Tamara Fields v. Twitter, Inc. ( 2018 )

FOR PUBLICATION
UNITED STATES COURT OF APPEALS
FOR THE NINTH CIRCUIT
TAMARA FIELDS, on behalf of No. 16-17165
herself, as a representative of the
Estate of Lloyd Fields, Jr.; HEATHER
CREACH, on behalf of herself and as D.C. No.
a representative of the Estate of 3:16-cv-00213-
James Damon Creach; J.C. (1), a WHO
minor; J.C. (2), a minor,
Plaintiffs-Appellants,
OPINION
v.
TWITTER, INC.,
Defendant-Appellee.
Appeal from the United States District Court
for the Northern District of California
William Horsley Orrick, District Judge, Presiding
Argued and Submitted December 6, 2017
San Francisco, California
Filed January 31, 2018
2 FIELDS V. TWITTER
Before: MILAN D. SMITH, JR. and SANDRA S. IKUTA,
Circuit Judges, and STEVEN J. MCAULIFFE, *
District Judge.
Opinion by Judge Milan D. Smith, Jr.
SUMMARY **
Anti-Terrorism Act
The panel affirmed the district court’s dismissal of an
action seeking civil remedies under the Anti-Terrorism Act
against Twitter, Inc.
Plaintiffs, whose family members were killed while
working as government contractors in Jordan in an attack for
which ISIS claimed credit, alleged injury “by reason of”
Twitter’s knowing provision of material support to ISIS.
The panel affirmed the district court’s holding that
plaintiffs failed to adequately plead proximate causation
because, to satisfy the ATA’s “by reason of” requirement, a
plaintiff must show at least some direct relationship between
the injuries that he or she suffered and the defendant’s acts.
*
The Honorable Steven J. McAuliffe, Senior United States District
Judge for the District of New Hampshire, sitting by designation.
**
This summary constitutes no part of the opinion of the court. It
has been prepared by court staff for the convenience of the reader.
FIELDS V. TWITTER 3
The panel declined to reach the district court’s additional
holding that Twitter’s liability was precluded by § 230 of the
Communications Decency Act because plaintiffs’ claims
sought to treat Twitter as the publisher of ISIS’s content.
COUNSEL
Joshua D. Arisohn (argued), L. Timothy Fisher, and Scott A.
Bursor, Bursor & Fisher P.A., Walnut Creek, California,
New York, New York, for Plaintiffs-Appellants.
Seth P. Waxman (argued), Patrick J. Carome, and Ari
Holtzblatt, Wilmer Cutler Pickering Hale and Dorr LLP,
Washington, D.C.; Mark D. Flanagan, Wilmer Cutler
Pickering Hale and Dorr LLP, Palo Alto, California; for
Defendant-Appellee.
Aaron Mackey, Jamie Williams, and Sophia Cope,
Electronic Frontier Foundation, San Francisco, California,
for Amici Curiae Electronic Frontier Foundation and Center
for Democracy & Technology.
Sonali D. Maitra, Durie Tangri LLP, San Francisco,
California, for Amicus Curiae Internet Association.
Catherine R. Gellis, Sausalito, California, for Amicus Curiae
Floor 64 Inc. DBA The Copia Institute.
4 FIELDS V. TWITTER
OPINION
M. SMITH, Circuit Judge:
After Lloyd “Carl” Fields, Jr., and James Damon Creach
were killed while working as government contractors in
Jordan in an attack for which ISIS claimed credit, Plaintiffs-
Appellants sued Defendant-Appellee Twitter, Inc. (Twitter)
pursuant to 18 U.S.C. § 2333(a), the civil remedies provision
of the Anti-Terrorism Act (ATA), alleging that they were
injured “by reason of” Twitter’s knowing provision of
material support to ISIS. Twitter moved to dismiss the case,
and its motion was granted. The district court held that
Plaintiffs-Appellants had failed to plead that they were
injured “by reason of” Twitter’s conduct. The district court
also ruled that Twitter’s liability was precluded by § 230 of
the Communications Decency Act (CDA), 47 U.S.C.
