Twitter, Inc. v. Taamneh

598 U.S. 471

• Court: Supreme Court of the United States
• Decided: May 18, 2023
• Docket: 21-1496
• Status: Published

Opinion

PRELIMINARY PRINT

Volume 598 U. S. Part 2 Pages 471–507

OFFICIAL REPORTS OF

THE SUPREME COURT May 18, 2023

REBECCA A. WOMELDORF reporter of decisions

NOTICE: This preliminary print is subject to formal revision before the bound volume is published. Users are requested to notify the Reporter of Decisions, Supreme Court of the United States, Washington, D.C. 20543, pio@supremecourt.gov, of any typographical or other formal errors. OCTOBER TERM, 2022 471

Syllabus

TWITTER, INC. v. TAAMNEH et al.

certiorari to the united states court of appeals for the ninth circuit No. 21–1496. Argued February 22, 2023—Decided May 18, 2023 In 2017, Abdulkadir Masharipov carried out a terrorist attack on the Reina nightclub in Istanbul, Turkey, on behalf of the Islamic State of Iraq and Syria (ISIS), a designated Foreign Terrorist Organization. Masharipov killed Nawras Alassaf and 38 others. Alassaf 's family then brought this suit under 18 U. S. C. § 2333, an Antiterrorism Act (ATA) provision that permits U. S. nationals who have been “injured . . . by reason of an act of international terrorism” to fle a civil suit for damages. Instead of suing ISIS directly under § 2333(a), the plaintiffs (respondents here) invoked § 2333(d)(2) to sue three of the largest social-media companies in the world—Facebook, Twitter (petitioner here), and Google (which owns YouTube)—for aiding and abetting ISIS. The parties today agree on the basic aspects of these platforms: Bil- lions of people from around the world have signed up for them and upload massive amounts of content each day. Defendants proft from that content by placing advertisements on or near it and use “recom- mendation” algorithms that match content, advertisements, and users based on information about the use, advertisement, and content being viewed. As the parties represent things, the algorithms here match any content with any user who is more likely to view that content, and the platforms perform little to no front-end screening on any content before it is uploaded. Plaintiffs, however, allege that for several years the companies have knowingly allowed ISIS and its supporters to use their platforms and “recommendation” algorithms as tools for recruiting, fundraising, and spreading propaganda; plaintiffs further allege that these companies have, in the process, profted from the advertisements placed on ISIS' tweets, posts, and videos. The District Court dismissed the complaint for failure to state a claim, but the Ninth Circuit reversed. Held: Plaintiffs' allegations that these social-media companies aided and abetted ISIS in its terrorist attack on the Reina nightclub fail to state a claim under 18 U. S. C. § 2333(d)(2). Pp. 482–507. (a) In 2016, Congress enacted the Justice Against Sponsors of Terror- ism Act (JASTA) to impose secondary civil liability on anyone “who aids and abets, by knowingly providing substantial assistance, or who conspires with the person who committed such an act of international 472 TWITTER, INC. v. TAAMNEH

terrorism.” § 2333(d)(2). The question here is whether the conduct of the social-media company defendants gives rise to aiding-and-abetting liability for the Reina nightclub attack. Pp. 482–484. (b) The text of JASTA begs two questions: What does it mean to “aid and abet”? And, what precisely must the defendant have “aided and abetted”? Pp. 484–497. (1) Nothing in the statute defnes any of the critical terms in the phrase “aids and abets, by knowingly providing substantial assistance.” The term “aids and abets,” however, is a familiar common-law term and thus presumably “brings the old soil” with it. Sekhar v. United States, 570 U. S. 729, 733. Congress also provided additional context in JASTA by pointing to Halberstam v. Welch, 705 F. 2d 472, as “provid[ing] the proper legal framework” for “civil aiding and abetting and conspiracy liability.” 130 Stat. 852. Halberstam's legal framework, viewed in context of the common-law tradition from which it arose, confrms that “aids and abets” in § 2333(d)(2) refers to a conscious, voluntary, and cul- pable participation in another's wrongdoing. Pp. 485–493. (i) In Halberstam, the D. C. Circuit undertook an extensive sur- vey of the common law with respect to aiding and abetting and synthe- sized the surveyed cases as resting on three main elements: (1) there must be a wrongful act causing an injury performed by the person whom the defendant aided; (2) at the time assistance was provided, the defendant must have been “generally aware of his role as part of an overall illegal or tortious activity;” and (3) the defendant must have “knowingly and substantially assist[ed] the principal violation.” 705 F. 2d, at 477. The court then articulated six factors to help determine whether a defendant's assistance was “substantial.” They are (1) “the nature of the act assisted,” (2) the “amount of assistance” provided, (3) whether the defendant was “present at the time” of the principal tort, (4) the defendant's “relation to the tortious actor,” (5) the “defendant's state of mind,” and (6) the “duration of the assistance” given. Id., at 488 (emphasis deleted). Halberstam also clarifed that those who aid and abet “a tortious act may be liable” not only for the act itself but also “for other reasonably foreseeable acts done in connection with it.” Id., at 484. Finally, the court warned that its formulations should “not be accepted as immutable components” but should be “adapted as new cases test their usefulness in evaluating vicarious liability.” Id., at 489. Pp. 485–487. (ii) Because the allegations here—involving international terror- ist networks and world-spanning internet platforms—are a far cry from the facts of Halberstam, the Court must ascertain the basic thrust of Halberstam's elements to determine how to adapt them to the facts of this case. To do so, the Court turns to the common law of aiding and Cite as: 598 U. S. 471 (2023) 473

abetting upon which Halberstam rested, and to which JASTA's common- law terminology points. At common law, the basic “view of culpability” animating aiding and abetting liability is that “a person may be respon- sible for a crime he has not personally carried out if he helps another to complete its commission.” Rosemond v. United States, 572 U. S. 65, 70. However, the concept of “helping” in the commission of a crime or a tort has never been boundless and ordinarily requires some level of blameworthy conduct; those limits ensure that aiding and abetting does not sweep in mere passive bystanders or those who, for example, simply deliver mail that happens to aid criminals. In tort law, many cases have thus required a voluntary, conscious, and culpable participation in the wrongful conduct to establish aiding and abetting. In doing so, they further articulated Halberstam's framework to capture those lim- its. As above, that framework requires that the defendant give know- ing and substantial assistance to the primary tortfeasor; notably, courts often viewed those twin requirements as working in tandem, with a lesser showing of one demanding a greater showing of the other to es- tablish a conscious, culpable participation in the tort. Pp. 487–492. (iii) Halberstam differentiated types of aid along the same culpa- bility axis that grounded the common law. And its six factors for “sub- stantial assistance” call for the same balancing that courts had under- taken previously between the nature and amount of assistance and the defendant's scienter. Pp. 492–493. (2) The parties then vigorously dispute what precisely a defendant must aid and abet under § 2333(d)(2). Plaintiffs assert that it is “the person,” while defendants insist that it is the “act of international ter- rorism.” That syntactic dispute makes little difference here, because aiding and abetting is inherently a rule of secondary liability for specifc wrongful acts. In the tort context, liability is imposed only when some- one commits (not merely agrees to commit) an actual tort. And in this case, the ATA limits that liability to injuries caused by an “act of inter- national terrorism,” § 2333(a). It thus is not enough for a defendant to have given substantial assistance to a transcendent enterprise. A defendant must have aided and abetted (by knowingly providing sub- stantial assistance) another person in the commission of the actionable wrong—here, an act of international terrorism. However, that does not require a strict nexus between the assistance and the wrongful act; de- fendants are liable for other torts that are the foreseeable risk of an intended tort, and an aider and abettor can assist someone without knowing all the details of his plan. Plus, in appropriate circumstances, a defendant's role in an illicit enterprise can be so systemic and inten- tional that the defendant aids and abets each act of the enterprise—as in Halberstam itself. 474 TWITTER, INC. v. TAAMNEH

