial. (6) Construction.— Nothing in this subsection shall prevent the United States from seeking protective orders or asserting privileges ordinarily available to the United States to protect against the disclosure of classified information, including the invocation of the military and State secrets privilege. (g) Definitions.—As used in this section— (1) the term “classified information” has the meaning given that term in section 1(a) of the Classified Information Procedures Act (18 U.S.C. App.); (2) the term “financial institution” has the same meaning as in section 5312(a)(2) of title 31, United States Code; (3) the term “funds” includes coin or currency of the United States or any other country, traveler's checks, personal checks, bank checks, money orders, stocks, bonds, debentures, drafts, letters of credit, any other negotiable instrument, and any electronic representation of any of the foregoing; (4) the term “material support or resources” has the same meaning given that term in section 2339A (including the definitions of “training” and “expert advice or assistance” in that section); (5) the term “Secretary” means the Secretary of the Treasury; and (6) the term “terrorist organization” means an organization designated as a terrorist organization under section 219 of the Immigration and Nationality Act. Add. 16 Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 89 of 90 (h) Provision of personnel.— No person may be prosecuted under this section in connection with the term “personnel” unless that person has knowingly provided, attempted to provide, or conspired to provide a foreign terrorist organization with 1 or more individuals (who may be or include himself) to work under that terrorist organization's direction or control or to organize, manage, supervise, or otherwise direct the operation of that organization. Individuals who act entirely independently of the foreign terrorist organization to advance its goals or objectives shall not be considered to be working under the foreign terrorist organization's direction and control. (i) Rule of construction.— Nothing in this section shall be construed or applied so as to abridge the exercise of rights guaranteed under the First Amendment to the Constitution of the United States. (j) Exception.— No person may be prosecuted under this section in connection with the term “personnel”, “training”, or “expert advice or assistance” if the provision of that material support or resources to a foreign terrorist organization was approved by the Secretary of State with the concurrence of the Attorney General. The Secretary of State may not approve the provision of any material support that may be used to carry out terrorist activity (as defined in section 212(a)(3)(B)(iii) of the Immigration and Nationality Act). Add. 17 Case: 16-17165, 05/31/2017, ID: 10455211, DktEntry: 23, Page 90 of 90 CERTIFICATE OF SERVICE I hereby certify that on this 31st day of May 2017 I electronically filed the foregoing with the Clerk of the Court for the United States Court of Appeals for the Ninth Circuit using the appellate CM/ECF system. Counsel for all parties to the case are registered CM/ECF users and will be served by the appellate CM/ECF system. /s/ Seth P. Waxman SETH P. WAXMAN