§ 230(b), because Plaintiffs-Appellants’ claims sought to
treat Twitter as the publisher of ISIS’s content. Plaintiffs-
Appellants have appealed both holdings. We affirm on the
ground that Plaintiffs-Appellants have failed to adequately
plead proximate causation.
FACTUAL AND PROCEDURAL BACKGROUND
I. The Deaths of Lloyd “Carl” Fields, Jr., and James
Damon Creach
Plaintiffs-Appellants Tamara Fields and Heather Creach
brought this lawsuit on behalf of themselves and as
representatives of the estates of their husbands, Lloyd “Carl”
Fields, Jr. (Fields), and James Damon Creach (Creach),
respectively. They are joined as Plaintiffs-Appellants by
J.C. (1) and J.C. (2), Creach’s two minor sons, who are
represented by their common legal guardian, Heather
Creach. All Plaintiffs-Appellants are American nationals.
FIELDS V. TWITTER 5
This case arises from the tragic deaths of Fields and
Creach in Jordan, on November 9, 2015. Fields had
“travelled to Jordan on June 12, 2015 as a government
contractor through DynCorp International.” “Creach [had]
arrived in Jordan on October 15, 2015[,] where he was
working through the government contractor DECO, Inc.”
While in Jordan, both men were assigned to work at the
International Police Training Center (the IPTC) in southeast
Amman, where they “both used their years of experience as
police officers to train law enforcement personnel from
Jordan, Iraq and the Palestinian Territories in basic police
and security skills.” Fields “had previously served as a
Deputy Sheriff in Calcasieu Parish, Louisiana, and as a
police advisor in both Iraq and Afghanistan.” Creach “was
a graduate of the Virginia Beach Polic[e] Academy, a former
police officer in the Virginia Police Department and had
trained police officers in Afghanistan, Kenya and other
locations.”
One of the students at the IPTC was Anwar Abu Zaid
(Abu Zaid), a “28-year old Jordanian police captain.” “On
November 9, 2015, Abu Zaid arrived at [the] IPTC[,]
smuggling [in] a Kalashnikov assault rifle with 120 bullets
and two handguns in his car.” Because he was an officer, he
was not searched upon entry. “After the noontime prayer,
Abu Zaid shot a truck that was moving through the facility,
killing [Creach]. Abu Zaid then entered the facility’s
cafeteria where he killed an additional four people eating
lunch, including [Fields].” Israeli military intelligence
ultimately determined that “Abu Zaid was a graduate of al-
Mutah University in al-Karak, Jordan where he was part of
6 FIELDS V. TWITTER
a clandestine . . . terror cell” associated with the Islamic
State of Iraq and Syria (al-Sham) (ISIS). 1
“ISIS claimed responsibility for the attack in a statement
issued through [its] al-Battar Media Foundation: ‘Yes . . . we
kill the Americans in Amman,’ the terror group said.” ISIS
reiterated the claim in its Dabiq Magazine, Issue 12:
“And on 9 November 2015, Anwar Abu Zeid
— after repenting from his former occupation
— attacked the American crusaders and their
apostate allies, killing two American
crusaders, two Jordanian apostates, and one
South African crusader. These are the deeds
of those upon the methodology of the revived
Khilāfah. They will not let its enemies enjoy
rest until enemy blood is spilled in revenge
for the religion and the Ummah.”
II. Twitter’s Conduct
Plaintiffs-Appellants have sued Twitter pursuant to
18 U.S.C. § 2333(a), the ATA’s civil remedies provision.
Plaintiffs-Appellants accuse Twitter of violating § 2333(a)
by knowingly providing material support to ISIS, in
violation of 18 U.S.C. § 2339A and 18 U.S.C. § 2339B. 2
Specifically, Plaintiffs-Appellants allege that Twitter
(1) provided material support to ISIS, a Foreign Terrorist
1
ISIS is also known as “the Islamic State of Iraq and the Levant,”
the “Islamic State,” “ad-Dawlah al-Islāmiyah fīl-ʿIrāq wash-Shām,” and
“al-Qaeda in Iraq.”