To summarize the requirements of § 2333(d)(2), the phrase “aids and abets, by knowingly providing substantial assistance” points to the ele- ments and factors articulated by Halberstam. Those elements and fac- tors should not be taken as infexible codes but should be understood in light of the common law and applied as a framework designed to hold defendants liable when they consciously and culpably “participate[d] in” a tortious act in such a way as to help “make it succeed.” Nye & Nissen v. United States, 336 U. S. 613, 619. Pp. 493–497. (c) Plaintiffs have satisfed Halberstam's frst two elements by alleg- ing both that ISIS committed a wrong and that defendants knew they were playing some sort of role in ISIS' enterprise. But plaintiffs' alle- gations do not show that defendants gave such knowing and substantial assistance to ISIS that they culpably participated in the Reina attack. Pp. 497–506. (1) Plaintiffs allege that defendants aided and abetted ISIS in the following ways: First, they provided social-media platforms, which are generally available to the internet-using public; ISIS was able to upload content to those platforms and connect with third parties on them. Second, defendants' recommendation algorithms matched ISIS-related content to users most likely to be interested in that content. And, third, defendants knew that ISIS was uploading this content but took insuffcient steps to ensure that its content was removed. Plaintiffs do not allege that ISIS or Masharipov used defendants' platforms to plan or coordinate the Reina attack. Nor do plaintiffs allege that defendants gave ISIS any special treatment or words of encouragement. Nor is there reason to think that defendants carefully screened any content before allowing users to upload it onto their platforms. None of plaintiffs' allegations suggest that defendants culpably “asso- ciate[d themselves] with” the Reina attack, “participate[d] in it as some- thing that [they] wishe[d] to bring about,” or sought “by [their] action to make it succeed.” Nye & Nissen, 336 U. S., at 619 (internal quota- tion marks omitted). Defendants' mere creation of their media plat- forms is no more culpable than the creation of email, cell phones, or the internet generally. And defendants' recommendation algorithms are merely part of the infrastructure through which all the content on their platforms is fltered. Moreover, the algorithms have been presented as agnostic as to the nature of the content. At bottom, the allegations here rest less on affrmative misconduct and more on passive nonfea- sance. To impose aiding-and-abetting liability for passive nonfeasance, plaintiffs must make a strong showing of assistance and scienter. Plaintiffs fail to do so. First, the relationship between defendants and the Reina attack is highly attenuated. Plaintiffs make no allegations that defendants' rela- Cite as: 598 U. S. 471 (2023) 475

tionship with ISIS was signifcantly different from their arm's length, passive, and largely indifferent relationship with most users. And their relationship with the Reina attack is even further removed, given the lack of allegations connecting the Reina attack with ISIS' use of these platforms. Second, plaintiffs provide no reason to think that defend- ants were consciously trying to help or otherwise participate in the Reina attack, and they point to no actions that would normally support an aiding-and-abetting claim. Plaintiffs' complaint rests heavily on defendants' failure to act; yet plaintiffs identify no duty that would require defendants or other communication-providing services to terminate customers after dis- covering that the customers were using the service for illicit ends. Even if such a duty existed in this case, it would not transform defend- ants' distant inaction into knowing and substantial assistance that could establish aiding and abetting the Reina attack. And the expansive scope of plaintiffs' claims would necessarily hold defendants liable as having aided and abetted each and every ISIS terrorist act committed anywhere in the world. The allegations plaintiffs make here are not the type of pervasive, systemic, and culpable assistance to a series of terrorist activities that could be described as aiding and abetting each

terrorist act by ISIS. In this case, the failure to allege that the platforms here do more than transmit information by billions of people—most of whom use the platforms for interactions that once took place via mail, on the phone, or in public areas—is insuffcient to state a claim that defendants know- ingly gave substantial assistance and thereby aided and abetted ISIS' acts. A contrary conclusion would effectively hold any sort of commu- nications provider liable for any sort of wrongdoing merely for knowing that the wrongdoers were using its services and failing to stop them. That would run roughshod over the typical limits on tort liability and unmoor aiding and abetting from culpability. Pp. 498–503. (2) The Ninth Circuit's analysis obscured the essence of aiding-and- abetting liability. First, the Ninth Circuit framed the issue of substan- tial assistance as turning on defendants' assistance to ISIS' activities in general, rather than with respect to the Reina attack. Next, the Ninth Circuit misapplied the “knowing” half of “knowing and substantial as- sistance,” which is designed to capture the defendants' state of mind with respect to their actions and the tortious conduct (even if not always the particular terrorist act). Finally, the Ninth Circuit appears to have regarded Halberstam's six substantiality factors as a sequence of dispar- ate, unrelated considerations without a common conceptual core. In doing so, the Ninth Circuit focused primarily on the value of defendants' 476 TWITTER, INC. v. TAAMNEH

platforms to ISIS, rather than whether defendants culpably associated themselves with ISIS' actions. Pp. 503–505. (3) There is also one set of allegations specifc to Google: that Google reviewed and approved ISIS videos on YouTube as part of a revenue-sharing system and thereby shared advertising revenue with ISIS. But the complaint here alleges nothing about the amount of money that Google supposedly shared with ISIS, the number of accounts approved for revenue sharing, or the content of the videos that were approved. Nor does it give any other reason to view Google's revenue sharing as substantial assistance. Without more, plaintiffs thus have not plausibly alleged that Google knowingly provided substantial assist- ance to the Reina attack, let alone (as their theory of liability would require) every single terrorist act committed by ISIS. Pp. 505–506. (d) The concepts of aiding and abetting and substantial assistance do not lend themselves to crisp, bright-line distinctions. Applying the guideposts provided by the common law and Halberstam, the nexus be- tween defendants and the Reina attack is far removed. As alleged by plaintiffs, defendants designed virtual platforms and knowingly failed to do “enough” to remove ISIS-affliated users and ISIS-related content from their platforms. Yet, plaintiffs have failed to allege that defend- ants intentionally provided any substantial aid to the Reina attack or otherwise consciously participated in it—much less that defendants so pervasively and systemically assisted ISIS as to render them liable for every ISIS attack. Plaintiffs accordingly have failed to state a claim under § 2333(d)(2). Pp. 506–507. 2 F. 4th 871, reversed.

Thomas, J., delivered the opinion for a unanimous Court. Jackson, J., fled a concurring opinion, post, p. 507.

Seth P. Waxman argued the cause for petitioner. With him on the briefs were Patrick J. Carome, Ari Holtzblatt, Claire H. Chung, Amy Lishinski, Andres C. Salinas, and Nathaniel W. Reisinger. Paul D. Clement fled briefs for respondents Facebook, Inc., et al. supporting petitioner under this Court's Rule 12.6. With him on the briefs were Erin E. Murphy, Theodore J. Boutrous, Jr., Helgi C. Walker, Jonathan C. Bond, Jacob T. Spencer, Lisa S. Blatt, Sarah M. Harris, Aaron Z. Roper, Brian M. Willen, Steffen N. John- son, Lauren Gallo White, Cassandra Knight, Nora Puckett, and Jennifer Newstead. Cite as: 598 U. S. 471 (2023) 477

Opinion of the Court

Deputy Solicitor General Kneedler argued the cause for the United States as amicus curiae urging reversal. With him on the brief were Acting Solicitor General Fletcher, Deputy Assistant Attorney General Granston, Erica L. Ross, and Sharon Swingle. Eric Schnapper argued the cause for respondents. With him on the brief were Kei th L. Altman and Robert J. Tolchin.*

Justice Thomas delivered the opinion of the Court. Under 18 U. S. C. § 2333, United States nationals who have been “injured . . . by reason of an act of international terror-

*Briefs of amici curiae urging reversal were fled for the Center for Democracy & Technology et al. by Kathleen R. Hartnett, David D. Cole, Aaron Mackey, Bruce D. Brown, Alex Abdo, Jennifer Stisa Granick, and Patrick Toomey; for the Chamber of Commerce of the United States of America et al. by Andrew J. Pincus and Paul Lettow; for the Computer & Communications Industry Association et al. by Scott A. Keller and Steven P. Lehotsky; for the Institute of International Bankers et al. by Marc J. Gottridge, Lisa J. Fried, and Barron M. Flood; for InterAction et al. by Timothy P. Harkness, Linda H. Martin, David Y. Livshiz, and Scott A. Eisman; and for the Pharmaceutical Research and Manufacturers of America by Beth S. Brinkman, David M. Zionts, Michael X. Imbroscio, and S. Conrad Scott. Briefs of amici curiae urging affrmance were fled for the American Association of Jewish Lawyers and Jurists et al. by Hal R. Morris; for the Anti-Defamation League by James P. Bonner, Patrick L. Rocco, and Steven M. Freeman; for Anti-Terrorism Act Scholars by Stephen I. Vla- deck, pro se; for Concerned Women for America by Mario Diaz, Gavriel Mairone, and Adora Sauer; for Former National Security Offcials by Gary M. Osen, Michael J. Radine, and Ari Ungar; for the Plaintiffs' Gen- eral Committee by Carter G. Phillips, Jacqueline G. Cooper, Sean P. Car- ter, Stephen A. Cozen, Richard D. Klingler, Jodi Westbrook, and John M. Eubanks; for Retired United States Generals by Glenn A. Danas; and for Sen. Charles E. Grassley by Michael A. Petrino. Briefs of amici curiae were fled for Former State Department Legal Advisers by Elbert Lin and Kevin S. Elliker; for 123 Victims of Terrorist Attacks by Geoffrey P. Eaton and Tejinder Singh; and for 470 Victims of Terrorist Attacks by Daniel Woofter and Erica Oleszczuk Evans. 478 TWITTER, INC. v. TAAMNEH