2
“Material support or resources” is defined statutorily to include,
among other things, any “service” and “communications equipment.” 18
U.S.C. § 2339A(b); see also id. § 2339B(g)(4).
FIELDS V. TWITTER 7
Organization (FTO), in the form of Twitter accounts and
direct-messaging services, (2) did so knowingly and
recklessly, and (3) thereby proximately caused Plaintiffs-
Appellants’ injuries.
These allegations are segregated into three sections of
Plaintiffs-Appellants’ Second Amended Complaint (SAC).
First, in support of their material-support allegation,
Plaintiffs-Appellants claim that Twitter has provided ISIS
with Twitter accounts. Plaintiffs-Appellants identify three
examples of ISIS-affiliated Twitter accounts with large
numbers of followers, which constitute a fraction of the
“estimated 70,000 Twitter accounts, at least 79 of which
were ‘official,’” that ISIS had as of December 2014.
Plaintiffs-Appellants contend that “[s]ince 2010, Twitter has
provided ISIS with dozens of accounts on its social
network,” and until recently did nothing while “the number
of ISIS accounts on Twitter grew at an astonishing rate.”
Second, in support of their scienter allegation, Plaintiffs-
Appellants assert that ISIS is a well-known FTO that openly
uses Twitter. Plaintiffs-Appellants claim that “[t]he United
Nations and international NGOs have condemned ISIS for
war crimes and ethnic cleansing,” and that the United States
designated ISIS an FTO on December 17, 2004. Plaintiffs-
Appellants also provide numerous examples of media
reports documenting ISIS’s Twitter usage, which reports
have prompted many government officials and technology
experts to urge Twitter to be more aggressive in combating
terrorism. Only recently has Twitter responded by changing
its rules to prohibit threats of violence or terrorism and
terrorism promotion, and by suspending terrorism-
promoting accounts.
Third, in support of their causation allegation, Plaintiffs-
Appellants allege ways in which ISIS uses Twitter.
8 FIELDS V. TWITTER
Plaintiffs-Appellants indicate that ISIS uses Twitter’s Direct
Messaging feature to communicate with potential recruits
and “for fundraising and operational purposes.” They assert
that ISIS also uses Twitter to recruit more publicly, by
posting “instructional guidelines and promotional videos,
referred to as ‘mujatweets.’” Plaintiffs-Appellants also
claim that within the year preceding August 2016 alone,
Twitter allowed ISIS to attract “more than 30,000 foreign
recruits,” and that ISIS uses Twitter to fundraise and to
“spread propaganda and incite fear by posting graphic
photos and videos of its terrorist feats.”
III. Prior Proceedings
Plaintiffs-Appellants filed their complaint on January 13,
2016. After Twitter filed its first motion to dismiss on March
10, 2016, Plaintiffs-Appellants filed an Amended Complaint
on March 24, 2016. Twitter again moved to dismiss on April
6, 2016, and its motion was granted with leave to amend on
August 10, 2016. Plaintiffs-Appellants filed the SAC on
August 30, 2016, and Twitter filed a motion to dismiss it on
September 13, 2016. The motion was granted with prejudice
on November 18, 2016, and a final judgment was entered.
This appeal timely followed.
STANDARD OF REVIEW
“We review de novo the district court’s grant of a motion
to dismiss under Rule 12(b)(6), accepting all factual
allegations in the complaint as true and construing them in
the light most favorable to the nonmoving party.”
Campidoglio LLC v. Wells Fargo & Co., 870 F.3d 963, 970
(9th Cir. 2017) (quoting Skilstaf, Inc. v. CVS Caremark
Corp., 669 F.3d 1005, 1014 (9th Cir. 2012)). “We also
review de novo questions of statutory interpretation.” Doe
v. Internet Brands, Inc., 824 F.3d 846, 850 (9th Cir. 2016).
FIELDS V. TWITTER 9
ANALYSIS
I. The ATA’s Proximate Causation Requirement
The civil remedies section of the ATA allows any United
States national “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,” to sue in federal court and
recover treble damages and attorney’s fees. 18 U.S.C.
§ 2333(a). 3
The ATA also contains criminal provisions, the violation
of which can provide the basis for a cause of action under
§ 2333(a). As relevant here, § 2339A prohibits the provision
of “material support or resources” by anyone “knowing or
intending that they are to be used in preparation for, or in
carrying out” any of several enumerated crimes. 18 U.S.C.
§ 2339A. Section 2339B prohibits the knowing provision of
“material support or resources to a foreign terrorist
organization.” Id. § 2339B(a)(1).
Plaintiffs-Appellants allege that Twitter violated both
§ 2339A and § 2339B when it knowingly provided Twitter
3
The term “international terrorism” is statutorily defined to include
activities occurring abroad 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” and “(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 . . . .”
18 U.S.C. § 2331(1). Twitter does not argue that its acts do not qualify
as “international terrorism” under this definition — indeed, Twitter
mentions this issue only in a footnote of its brief. Therefore, we do not
address it. For purposes of this appeal, we assume, arguendo, that the
SAC pleads properly the ATA’s “international terrorism” element.
10 FIELDS V. TWITTER
accounts and its Direct Messaging services to ISIS, and is
therefore liable under § 2333(a). Twitter argues that
Plaintiffs-Appellants have failed to show that they were
injured “by reason of” its alleged acts, as is required for
liability under that section.
The gravamen of the parties’ disagreement is the scope
of the ATA’s “by reason of” requirement. Appropriately,
the parties do not dispute that the ATA’s “by reason of”
language requires a showing of proximate causation. Rather,
they disagree concerning what such a showing entails. 4
Plaintiffs-Appellants contend that proximate causation is
established under the ATA when a defendant’s “acts were a
substantial factor in the sequence of responsible causation,”
and the injury at issue “was reasonably foreseeable or
anticipated as a natural consequence.” See Rothstein v. UBS
AG, 708 F.3d 82, 91 (2d Cir. 2013)). Twitter argues that the
standard is higher, requiring Plaintiffs-Appellants to show
that Twitter’s conduct “led directly” to their injuries. The
district court declined to decide the question because it
concluded that Plaintiffs-Appellants’ pleading was
insufficient under either standard. We conclude that Twitter
has the better of the argument, and hold that to satisfy the
ATA’s “by reason of” requirement, a plaintiff must show at
least some direct relationship between the injuries that he or
she suffered and the defendant’s acts.
4
In footnotes in their briefing, the parties suggest that there might
be a second dispute regarding the issue of but-for causation. However,
because this dispute was not “argued specifically and distinctly” in either
party’s brief, we do not address it. Greenwood v. F.A.A., 28 F.3d 971,
977 (9th Cir. 1994); see also Rodriguez v. Airborne Express, 265 F.3d
890, 894 n.2 (9th Cir. 2001).
FIELDS V. TWITTER 11
A. The ATA’s “By Reason Of” Language
We undertake our construction of the ATA’s “by reason
of” provision against the backdrop of two key assumptions
mandated by the Supreme Court: First, we assume that
Congress is familiar with the “‘well established principle of
the common law that in all cases of loss, we are to attribute
it to the proximate cause, and not to any remote cause . . .