ism” may sue for damages. § 2333(a). They are not limited to suing the individual terrorists or organizations that di- rectly carried out the attack, however. That is because § 2333(d)(2) also imposes civil liability on “any person who aids and abets, by knowingly providing substantial assist- ance, or who conspires with the person who committed such an act of international terrorism.” Victims of terrorist acts therefore may seek to recover from those who aided and abetted the terrorist act that injured them. The plaintiffs (who are respondents) contend that they have stated a claim for relief under § 2333(d)(2). They were allegedly injured by a terrorist attack carried out by ISIS. But plaintiffs are not suing ISIS. Instead, they have brought suit against three of the largest social-media com- panies in the world—Facebook, Twitter (who is petitioner), and Google (which owns YouTube)—for allegedly aiding and abetting ISIS. As plaintiffs allege, ISIS has used defend- ants' social-media platforms to recruit new terrorists and to raise funds for terrorism. Defendants allegedly knew that ISIS was using their platforms but failed to stop it from doing so. Plaintiffs accordingly seek to hold Facebook, Twitter, and Google liable for the terrorist attack that alleg- edly injured them. We conclude, however, that plaintiffs' al- legations are insuffcient to establish that these defendants aided and abetted ISIS in carrying out the relevant attack. I Plaintiffs' case arises from a 2017 terrorist attack on the Reina nightclub in Istanbul, Turkey. The attack was car- ried out by Abdulkadir Masharipov on behalf of the Islamic State of Iraq and Syria (ISIS).1 Born in Uzbekistan, Mash- aripov had received military training with al Qaeda in Af- ghanistan in 2011 and eventually became affliated with ISIS. 1 In one form or another, ISIS has been designated a Foreign Terrorist Organization since 2004; it has also been known as the Islamic State of Iraq and the Levant, al Qaeda in Iraq, and the al-Zarqawi Network. Cite as: 598 U. S. 471 (2023) 479

In 2016, he was ordered by ISIS to travel to Turkey and launch an attack in Istanbul on New Year's Eve. After planning and coordinating the attack with ISIS emir Abu Shuhada, Masharipov entered the Reina nightclub in the early hours of January 1, 2017, and fred over 120 rounds into a crowd of more than 700 people. In total, Masharipov killed 39 people and injured 69 others. The next day, ISIS released a statement claiming responsibility for the attack. Two weeks later, Masharipov was arrested in Istanbul after hiding out in ISIS safe houses. One of Masharipov's victims was Nawras Alassaf, who was killed in the attack. Several members of Alassaf 's family then brought the present lawsuit under § 2333, alleging that they had been injured by the attack.2 Invoking § 2333(d)(2), plaintiffs sued three major social-media companies—Face- book, Inc., Google, Inc., and Twitter, Inc.—claiming that they aided and abetted ISIS and thus were liable for the Reina nightclub attack.3 As is common knowledge, these three companies control three of the largest and most ubiquitous platforms on the internet: Facebook, YouTube, and Twitter. At the time of the Reina attack, Facebook had over 2 billion active users each month, YouTube had over 1 billion, and Twitter had around 330 million. See Facebook, Inc., Form 10–K for Fis- cal Year Ended Dec. 31, 2017, p. 34; Twitter, Inc., Form 10– K for Fiscal Year Ended Dec. 31, 2017, p. 47; YouTube, You- Tube Hits a Billion Monthly Users (Mar. 21, 2013), https:// blog.youtube/news-and-events/onebillionstrong. At least for

2 Plaintiffs appear to offer several theories as to the nature of their in- jury; because that question is not before us, we take no position as to any of those theories. 3 Although Twitter, Inc., is the named petitioner and defendant, Twitter, Inc., has since been merged into X Corp., a subsidiary of X Holdings Corp. Similarly, although Facebook, Inc., and Google, Inc., are the named defend- ants, Facebook, Inc., is now known as Meta Platforms, Inc., and Google, Inc., is now Google LLC, a subsidiary of Alphabet, Inc. 480 TWITTER, INC. v. TAAMNEH

Facebook and YouTube, those numbers are even higher today. See Meta Platforms, Inc., Form 10–K for Fiscal Year Ended Dec. 31, 2022, p. 56 (nearly 3 billion); L. Ceci, Stat- ista, YouTube–Statistics & Facts (Mar. 22, 2023), https:// www.statista.com/topics/2019/youtube/#topicOverview (2.56 billion). Everyone before us today agrees on the basic aspects of these platforms' business models. People from around the world can sign up for the platforms and start posting content on them, free of charge and without much (if any) advance screening by defendants. Once on the platforms, users can upload messages, videos, and other types of content, which others on the platform can then view, respond to, and share. As noted above, billions of people have done just that. As a result, the amount of content on defendants' platforms is staggering. It appears that for every minute of the day, approximately 500 hours of video are uploaded to YouTube, 510,000 comments are posted on Facebook, and 347,000 tweets are sent on Twitter. See Statista, Media Usage in an Internet Minute as of April 2022 (2023), https://www. statista.com/statistics/195140/new-user-generated-content- uploaded-by-users-per-minute; Statista, YouTube–Statis- tics & Facts; B. Marr, How Much Data Do We Create Every Day? Forbes, May 21, 2018. On YouTube alone, users col- lectively watch more than 1 billion hours of video every day. See YouTube Advertising, Reach Your Customers—and Dis- cover New Ones, https://youtube.com/intl/en_us/ads/how-it- works/set-up-a-campaign/audience. Defendants proft from this content largely by charging third parties to advertise on their platforms. Those adver- tisements are placed on or near the billions of videos, posts, comments, and tweets uploaded by the platforms' users. To organize and present all those advertisements and pieces of content, defendants have developed “recommendation” algo- rithms that automatically match advertisements and content with each user; the algorithms generate those outputs based Cite as: 598 U. S. 471 (2023) 481

on a wide range of information about the user, the advertise- ment, and the content being viewed. So, for example, a per- son who watches cooking shows on YouTube is more likely to see cooking-based videos and advertisements for cookbooks, whereas someone who likes to watch professorial lectures might see collegiate debates and advertisements for TED Talks. But not all of the content on defendants' platforms is so benign. As plaintiffs allege, ISIS and its adherents have used these platforms for years as tools for recruiting, fund- raising, and spreading their propaganda. Like many others around the world, ISIS and its supporters opened accounts on Facebook, YouTube, and Twitter and uploaded videos and messages for others to see. Like most other content on those platforms, ISIS' videos and messages were then matched with other users based on those users' information and use history. And, like most other content, advertise- ments were displayed with ISIS' messages, posts, and videos based on information about the viewer and the content being viewed. Unlike most other content, however, ISIS' videos and messages celebrated terrorism and recruited new terror- ists. For example, ISIS uploaded videos that fundraised for weapons of terror and that showed brutal executions of sol- diers and civilians alike. And plaintiffs allege that these platforms have been crucial to ISIS' growth, allowing it to reach new audiences, gain new members, and spread its mes- sage of terror. Plaintiffs also allege that defendants have known that ISIS has used their platforms for years. Yet, plaintiffs claim that defendants have failed to detect and remove a substantial number of ISIS-related accounts, posts, and videos. (For example, plaintiffs aver that defendants “have failed to im- plement . . . a basic account detection methodology” to pre- vent ISIS supporters from generating multiple accounts on their platforms. App. 150.) Accordingly, plaintiffs assert that defendants aided and abetted ISIS by knowingly allow- 482 TWITTER, INC. v. TAAMNEH

ing ISIS and its supporters to use their platforms and beneft from their “recommendation” algorithms, enabling ISIS to connect with the broader public, fundraise, and radicalize new recruits. And, in the process, defendants allegedly have profted from the advertisements placed on ISIS' tweets, posts, and videos. Plaintiffs also provide a set of allegations specifc to Google. According to plaintiffs, Google has established a system that shares revenue gained from certain advertise- ments on YouTube with users who posted the videos watched with the advertisement. As part of that system, Google allegedly reviews and approves certain videos before Google permits ads to accompany that video. Plaintiffs al- lege that Google has reviewed and approved at least some ISIS videos under that system, thereby sharing some amount of revenue with ISIS. The District Court dismissed plaintiffs' complaint for fail- ure to state a claim.4 But the Ninth Circuit reversed, fnd- ing that plaintiffs had plausibly alleged that defendants aided and abetted ISIS within the meaning of § 2333(d)(2) and thus could be held secondarily liable for the Reina nightclub at- tack. Gonzalez v. Google, 2 F. 4th 871 (2021). We granted certiorari to resolve whether plaintiffs have adequately stated such a claim under § 2333(d)(2). See 598 U. S. ––– (2022). II Section 2333 was originally enacted as part of the Antiter- rorism Act (ATA) in 1990. 104 Stat. 2250. At that time, Congress authorized United States nationals or their “estate, survivors, or heirs” to bring civil lawsuits when “injured in [their] person, property, or business by reason of an act of

4 Plaintiffs also raised other claims, including that defendants were di- rectly liable for having provided material support to ISIS. See, e. g., 18 U. S. C. §§ 2333(a), 2339A, 2339B, 2339C. The District Court dismissed those claims as well, and plaintiffs did not appeal them. Cite as: 598 U. S. 471 (2023) 483

international terrorism.” Id., at 2251.5 In such a lawsuit, the plaintiff could recover treble damages and the cost of the suit, including attorney's fees. See 18 U. S. C. § 2333(a). But the ATA did not explicitly impose liability on anyone who only helped the terrorists carry out the attack or con- spired with them. Prior to 2016, some courts accordingly determined that the ATA did not authorize that sort of sec- ondary civil liability. See, e. g., Rothstein v. UBS AG, 708 F. 3d 82, 97–98 (CA2 2013). Then, in 2016, Congress enacted the Justice Against Spon- sors of Terrorism Act (JASTA) to provide for a form of sec- ondary civil liability. 130 Stat. 852. Thus, as the law now stands, those injured by an act of international terrorism can sue the relevant terrorists directly under § 2333(a)—or they can sue anyone “who aids and abets, by knowingly providing substantial assistance, or who conspires with the person who committed such an act of international terrorism” under § 2333(d)(2). For such a secondary-liability claim, there is an

5 The ATA defnes “international terrorism” to mean “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 infuence the policy of a government by intimidation or coer- cion; or “(iii) to affect the conduct of a government by mass destruction, assassi- nation, 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.” § 2331(1). In short, the act generally must be violent, criminal, intended to in- timidate or coerce civilians or a government, and occur either primarily outside the United States or transcending national boundaries. See ibid. 484 TWITTER, INC. v. TAAMNEH

additional condition: The “act of international terrorism” must have been “committed, planned, or authorized by an organization that had been designated as a foreign terrorist organization under [8 U. S. C. § 1189] as of the date on which such act of international terrorism was committed, planned, or authorized.” Ibid. Plaintiffs seeking secondary liability can likewise recover treble damages and the cost of the suit, including attorney's fees. See §§ 2333(a), (d)(2). The parties here do not dispute that the frst three compo- nents of § 2333(d)(2) have been adequately alleged: The Reina nightclub attack was an “act of international terrorism”; the attack was “committed, planned, or authorized” by ISIS; and ISIS was “designated as a foreign terrorist organization” as of the date of the Reina nightclub attack. § 2333(d)(2). The central question is thus whether defendants' conduct consti- tutes “aid[ing] and abett[ing], by knowingly providing sub- stantial assistance,” such that they can be held liable for the Reina nightclub attack. Ibid. III As always, we start with the text of § 2333. See Barten- werfer v. Buckley, 598 U. S. 69, 74 (2023). Here, that text immediately begs two questions: First, what exactly does it mean to “aid and abet”? Second, what precisely must the defendant have “aided and abetted”?

A We turn frst to the meaning of the phrase “aids and abets, by knowingly providing substantial assistance.” Nothing in the statute defnes any of those critical terms. Yet terms like “aids and abets” are familiar to the common law, which has long held aiders-and-abettors secondarily liable for the wrongful acts of others. See Central Bank of Denver, N. A. v. First Interstate Bank of Denver, N. A., 511 U. S. 164, 181 (1994). We generally presume that such common-law terms “brin[g] the old soil” with them. Sekhar v. United States, Cite as: 598 U. S. 471 (2023) 485

570 U. S. 729, 733 (2013) (internal quotation marks omitted). In enacting JASTA, Congress provided additional context by pointing to Halberstam v. Welch, 705 F. 2d 472 (CADC 1983), as “provid[ing] the proper legal framework” for “civil aiding and abetting and conspiracy liability.” § 2(a)(5), 130 Stat. 852.6 We thus begin with Halberstam's “legal framework,” viewed in context of the common-law tradition from which it arose. 1 Long regarded as a leading case on civil aiding-and-abet- ting and conspiracy liability, see 130 Stat. 852, Halberstam arose from a distinctive fact pattern. Bernard Welch was a serial burglar who had killed Michael Halberstam during a break-in. 705 F. 2d, at 474. Halberstam's estate then sued Welch's live-in partner, Linda Hamilton, for aiding and abet- ting and conspiring with Welch. Id., at 474, 476.7 Hamilton was not present for Halberstam's murder, or even allegedly aware of the murder. See id., at 474–476. But the facts made clear that “[s]he was a willing partner in [Welch's] criminal activities.” Id., at 474 (internal quotation marks omitted). Hamilton had lived with Welch for fve years, dur- ing which time the couple had risen from modest circum- stances to possess a substantial fortune. Ibid. This rapid ascent was remarkable because Welch had no outside em- ployment. Id., at 475. Rather, he left the house most eve- nings and returned with antiques, jewelry, and precious met-

6 The provision reads in full: “The decision of the United States Court of Appeals for the District of Columbia in Halberstam v. Welch, 705 F. 2d 472 (D. C. Cir. 1983), which has been widely recognized as the leading case regarding Federal civil aiding and abetting and conspiracy liability, including by the Supreme Court of the United States, provides the proper legal framework for how such liability should function in the context of chapter 113B of title 18, United States Code.” § 2(a)(5), 130 Stat. 852. 7 Halberstam's estate also sued Welch himself; a default judgment was entered against Welch, who did not appeal. See Halberstam, 705 F. 2d, at 474. 486 TWITTER, INC. v. TAAMNEH

als—some of which he melted down into gold and silver ingots by using a smelting furnace that he had installed in their garage. Meanwhile, Hamilton did bookkeeping work for Welch's “business,” facilitating the sale of those stolen goods. Ibid. She had Welch's customers make checks pay- able to her, falsifed her tax returns at Welch's direction, and kept records of incoming payments from Welch's custom- ers—with no records of outgoing funds to his “suppliers.” Ibid. Their arrangement continued until Welch was ar- rested after he killed Halberstam while burglarizing Halber- stam's home. Ibid. To determine Hamilton's liability, the D. C. Circuit under- took an extensive survey of the common law, examining a series of state and federal cases, the Restatement (Second) of Torts, and prominent treatises that discussed secondary liability in tort. Id., at 476–478, 481–486. With respect to aiding and abetting, the court synthesized the cases as rest- ing on three main elements: First, “the party whom the de- fendant aids must perform a wrongful act that causes an injury. ” Id., at 477. Second, “ the defendant must be generally aware of his role as part of an overall illegal or tortious activity at the time that he provides the assistance.” Ibid. And, third, “the defendant must knowingly and sub- stantially assist the principal violation.” Ibid. (citing, e. g., Landy v. Federal Deposit Ins. Corp., 486 F. 2d 139, 162–163 (CA3 1973); Woodward v. Metro Bank of Dallas, 522 F. 2d 84, 94–95 (CA5 1975)). Halberstam then articulated six fac- tors to help determine whether a defendant's assistance was “substantial.” 705 F. 2d, at 486–488. Those factors are (1) “the nature of the act assisted,” (2) the “amount of assist- ance” provided, (3) whether the defendant was “present at the time” of the principal tort, (4) the defendant's “relation to the tortious actor,” (5) the “defendant's state of mind,” and (6) the “duration of the assistance” given. Id., at 488 (emphasis deleted). Last, Halberstam clarifed that those who aid and abet “a tortious act may be liable” not only for Cite as: 598 U. S. 471 (2023) 487

the act itself but also “for other reasonably foreseeable acts done in connection with it.” Id., at 484. Applying that framework, the D. C. Circuit held that Ham- ilton was liable for aiding and abetting Halberstam's murder. Id., at 489.8 The court frst determined that Welch had com- mitted a wrong (in killing Halberstam during the burglary) and that Hamilton was generally aware of her role in Welch's criminal enterprise. Id., at 488. It then explained that Hamilton had given knowing and substantial assistance to Welch's activities by helping him turn his “stolen goods into `legitimate' wealth,” thereby intending to help Welch suc- ceed by performing a function crucial to any thief. Ibid. And it clarifed that Hamilton knew Welch was commit- ting some sort of “personal property crime,” the “foresee- able risk” of which was “violence and killing.” Ibid. The court therefore concluded that Hamilton substantially helped Welch commit personal property crimes and was liable for Halberstam's death, which was a foreseeable result of such crimes. Ibid. That articulation of the common law thus resolved Halber- stam. But Halberstam recognized that the elements and factors it provided could “be merged or articulated some- what differently without affecting their basic thrust.” Id., at 478, n. 8. It thus cautioned—in a typical common-law fashion—that its formulations should “not be accepted as im- mutable components.” Id., at 489. Rather, Halberstam suggested that its framework should be “adapted as new cases test their usefulness in evaluating vicarious liability.” Ibid. 2 The allegations before us today are a far cry from the facts of Halberstam. Rather than dealing with a serial burglar and his live-in partner-in-crime, we are faced with interna-