and does not mean to displace it sub silentio’ in federal
causes of action.” Bank of Am. Corp. v. City of Miami,
137 S. Ct. 1296, 1305 (2017) (alteration omitted) (quoting
Lexmark Int’l, Inc. v. Static Control Components, Inc.,
134 S. Ct. 1377, 1390 (2014)); see also Rothstein, 708 F.3d
at 95 (“[I]f, in creating civil liability through § 2333,
Congress had intended to allow recovery upon a showing
lower than proximate cause, we think it either would have so
stated expressly or would at least have chosen language that
had not commonly been interpreted to require proximate
cause for the prior 100 years.”). Second, we assume that
because Congress has used “the same words” — “by reason
of” — in the ATA as it used previously in the Sherman,
Clayton, and Racketeer Influenced and Corrupt
Organizations (RICO) Acts, Congress intended these words
to “have the same meaning that courts had already given
them” in those contexts. Holmes v. Sec. Inv’r Prot. Corp.,
503 U.S. 258, 268 (1992). 5
5
All four statutes use near identical language to create a private right
of action. Compare Sherman Act, ch. 647, § 7, 26 Stat. 209, 210 (1890)
(permitting “[a]ny person who shall be injured in his business or property
by any other person or corporation by reason of anything forbidden or
declared to be unlawful by this act” to sue in federal court and recover
treble damages and attorney’s fees), with 15 U.S.C. § 15(a) (the Clayton
Act) (permitting “any person who shall be injured in his business or
12 FIELDS V. TWITTER
In light of these assumptions, we understand that the
phrase “by reason of” connotes some degree of directness.
We begin our construction of the statutory language with the
Court’s decision in Holmes v. Securities Investor Protection
Corporation, 503 U.S. 258 (1992). In that case, the Court
was charged with interpreting the “by reason of” language
that Congress had included in the civil RICO statute. Id. at
265–70. To do so, the Court looked first to the history of the
language as used in the Sherman and Clayton Acts, and
determined that a longstanding, central element of proximate
causation in the Clayton Act context was “some direct
relation between the injury asserted and the injurious
conduct alleged.” Id. at 268. The Court reasoned that courts
coalesced around this directness requirement because not
requiring “some direct relation” (1) would make it more
difficult to determine the amount of damages “attributable to
the violation, as distinct from other, independent factors”;
(2) would force courts to develop complicated damages-
apportionment rules to avoid multiple recoveries; and
(3) would create these difficulties needlessly, because
requiring some direct relation would never prevent directly
injured victims from utilizing the law. Id. at 268–70. The
Court then held that “these reasons appl[ied] with equal
force to suits” brought under the civil RICO provision, id. at
270, and convinced the Court to hold that civil RICO
liability required a showing of “some direct relation between
the injury asserted and the injurious conduct alleged,” id. at
268.
property by reason of anything forbidden in the antitrust laws” to sue in
federal court and recover treble damages and attorney’s fees), and
18 U.S.C. § 1964(c) (the RICO Act) (permitting “[a]ny person injured in
his business or property by reason of a violation of section 1962” to sue
in federal court and recover treble damages and attorney’s fees).
FIELDS V. TWITTER 13
Subsequent civil RICO cases affirmed this requirement.
In Anza v. Ideal Steel Supply Corp., 547 U.S. 451 (2006), the
Court reiterated that “[w]hen a court evaluates a RICO claim
for proximate causation, the central question it must ask is
whether the alleged violation led directly to the plaintiff’s
injuries.” Id. at 461. And in Hemi Group, LLC v. City of
New York, 559 U.S. 1 (2010), the Court emphasized that a
claim would not meet RICO’s direct relationship
requirement if it required the Court to move beyond the first
step in the causal chain. Id. at 8–12; see also Holmes,
503 U.S. at 271 (“The general tendency of the law, in regard
to damages at least, is not to go beyond the first step.”
(quoting Associated Gen. Contractors of Cal., Inc. v. Cal.
State Council of Carpenters, 459 U.S. 519, 534 (1983)).
More recently, outside the RICO context, the Court has
explored the necessity and scope of the “first step”
limitation. Echoing Holmes’s three concerns, the Court
explained in Lexmark International that the first-step
limitation was crucial in most cases because “there
ordinarily is a ‘discontinuity’ between the injury to the direct
victim and the injury to the indirect victim, so that the latter
is not surely attributable to the former (and thus also to the
defendant’s conduct), but might instead have resulted from
‘any number of [other] reasons.’” 134 S. Ct. at 1394(alteration in original) (quoting Anza, 547 U.S. at 458–59).