8 The D. C. Circuit also held that Hamilton was liable as a co-conspirator. Id., at 489. 488 TWITTER, INC. v. TAAMNEH

tional terrorist networks and world-spanning internet plat- forms. By Halberstam's own lights, its precise three- element and six-factor test thus may not be entirely ade- quate to resolve these new facts. Ibid. And JASTA itself points only to Halberstam's “framework,” not its facts or its exact phrasings and formulations, as the benchmark for aid- ing and abetting. § 2(a)(5), 130 Stat. 852. We therefore must ascertain the “basic thrust” of Halberstam's elements and determine how to “adap[t]” its framework to the facts before us today. See 705 F. 2d, at 478, 489, and n. 8. To do so, we turn to the common law of aiding and abetting upon which Halberstam rested and to which JASTA's common-law terminology points. As we have recognized, “[a]iding and abetting is an ancient criminal law doctrine” that has substantially infuenced its analog in tort. Central Bank of Denver, 511 U. S., at 181. In one early statement of the criminal-law doctrine, William Blackstone explained that those who were “present, aiding and abetting the fact to be done,” or “procure[d], counsel[ed], or command[ed] another to commit a crime,” were guilty and punishable. 4 Commentaries on the Laws of England 34, 36 (1795). Over the years, many statutes and courts have of- fered variations on that basic rule. See United States v. Peoni, 100 F. 2d 401, 402 (CA2 1938) (L. Hand, J., for the court) (collecting authorities). Yet, to this day, the basic “view of culpability” that animates the doctrine is straight- forward: “[A] person may be responsible for a crime he has not personally carried out if he helps another to complete its commission.” Rosemond v. United States, 572 U. S. 65, 70 (2014). Importantly, the concept of “helping” in the commission of a crime—or a tort—has never been boundless. That is because, if it were, aiding-and-abetting liability could sweep in innocent bystanders as well as those who gave only tan- gential assistance. For example, assume that any assistance of any kind were suffcient to create liability. If that were Cite as: 598 U. S. 471 (2023) 489

the case, then anyone who passively watched a robbery could be said to commit aiding and abetting by failing to call the police. Yet, our legal system generally does not impose lia- bility for mere omissions, inactions, or nonfeasance; although inaction can be culpable in the face of some independent duty to act, the law does not impose a generalized duty to rescue. See 1 W. LaFave, Substantive Criminal Law § 6.1 (3d ed. 2018) (LaFave); W. Keeton, D. Dobbs, R. Keeton, & D. Owen, Prosser and Keeton on Law of Torts 373–375 (5th ed. 1984) (Prosser & Keeton). Moreover, both criminal and tort law typically sanction only “wrongful conduct,” bad acts, and misfeasance. J. Goldberg, A. Sebok, & B. Zipursky, Tort Law: Responsibilities and Redress 31 (2004). Some level of blameworthiness is therefore ordinarily required. But, again, if aiding-and-abetting liability were taken too far, then ordinary merchants could become liable for any misuse of their goods and services, no matter how attenuated their re- lationship with the wrongdoer. And those who merely de- liver mail or transmit emails could be liable for the tortious messages contained therein. See Restatement (Second) of Torts § 876, Comment d, Illus. 9, p. 318 (1979) (cautioning against this result). For these reasons, courts have long recognized the need to cabin aiding-and-abetting liability to cases of truly culpable conduct. They have cautioned, for example, that not “all those present at the commission of a trespass are liable as principals” merely because they “make no opposition or man- ifest no disapprobation of the wrongful” acts of another. Brown v. Perkins, 83 Mass. 89, 98 (1861); see also Hilmes v. Stroebel, 59 Wis. 74, 17 N. W. 539 (1883); Duke v. Feldman, 245 Md. 454, 457–458, 226 A. 2d 345, 347 (1967). Put another way, overly broad liability would allow for “one person [to] be made a trespasser and even a felon against his or her consent, and by the mere rashness or precipitancy or over- heated zeal of another.” Bird v. Lynn, 49 Ky. 422, 423 (1850). Moreover, unlike its close cousin conspiracy, aiding 490 TWITTER, INC. v. TAAMNEH

and abetting does not require any agreement with the pri- mary wrongdoer to commit wrongful acts, thus eliminating a signifcant limiting principle. See Nye & Nissen v. United States, 336 U. S. 613, 620 (1949). To keep aiding-and-abetting liability grounded in culpable misconduct, criminal law thus requires “that a defendant `in some sort associate himself with the venture, that he partici- pate in it as in something that he wishes to bring about, that he seek by his action to make it succeed' ” before he could be held liable. Id., at 619 (quoting Peoni, 100 F. 2d, at 402). In other words, the defendant has to take some “affrmative act” “with the intent of facilitating the offense's commission.” Rosemond, 572 U. S., at 71. Such intentional participation can come in many forms, including abetting, inducing, en- couraging, soliciting, or advising the commission of the of- fense, such as through words of encouragement or driving the getaway car. 2 LaFave § 13.2(a), at 457–460; see also J. Hawley & M. McGregor, The Criminal Law 81 (3d ed. 1899). Regardless of the particulars, however, it is clear that some culpable conduct is needed. See Rosemond, 572 U. S., at 73, 77.9 Similar principles and concerns have shaped aiding-and- abetting doctrine in tort law, with numerous cases directly employing them to help articulate the standard for tortious aiding and abetting. See, e. g., Zoelsch v. Arthur Ander- sen & Co., 824 F. 2d 27, 35–36 (CADC 1987); Woodward, 522 F. 2d, at 95, n. 23; Landy, 486 F. 2d, at 163–164 (all relying on Nye & Nissen, 336 U. S., at 619); see also Camp v. Dema, 948 F. 2d 455, 459 (CA8 1991) (relying on People v. Terman, 4 Cal. App. 2d 345, 346–347 (1935), another criminal case).

9 Conversely, conspiracy liability could be premised on a “more attenu- ated relation with the principal violation” because the defendant and the principal wrongdoer had agreed to a wrongful enterprise. Halberstam, 705 F. 2d, at 485; see also Restatement (Second) of Torts § 876, Comment a, p. 316 (1979) (“The theory of the early common law was that there was a mutual agency of each to act for the others”). Cite as: 598 U. S. 471 (2023) 491

Similar to the criminal-law rule, some cases have required that the defendant's assistance “must have had a direct rela- tion to the trespass, and have been calculated and intended to produce it” to warrant liability for the resulting tort. Bird, 49 Ky., at 423; see also Smith v. Thompson, 103 Idaho 909, 911, 655 P. 2d 116, 118 (App. 1982); Brown, 83 Mass., at 98. Other cases have emphasized the need for some “culpa- ble conduct” and “some degree of knowledge that [a defend- ant's] actions are aiding the primary violator” before holding the defendant secondarily liable. Camp, 948 F. 2d, at 460. Still others have explained that “[c]ulpability of some sort is necessary to justify punishment of a secondary actor,” lest mostly passive actors like banks become liable for all of their customers' crimes by virtue of carrying out routine transac- tions. Monsen v. Consolidated Dressed Beef Co., 579 F. 2d 793, 799 (CA3 1978). And others have suggested that “inac- tion cannot create liability as an aider and abettor” absent a duty to act. Zoelsch, 824 F. 2d, at 36; see also Woodward, 522 F. 2d, at 96. In articulating those limits, courts simultaneously began to crystalize the framework for aiding and abetting that Halberstam identifed and applied. See, e. g., Monsen, 579 F. 2d, at 799 (stating a similar three-part test).10 As in Halberstam, that framework generally required what the text of § 2333(d)(2) demands: that the defendant have given knowing and substantial assistance to the primary tortfea- sor. See, e. g., Monsen, 579 F. 2d, at 799; Landy, 486 F. 2d, at 162–163. Notably, courts often viewed those twin re- quirements as working in tandem, with a lesser showing of

10 Others, however, have wondered whether any of these “elaborate dis- cussions of the aiding and abetting standard . . . `have added anything except unnecessary detail' ” to the formulation set forth by Judge Learned Hand in United States v. Peoni, 100 F. 2d 401 (CA2 1938), and adopted by this Court in Nye & Nissen v. United States, 336 U. S. 613 (1949). SEC v. Apuzzo, 689 F. 3d 204, 212, n. 9 (CA2 2012) (quoting IIT, an Int'l Inv. Trust v. Cornfeld, 619 F. 2d 909, 922 (CA2 1980)). 492 TWITTER, INC. v. TAAMNEH

one demanding a greater showing of the other. E. g., Wood- ward, 522 F. 2d, at 97; Woods v. Barnette Bank of Ft. Lau- derdale, 765 F. 2d 1004, 1010 (CA11 1985). In other words, less substantial assistance required more scienter before a court could infer conscious and culpable assistance. See Woodward, 522 F. 2d, at 97. And, vice versa, if the assist- ance were direct and extraordinary, then a court might more readily infer conscious participation in the underlying tort. See ibid. In moving back and forth between all these guide- posts, the courts thus largely tracked the same distinctions drawn above to ensure that liability fell only on those who had abetted the underlying tort through conscious, “culpable conduct.” Camp, 948 F. 2d, at 460.