The Court observed that the general purpose of requiring
proximate causation is to “bar[] suits for alleged harm that is
‘too remote’ from the defendant’s unlawful conduct,” and
limit recovery to those cases where “the harm alleged has a
sufficiently close connection to the conduct the statute
prohibits.” Id. at 1390.
The ATA presents precisely the risks with which the
Court was concerned in Holmes and Lexmark. On its face,
14 FIELDS V. TWITTER
the ATA does not limit recovery to those directly injured.
Rather, it allows “[a]ny national of the United States injured
. . . by reason of an act of international terrorism, or his or
her estate, survivors, or heirs” to sue in federal court.
18 U.S.C. § 2333(a) (emphasis added). Thus, where an act
of international terrorism causes an injury indirectly, there is
inevitably a “discontinuity” between the injury and the
defendant’s conduct. See Lexmark Int’l, 134 S. Ct. at 1390(explaining injuries that are “too remote” from a defendant’s
unlawful conduct “ordinarily” arise where the harm for
which recovery is permitted is “purely derivative of
‘misfortunes visited upon a third person by the defendant’s
acts’” (quoting Holmes, 503 U.S. at 268)); see also Linde v.
Arab Bank, PLC, 97 F. Supp. 3d 287, 324 (E.D.N.Y. 2015)
(“The tort [that the ATA] condemns is one of secondary
action, not primary action. It assumes the existence of a tort
by a third party, and then renders the defendant liable for
providing support to that third party.”).
Accordingly, the same three reasons that compelled the
Holmes Court to adopt the Clayton Act’s “some direct
relation” requirement in the RICO context now motivate us
to adopt that requirement in the context of the ATA. As
relevant here, the conduct the ATA prohibits is the provision
of material support to international terrorists. See 18 U.S.C.
§§ 2333(a), 2339A, 2339B; see also Bank of Am. Corp.,
137 S. Ct. at 1305 (2017) (identifying what alleged conduct
the FHA prohibited in order to analyze proximate cause).
Not requiring that this provision of support have some direct
relation to a plaintiff’s injuries (1) would make it extremely
difficult to attribute damages to the provision of material
support as distinct from other intervening factors, (2) would
force courts to develop complicated damages-apportionment
rules to avoid multiple recoveries, and (3) would create these
difficulties needlessly, because victims injured more directly
FIELDS V. TWITTER 15
by the provision of material support would not be prevented
from recovery by a “direct relation” requirement. See
Holmes, 503 U.S. at 268–70.
B. Plaintiffs-Appellants’ Foreseeability Argument
Plaintiffs-Appellants urge us to break with Holmes and
its progeny and adopt a different standard for proximate
causation. Citing the Second Circuit’s decision in Rothstein
and the district court decision in Linde, Plaintiffs-Appellants
argue that “[p]roximate causation is established under the
ATA when a defendant’s ‘acts were a substantial factor in
the sequence of responsible causation,’ and the injury at
issue ‘was reasonably foreseeable or anticipated as a natural
consequence.’”
Plaintiffs-Appellants’ authorities are neither
authoritative nor persuasive. With regard to Rothstein,
Plaintiffs-Appellants are correct that in distinguishing the
traceability requirement of Article III standing from
proximate cause in that case, the Second Circuit quoted its
own prior observation that
[c]entral to the notion of proximate cause is
the idea that a person is not liable to all those
who may have been injured by his conduct,
but only to those with respect to whom his
acts were a substantial factor in the sequence
of responsible causation and whose injury
was reasonably foreseeable or anticipated as
a natural consequence.
708 F.3d at 91 (quoting Lerner v. Fleet Bank, N.A., 318 F.3d
113, 123 (2d Cir. 2003), as amended (Apr. 16, 2003)). But
the Second Circuit also quoted the Supreme Court’s holding
in Anza that with respect to “proximate causation, the central
16 FIELDS V. TWITTER
question is whether the alleged violation led directly to the
plaintiff’s injuries,” id. at 91–92 (alteration omitted)
(quoting Anza, 547 U.S. at 461), and held that the ATA’s “by
reason of” language should be interpreted to require a
showing of “proximate cause, as the term is ordinarily used,”
in the Sherman, Clayton, and RICO Act contexts, id. at 95.