3 Halberstam's framework refected and distilled those common-law principles. Indeed, Halberstam started with a survey of many earlier common-law cases, including many of the same cases cited above. 705 F. 2d, at 477, 483 (citing, e. g., Woodward, 522 F. 2d, at 94–95; Landy, 486 F. 2d, at 162– 163; Duke, 245 Md., at 457–458, 226 A. 2d, at 347). As part of that survey, Halberstam explicitly distinguished different types of aid along the same culpability axis that grounded the common law. 705 F. 2d, at 481– 483. For example, Halberstam recognized that giving verbal encouragement (such as yelling “ `Kill him!' ”) could be substantial assistance, id., at 481, but that passively watching an assault after hear- ing an assailant threaten the victim likely would not be, id., at 483. Those same lines have long been drawn for aiding- and-abetting liability under the common law. See Rose- mond, 572 U. S., at 70; see also, e. g., 2 LaFave § 13.2(a), at 457–460; Restatement (Second) of Torts § 876, Comment d, Illus. 9–10, at 318; Brown, 83 Mass., at 98. And Halbers- tam's six factors for “substantial assistance” call for the same balancing that courts had undertaken previously between the nature and amount of assistance on the one hand and Cite as: 598 U. S. 471 (2023) 493

the defendant's scienter on the other. 705 F. 2d, at 484–485, 487–488. Despite that deep-rooted common-law basis, the Ninth Cir- cuit appears to have understood JASTA's approval of Halber- stam's “legal framework” as requiring it to hew tightly to the precise formulations that Halberstam used. The parties before us similarly make a conscious effort to draw analogies to the facts of that case. But any approach that too rigidly focuses on Halberstam's facts or its exact phraseology risks missing the mark. Halberstam is by its own terms a common-law case and provided its elements and factors as a way to synthesize the common-law approach to aiding and abetting. And JASTA employs the common-law terms “aids and abets,” pointing to Halberstam's common-law “frame- work” as the primary guidepost for understanding the scope of § 2333(d)(2). At bottom, both JASTA and Halberstam's elements and factors rest on the same conceptual core that has animated aiding-and-abetting liability for centuries: that the defendant consciously and culpably “participate[d]” in a wrongful act so as to help “make it succeed.” Nye & Nissen, 336 U. S., at 619. To be sure, nuances may establish daylight between the rules for aiding and abetting in criminal and tort law; we have described the doctrines as “rough[ly] simila[r],” not identical. Central Bank of Denver, 511 U. S., at 181. But we need not resolve the extent of those differences today; it is enough for our purposes to recognize the framework that Halberstam set forth and the basis on which it rests. The phrase “aids and abets” in § 2333(d)(2), as elsewhere, refers to a conscious, voluntary, and culpable participation in anoth- er's wrongdoing. B The next question, however, is what precisely a defendant must aid and abet. As noted above, JASTA imposes liability on anyone “who aids and abets, by knowingly providing substantial assistance, or who conspires with the person 494 TWITTER, INC. v. TAAMNEH

who committed such an act of international terrorism.” § 2333(d)(2). The parties vigorously dispute the textual ob- ject of the term “aids and abets”: Plaintiffs assert that it is “the person,” and defendants insist that it is the “act of international terrorism.” So, plaintiffs contend, defendants can be liable if they aided and abetted ISIS generally—there is no need for defendants to have aided and abetted the spe- cifc Reina nightclub attack. Conversely, defendants posit that they are liable only if they directly aided and abetted the Reina nightclub attack, with a strict nexus between their assistance and that attack. Neither side is quite right. To start, we fnd it unnecessary to parse whether the tex- tual object of “aids and abets” is “the person” or the “act of international terrorism.” That syntactic dispute makes lit- tle difference here, because aiding and abetting is inherently a rule of secondary liability for specifc wrongful acts. See Prosser & Keeton 323. As stated above, the rule imposes liability for a wrong on those who “hel[p] another to complete its commission.” Rosemond, 572 U. S., at 70 (emphasis added). Or, as Halberstam put it, the defendant must aid and abet “a tortious act.” 705 F. 2d, at 484. Nor would a contrary rule make sense for torts. That is because tort law imposes liability only when someone com- mits an actual tort; merely agreeing to commit a tort or suggesting a tortious act is not, without more, tortious. See Prosser & Keeton 324; Halberstam, 705 F. 2d, at 479.11 “Enterprises” or “conspiracies” alone are therefore not tor- tious—the focus must remain on the tort itself. The same

11 In this respect, tort law is different from criminal law, which does punish mere agreements to commit crimes. See, e. g., 2 LaFave § 12.2(b), at 372 (“At common law a conspiracy was punishable even though no act was done beyond the mere making of the agreement”); see also 1 J. Ohlin, Wharton's Criminal Law § 8.7, p. 242 (16th ed. 2021) (noting that statutory requirements of an “overt act” generally do not require that the overt act be criminal); Iannelli v. United States, 420 U. S. 770, 785, n. 17 (1975) (similar). Cite as: 598 U. S. 471 (2023) 495

is true here: The ATA opens the courthouse doors only if the plaintiff is “injured . . . by reason of an act of international terrorism.” § 2333(a). JASTA further restricts secondary liability by requiring that the “act of international terror- ism” be “committed, planned, or authorized by” a foreign ter- rorist organization designated as such “as of the date on which such act of international terrorism was committed, planned, or authorized.” § 2333(d). Thus, it is not enough, as plaintiffs contend, that a defendant have given substantial assistance to a transcendent “enterprise” separate from and foating above all the actionable wrongs that constitute it. Rather, a defendant must have aided and abetted (by know- ingly providing substantial assistance) another person in the commission of the actionable wrong—here, an act of interna- tional terrorism. See, e. g., Restatement (Second) of Torts § 876(b), and Comment d; Halberstam, 705 F. 2d, at 488. Plaintiffs insist that Halberstam proves the contrary, but their argument misses the gist of that case. To be sure, Linda Hamilton was not on the scene for the burglary of Halberstam's house and did not lend any specifc support to Halberstam's murder. Ibid. But Hamilton's assistance to Welch was so intentional and systematic that she assisted each and every burglary committed by Welch; any time that Welch left the house to burglarize, he would have relied on Hamilton's assistance in laundering the stolen goods and transforming them into usable wealth. See ibid. Thus, Hamilton did aid and abet Welch in burglarizing Halbers- tam's home—and, as noted above, killing Halberstam was a foreseeable consequence of that burglary. See ibid. On the other hand, defendants overstate the nexus that § 2333(d)(2) requires between the alleged assistance and the wrongful act. To start, aiding and abetting does not require the defendant to have known “all particulars of the primary actor's plan.” Restatement (Third) of Torts: Intentional Torts to Persons § 10, Comment c, p. 104 (Tent. Draft No. 3, Apr. 6, 2018). For example, a defendant might be held liable 496 TWITTER, INC. v. TAAMNEH

for aiding and abetting the burning of a building if he inten- tionally helped others break into the building at night and then, unknown to him, the others lit torches to guide them through the dark and accidentally started a fre. See Amer- ican Family Mutual Ins. Co. v. Grim, 201 Kan. 340, 345–347, 440 P. 2d 621, 625–626 (1968); Restatement (Second) of Torts § 876, Comment d, Illus. 10, at 318. That leads to the next problem: As Halberstam makes clear, people who aid and abet a tort can be held liable for other torts that were “a foreseeable risk” of the intended tort. 705 F. 2d, at 488. Accordingly, a close nexus between the assistance and the tort might help establish that the defendant aided and abet- ted the tort, but even more remote support can still consti- tute aiding and abetting in the right case. Moreover, in appropriate circumstances, a secondary de- fendant's role in an illicit enterprise can be so systemic that the secondary defendant is aiding and abetting every wrong- ful act committed by that enterprise—as in Halberstam it- self. Id., at 487–488. At this point, aiding-and-abetting lia- bility begins to blur with conspiracy liability, which typically holds co-conspirators liable for all reasonably foreseeable acts taken to further the conspiracy. See Pinkerton v. United States, 328 U. S. 640, 647–648 (1946); see also Halber- stam, 705 F. 2d, at 482–483 (noting the occasional overlap). Yet, as noted above, aiding and abetting lacks the requisite agreement that justifes such extensive conspiracy liability. See Restatement (Second) of Torts § 876, Comment a, at 316 (“The theory of the early common law was that there was mutual agency of each [conspirator] to act for the others”); Pinkerton, 328 U. S., at 646. Thus, while the facts of Hal- berstam are not totemic (lest courts strain to compare Linda Hamilton with international criminal syndicates), its facts are useful when determining whether a defendant has so consciously “participate[d] in” a series of tortious acts in order to “make [each one] succeed.” Nye & Nissen, 336 U. S., at 619 (internal quotation marks omitted). Cite as: 598 U. S. 471 (2023) 497

* * * To summarize the requirements of § 2333(d)(2), the phrase “aids and abets, by knowingly providing substantial as- sistance,” points to the elements and factors articulated by Halberstam. But, those elements and factors should not be taken as infexible codes; rather, they should be understood in light of the common law and applied as a framework de- signed to hold defendants liable when they consciously and culpably “participate[d] in” a tortious act in such a way as to help “make it succeed.” Nye & Nissen, 336 U. S., at 619 (internal quotation marks omitted). And the text requires that defendants have aided and abetted the act of interna- tional terrorism that injured the plaintiffs—though that re- quirement does not always demand a strict nexus between the alleged assistance and the terrorist act.

IV Page Proof Under the Pending appropriate framework, somePublication aspects of today's case become immediately clear: First, because they are try- ing to hold defendants liable for the Reina attack, plaintiffs must plausibly allege that defendants aided and abetted ISIS12 in carrying out that attack. Next, plaintiffs have sat- isfed Halberstam's frst two elements by alleging both that ISIS committed a wrong and that defendants knew they were playing some sort of role in ISIS' enterprise. The key question, therefore, is whether defendants gave such know- ing and substantial assistance to ISIS that they culpably par- ticipated in the Reina attack. The allegations here fall short of that showing under Halberstam's framework as properly 12 For purposes of this case, we need not resolve whether defendants must have aided and abetted ISIS, Masharipov, or some subgroup of ISIS operatives in committing the Reina attack. In other words, we need not resolve whether “the person” referred to in § 2333(d)(2) encompasses inter- national terrorist syndicates or is somehow otherwise limited; either way, defendants need to have aided and abetted that “person” in carrying out the Reina attack. 498 TWITTER, INC. v. TAAMNEH

understood by reference to the common-law principles it applied. A To start, recall the basic ways that defendants as a group allegedly helped ISIS. First, ISIS was active on defend- ants' social-media platforms, which are generally available to the internet-using public with little to no front-end screening by defendants. In other words, ISIS was able to upload con- tent to the platforms and connect with third parties, just like everyone else. Second, defendants' recommendation algo- rithms matched ISIS-related content to users most likely to be interested in that content—again, just like any other con- tent. And, third, defendants allegedly knew that ISIS was uploading this content to such effect, but took insuffcient steps to ensure that ISIS supporters and ISIS-related con- tent were removed from their platforms. Notably, plaintiffs never allege that ISIS used defendants' platforms to plan or coordinate the Reina attack; in fact, they do not allege that Masharipov himself ever used Facebook, YouTube, or Twitter. None of those allegations suggest that defendants cul- pably “associate[d themselves] with” the Reina attack, “par- ticipate[d] in it as something that [they] wishe[d] to bring about,” or sought “by [their] action to make it succeed.” Nye & Nissen, 336 U. S., at 619 (internal quotation marks ommitted). In part, that is because the only affrmative “conduct” defendants allegedly undertook was creating their platforms and setting up their algorithms to display content relevant to user inputs and user history. Plaintiffs never allege that, after defendants established their platforms, they gave ISIS any special treatment or words of encourage- ment. Nor is there reason to think that defendants selected or took any action at all with respect to ISIS' content (ex- cept, perhaps, blocking some of it).13 Indeed, there is not 13 Plaintiffs concede that defendants attempted to remove at least some ISIS-sponsored accounts and content after they were brought to their attention. Cite as: 598 U. S. 471 (2023) 499

even reason to think that defendants carefully screened any content before allowing users to upload it onto their plat- forms. If anything, the opposite is true: By plaintiffs' own allegations, these platforms appear to transmit most content without inspecting it. The mere creation of those platforms, however, is not cul- pable. To be sure, it might be that bad actors like ISIS are able to use platforms like defendants' for illegal—and some- times terrible—ends. But the same could be said of cell phones, email, or the internet generally. Yet, we generally do not think that internet or cell service providers incur cul- pability merely for providing their services to the public writ large. Nor do we think that such providers would normally be described as aiding and abetting, for example, illegal drug deals brokered over cell phones—even if the provider 's conference-call or video-call features made the sale easier. See Doe v. GTE Corp., 347 F. 3d 655, 659 (CA7 2003). To be sure, plaintiffs assert that defendants' “recommen- dation” algorithms go beyond passive aid and constitute ac- tive, substantial assistance. We disagree. By plaintiffs' own telling, their claim is based on defendants' “provision of the infrastructure which provides material support to ISIS.” App. 53. Viewed properly, defendants' “recommendation” algorithms are merely part of that infrastructure. All the content on their platforms is fltered through these algo- rithms, which allegedly sort the content by information and inputs provided by users and found in the content itself. As presented here, the algorithms appear agnostic as to the na- ture of the content, matching any content (including ISIS' content) with any user who is more likely to view that con- tent. The fact that these algorithms matched some ISIS content with some users thus does not convert defendants' passive assistance into active abetting. Once the platform and sorting-tool algorithms were up and running, defendants at most allegedly stood back and watched; they are not al- leged to have taken any further action with respect to ISIS. 500 TWITTER, INC. v. TAAMNEH

At bottom, then, the claim here rests less on affrmative misconduct and more on an alleged failure to stop ISIS from using these platforms. But, as noted above, both tort and criminal law have long been leery of imposing aiding-and- abetting liability for mere passive nonfeasance. To show that defendants' failure to stop ISIS from using these plat- forms is somehow culpable with respect to the Reina attack, a strong showing of assistance and scienter would thus be required. Plaintiffs have not made that showing. First, the relationship between defendants and the Reina attack is highly attenuated. As noted above, defendants' platforms are global in scale and allow hundreds of millions (or billions) of people to upload vast quantities of information on a daily basis. Yet, there are no allegations that defend- ants treated ISIS any differently from anyone else. Rather, defendants' relationship with ISIS and its supporters ap- pears to have been the same as their relationship with their billion-plus other users: arm's length, passive, and largely indifferent. Cf. Halberstam, 705 F. 2d, at 488. And their relationship with the Reina attack is even further removed, given the lack of allegations connecting the Reina attack with ISIS' use of these platforms. Second, because of the distance between defendants' acts (or failures to act) and the Reina attack, plaintiffs would need some other very good reason to think that defendants were consciously trying to help or otherwise “participate in” the Reina attack. Nye & Nissen, 336 U. S., at 619 (internal quo- tation marks omitted). But they have offered no such rea- son, let alone a good one. Again, plaintiffs point to no act of encouraging, soliciting, or advising the commission of the Reina attack that would normally support an aiding-and- abetting claim. See 2 LaFave § 13.2(a), at 457. Rather, they essentially portray defendants as bystanders, watching passively as ISIS carried out its nefarious schemes. Such allegations do not state a claim for culpable assistance or Cite as: 598 U. S. 471 (2023) 501

participation in the Reina attack. See Halberstam, 705 F. 2d, at 481, 483. Because plaintiffs' complaint rests so heavily on defend- ants' failure to act, their claims might have more purchase if they could identify some independent duty in tort that would have required defendants to remove ISIS' content. See Woodward, 522 F. 2d, at 97, 100. But plaintiffs identify no duty that would require defendants or other communication- providing services to terminate customers after discovering that the customers were using the service for illicit ends. See Doe, 347 F. 3d, at 659; People v. Brophy, 49 Cal. App. 2d 15, 33–34 (1942).14 To be sure, there may be situations where some such duty exists, and we need not resolve the issue today. Even if there were such a duty here, it would not transform defendants' distant inaction into knowing and substantial assistance that could establish aiding and abet- ting the Reina attack. If there were any doubt, the expansive scope of plaintiffs' claims would put it to rest. Given the lack of any concrete nexus between defendants' services and the Reina attack, plaintiffs' claims would necessarily hold defendants liable as having aided and abetted each and every ISIS terrorist act committed anywhere in the world. Under plaintiffs' theory, any U. S. national victimized by an ISIS attack could bring the same claim based on the same services allegedly pro- vided to ISIS. Plaintiffs thus must allege that defendants so systemically and pervasively assisted ISIS that defend- ants could be said to aid and abet every single ISIS attack. 14 Plaintiffs have not presented any case holding such a company liable for merely failing to block such criminals despite knowing that they used the company's services. Rather, when legislatures have wanted to impose a duty to remove content on these types of entities, they have apparently done so by statute. See, e. g., Telecommunications Act of 1996, § 502, 110 Stat. 133–134 (codifed, as amended, at 47 U. S. C. § 223); but see Reno v. American Civil Liberties Union, 521 U. S. 844, 857, 874 (1997) (holding parts of § 223 unconstitutional under the First Amendment). 502 TWITTER, INC. v. TAAMNEH