Ultimately, after considering both proximate cause
standards, the Rothstein court found the plaintiffs had failed
to plead proximate causation without ever defining precisely
its understanding of the concept. See id. at 95–97.
Similarly, while the district court in Linde denied the
defendant’s Rule 59 challenge to a jury instruction on
causation that had focused “solely on whether [the]
defendant’s acts were a substantial factor in causing [the]
plaintiffs’ injuries, and whether such injuries were a
foreseeable result of those acts,” the court also “instructed
the jury that ‘activities that are too remote, too indirect, or
too attenuated are insufficient’ to show proximate
causation.” 97 F. Supp. 3d at 328 (alteration omitted). This
language comes from Hemi Group, in which the Supreme
Court instructed that
[p]roximate cause for RICO purposes . . .
should be evaluated in light of its common-
law foundations; proximate cause thus
requires “some direct relation between the
injury asserted and the injurious conduct
alleged.” A link that is “too remote,” “purely
contingent,” or “indirect” is insufficient.
559 U.S. at 9 (alteration omitted) (quoting Holmes, 503 U.S.
at 268, 271, 274). As was the case in Rothstein, rather than
replacing a proximate cause definition based in directness
with one based in foreseeability, the Linde court blurred the
FIELDS V. TWITTER 17
two concepts. Thus, at most, Plaintiffs-Appellants’ cases
indicate that courts in a circuit other than ours have
considered foreseeability and directness when evaluating
showings of proximate cause in cases brought under the
ATA. These cases do not trump Holmes and its Supreme
Court progeny, and they do not establish that directness is
unnecessary, or that a showing of foreseeability is sufficient
on its own to demonstrate proximate causation.
To be clear, we do not hold that a consideration of
foreseeability is irrelevant to, or never required in, a
proximate cause analysis. Proximate cause is an infamously
nebulous concept that the Court has explained is meant to
serve as a generic label for “the judicial tools used to limit a
person’s responsibility for the consequences of that person’s
own acts.” Holmes, 503 U.S. at 268; see also Paroline v.
United States, 134 S. Ct. 1710, 1719 (2014) (“The idea of
proximate cause, as distinct from actual cause or cause in
fact, defies easy summary.”); United States v. Galan,
804 F.3d 1287, 1290 (9th Cir. 2015) (“As the Court
demonstrated, the phrase ‘proximate cause’ hides (or
encompasses) interpretive problems of its own.”). We
recognize that foreseeability is another of the “judicial tools”
in the proximate cause toolshed. See Hemi Grp., 559 U.S. at
12 (“The concepts of direct relationship and foreseeability
are of course two of the ‘many shapes proximate cause took
at common law.’” (alteration omitted) (quoting Holmes,
559 U.S. at 268)).
However, for purposes of the ATA, it is a direct
relationship, rather than foreseeability, that is required. “The
key to the better interpretation lies in [the] statutory history.”
Holmes, 503 U.S. at 267. In the ATA, Congress chose to use
the phrase “by reason of” to require a proximate cause
showing, and the Court has consistently rejected arguments
18 FIELDS V. TWITTER
that this language requires only foreseeability. See Hemi
Grp., 559 U.S. at 12 (affirming Anza’s rejection of
foreseeability in favor of focus “on the directness of the
relationship between the conduct and the harm”). And even
in other contexts, where the Court has given foreseeability
analysis more weight, the Court has still evinced a concern
for the presence of some direct relationship between the
defendant’s acts and the injury. See, e.g., Cty. of L.A. v.
Mendez, 137 S. Ct. 1539, 1548–49 (2017) (holding that
§ 1983 claim “required consideration of the ‘foreseeability
or the scope of the risk created by the predicate conduct,’ and
required the court to conclude that there was ‘some direct
relation between the injury asserted and the injurious
conduct alleged’” (quoting Paroline, 134 S. Ct. at 1719)).