Viewed in that light, the allegations here certainly fall short. Plaintiffs do not claim that defendants intentionally associ- ated themselves with ISIS' operations or affrmatively gave aid that would assist each of ISIS' terrorist acts. Nor have they alleged that defendants and ISIS formed a near- common enterprise of the kind that could establish such broad liability. These allegations are thus a far cry from the type of pervasive, systemic, and culpable assistance to a se- ries of terrorist activities that could be described as aiding and abetting each terrorist act. To be sure, we cannot rule out the possibility that some set of allegations involving aid to a known terrorist group would justify holding a secondary defendant liable for all of the group's actions or perhaps some defnable subset of ter- rorist acts. There may be, for example, situations where the provider of routine services does so in an unusual way or provides such dangerous wares that selling those goods to a terrorist group could constitute aiding and abetting a foreseeable terror attack. Cf. Direct Sales Co. v. United States, 319 U. S. 703, 707, 711–712, 714–715 (1943) (registered morphine distributor could be liable as a co-conspirator of an illicit operation to which it mailed morphine far in excess of normal amounts). Or, if a platform consciously and selec- tively chose to promote content provided by a particular ter- rorist group, perhaps it could be said to have culpably as- sisted the terrorist group. Cf. Passaic Daily News v. Blair, 63 N. J. 474, 487–488, 308 A. 2d 649, 656 (1973) (publishing employment advertisements that discriminate on the basis of sex could aid and abet the discrimination). In those cases, the defendants would arguably have of- fered aid that is more direct, active, and substantial than what we review here; in such cases, plaintiffs might be able to establish liability with a lesser showing of scienter. But we need not consider every iteration on this theme. In this case, it is enough that there is no allegation that the plat- forms here do more than transmit information by billions of Cite as: 598 U. S. 471 (2023) 503

people, most of whom use the platforms for interactions that once took place via mail, on the phone, or in public areas. The fact that some bad actors took advantage of these plat- forms is insuffcient to state a claim that defendants know- ingly gave substantial assistance and thereby aided and abet- ted those wrongdoers' acts. And that is particularly true because a contrary holding would effectively hold any sort of communication provider liable for any sort of wrongdoing merely for knowing that the wrongdoers were using its serv- ices and failing to stop them. That conclusion would run roughshod over the typical limits on tort liability and take aiding and abetting far beyond its essential culpability moorings. B In holding that plaintiffs had stated a claim, the Ninth Cir- cuit went astray through a series of missteps that, together, obscured the essence of aiding-and-abetting liability. To correct those errors, we proceed through the Ninth Circuit's application of Halberstam's framework. The Ninth Circuit framed the issue of substantial assist- ance as turning on defendants' assistance to ISIS' activities in general. See 2 F. 4th, at 909. But, as we explained above, the question is whether defendants gave substantial assistance to ISIS with respect to the Reina attack. The focus thus must remain on the Reina attack; plaintiffs' failure to allege any defnable nexus between the defendants' assist- ance and that attack therefore—at minimum—drastically in- creases their burden to show that defendants somehow con- sciously and culpably assisted the attack. Next, the Ninth Circuit misapplied the “knowing” half of “knowing and substantial assistance.” It frst separated the “knowing” and “substantial” subelements; it then analyzed the “knowing” subelement as a carbon copy of the ante- cedent element of whether the defendants were “generally aware” of their role in ISIS' overall scheme. Ibid.; see also id., at 908. But, as discussed above, “the knowledge and 504 TWITTER, INC. v. TAAMNEH

substantial assistance” components “should be considered relative to one another” as part of a single inquiry designed to capture conscious and culpable conduct. Camp, 948 F. 2d, at 459 (internal quotation marks omitted). The “knowing” part of that inquiry is therefore designed to capture the de- fendants' state of mind with respect to their actions and the tortious conduct (even if not always the particular terrorist act), not the same general awareness that defnes Halber- stam's second element. Finally, when applying Halberstam's six substantiality fac- tors, the Ninth Circuit appears to have regarded the factors as a sequence of disparate, unrelated considerations without a common conceptual core. See 2 F. 4th, at 909–910. That is incorrect. The point of those factors is to help courts cap- ture the essence of aiding and abetting: participation in another's wrongdoing that is both signifcant and culpable enough to justify attributing the principal wrongdoing to the aider and abettor. The Ninth Circuit thus erred in focusing (as it did) primarily on the value of defendants' platforms to ISIS, rather than whether defendants culpably associated themselves with ISIS' actions. For example, when applying the second factor (the amount and kind of assistance), the Ninth Circuit should have considered that defendants' plat- forms and content-sorting algorithms were generally avail- able to the internet-using public. That focus reveals that ISIS' ability to beneft from these platforms was merely inci- dental to defendants' services and general business models; it was not attributable to any culpable conduct of defendants directed toward ISIS. And, when considering the fourth and ffth factors (the defendants' relationship to ISIS and the defendants' state of mind), the Ninth Circuit should have given much greater weight to defendants' arm's-length rela- tionship with ISIS—which was essentially no different from their relationship with their millions or billions of other users—and their undisputed lack of intent to support ISIS. See id., at 910. Cite as: 598 U. S. 471 (2023) 505

Taken as a whole, the Ninth Circuit's analytic approach thus elided the fundamental question of aiding-and-abetting liability: Did defendants consciously, voluntarily, and culpa- bly participate in or support the relevant wrongdoing? As we have explained above, the answer in this case is no. Plaintiffs allege only that defendants supplied generally available virtual platforms that ISIS made use of, and that defendants failed to stop ISIS despite knowing it was using those platforms. Given the lack of nexus between that as- sistance and the Reina attack, the lack of any defendant in- tending to assist ISIS, and the lack of any sort of affrmative and culpable misconduct that would aid ISIS, plaintiffs' claims fall far short of plausibly alleging that defendants aided and abetted the Reina attack.

C That leaves the set of allegations specifc to Google. As explained above, plaintiffs allege that Google reviewed and approved ISIS videos on YouTube as part of its revenue- sharing system and thereby shared advertising revenue with ISIS. The Ninth Circuit briefy mentioned those allegations when analyzing plaintiffs' complaint here. However, in ad- dressing another, materially identical complaint, the Ninth Circuit held that the same allegations “failed to state a claim for aiding-and-abetting liability” because they were “devoid of any allegations about how much assistance Google pro- vided” and therefore did not plausibly allege “that Google's assistance was substantial.” Id., at 907. We think that the Ninth Circuit was correct in that hold- ing. The complaint here alleges nothing about the amount of money that Google supposedly shared with ISIS, the num- ber of accounts approved for revenue sharing, or the content of the videos that were approved. It thus could be the case that Google approved only one ISIS-related video and shared only $50 with someone affliated with ISIS; the complaint simply does not say, nor does it give any other reason to view 506 TWITTER, INC. v. TAAMNEH

Google's revenue sharing as substantial assistance. Without more, plaintiffs thus have not plausibly alleged that Google knowingly provided substantial assistance to the Reina at- tack, let alone (as their theory of liability requires) every single terrorist act committed by ISIS.

V By their very nature, the concepts of aiding and abetting and substantial assistance do not lend themselves to crisp, bright-line distinctions. However, both the common law and Halberstam provide some clear guideposts: The point of aid- ing and abetting is to impose liability on those who con- sciously and culpably participated in the tort at issue. The focus must remain on assistance to the tort for which plain- tiffs seek to impose liability. When there is a direct nexus between the defendant's acts and the tort, courts may more easily infer such culpable assistance. But, the more attenu- ated the nexus, the more courts should demand that plain- tiffs show culpable participation through intentional aid that substantially furthered the tort. And, if a plaintiff's theory would hold a defendant liable for all the torts of an enter- prise, then a showing of pervasive and systemic aid is re- quired to ensure that defendants actually aided and abetted each tort of that enterprise. Here, however, the nexus between defendants and the Reina attack is far removed. As alleged by plaintiffs, de- fendants designed virtual platforms and knowingly failed to do “enough” to remove ISIS-affliated users and ISIS-related content—out of hundreds of millions of users worldwide and an immense ocean of content—from their platforms. Yet, plaintiffs have failed to allege that defendants intentionally provided any substantial aid to the Reina attack or other- wise consciously participated in the Reina attack—much less that defendants so pervasively and systemically assisted ISIS as to render them liable for every ISIS attack. Plain- Cite as: 598 U. S. 471 (2023) 507

Jackson, J., concurring

tiffs accordingly have failed to state a claim under § 2333(d)(2). We therefore reverse the judgment of the Ninth Circuit.

It is so ordered.

Justice Jackson, concurring. I join the opinion of the Court with the understanding that today's decisions are narrow in important respects. In this case and its companion, Gonzalez v. Google, 598 U. S. 617 (2023) (per curiam), the Court has applied 18 U. S. C. § 2333(d)(2) to two closely related complaints, fled by the same counsel. Both cases came to this Court at the motion- to-dismiss stage, with no factual record. And the Court's view of the facts—including its characterizations of the social-media platforms and algorithms at issue—properly rests on the particular allegations in those complaints. Other cases presenting different allegations and different records may lead to different conclusions. The Court also draws on general principles of tort and criminal law to inform its understanding of § 2333(d)(2). General principles are not, however, universal. The common-law propositions this Court identifes in interpret- ing § 2333(d)(2) do not necessarily translate to other contexts. Reporter’s Note

The attached opinion has been revised to refect the usual publication and citation style of the United States Reports. The revised pagination makes available the offcial United States Reports citation in advance of publication. The syllabus has been prepared by the Reporter of Decisions for the convenience of the reader and constitutes no part of the opinion of the Court. A list of counsel who argued or fled briefs in this case, and who were members of the bar of this Court at the time this case was argued, has been inserted following the syllabus. Other revisions may include adjustments to formatting, captions, citation form, and any errant punctuation. The following additional edits were made:

None