Thus, while it is true that “[t]he proximate-cause inquiry is
not easy to define, and over the years . . . has taken various
forms,” we “have a great deal of experience applying it, and
. . . a wealth of precedent for [us] to draw upon in doing so.”
Lexmark, 134 S. Ct. at 1390. Here, the relevant precedents
analyzing the phrase “by reason of” dictate that it must
require a showing of at least some direct relationship
between a defendant’s acts and a plaintiff’s injuries.
C. Plaintiffs-Appellants’ Fungibility Argument
In support of their alternative proximate causation
definition, Plaintiffs-Appellants rely heavily on case law
emphasizing the fungibility of support to terrorist
organizations. For example, in Humanitarian Law Project
v. Reno, 205 F.3d 1130 (9th Cir. 2000), we held that “all
material support” given to terrorist organizations “aids their
unlawful goals.” Id. at 1136. Focusing our attention on
financial contributions in particular, we noted that there was
no way to distinguish among those given to terrorist
organizations for good versus for ill because (1) a donor has
FIELDS V. TWITTER 19
no way to tell how his donation is used; (2) “even
contributions earmarked for peaceful purposes can be used
to give aid to the families of those killed while carrying out
terrorist acts, thus making the decision to engage in terrorism
more attractive”; and (3) “money is fungible,” such that
“giving support intended to aid an organization’s peaceful
activities frees up resources that can be used for terrorist
acts.” Id. And, in Holder v. Humanitarian Law Project,
561 U.S. 1 (2010), the Supreme Court expanded on our
reasoning, holding that material support “in any form” could
be fungible. Id. at 29–36.
But these cases do not sway us to adopt Plaintiffs-
Appellants’ foreseeability standard. For one thing, their
concern is with what constitutes “material support,” not
proximate causation. For another, their emphasis on the
fungibility of support to terrorists only highlights the
insufficiency of foreseeability alone as the standard for
proximate causation in the ATA context. Indeed, the Court
recently rejected a foreseeability standard of proximate
causation for precisely this reason in Bank of America
Corporation v. City of Miami, holding that in the context of
the Fair Housing Act (FHA), “foreseeability alone [could]
not ensure the close connection that proximate cause
requires.” 137 S. Ct. at 1306. The Court reasoned that
because “[t]he housing market is interconnected with
economic and social life,” an FHA violation could “‘be
expected to cause ripples of harm to flow’ far beyond the
defendant’s misconduct.” Id. (quoting Associated Gen.
Contractors, 459 U.S. at 534). However, the Court found
“[n]othing in the statute [to] suggest[] that Congress
intended to provide a remedy wherever those ripples travel,”
and was troubled by the risk of “massive and complex
damages litigation” that “entertaining suits to recover
damages for any foreseeable result of an FHA violation”
20 FIELDS V. TWITTER
would pose. Id. (quoting Associated Gen. Contractors,
459 U.S. at 545). Thus, the Court held that the FHA should
be analogized to statutes with common law foundations, like
the RICO Act, to which the Court has “repeatedly applied
directness principles” and “require[d] ‘some direct relation
between the injury asserted and the injurious conduct
alleged.’” Id. (quoting Holmes, 503 U.S. at 268).
We reach the same conclusion. Communication services
and equipment are highly interconnected in modern
economic and social life, such that the provision of these
services and equipment to terrorists could be expected to
cause ripples of harm to flow far beyond the defendant’s
misconduct. Nothing in § 2333 indicates that Congress
intended to provide a remedy to every person reached by
these ripples; instead, Congress intentionally used the “by
reason of” language to limit recovery. Moreover, we are
troubled by the seemingly boundless litigation risks that
would be posed by extending the ATA’s bounds as far as
foreseeability may reach.
We also note that even where courts have accepted a
fungibility theory with regard to the provision of material
support to terrorists, they have emphasized that the fact of
fungibility does not relieve claimants of their burden to show
causation. Accepting the fungible nature of material support
does not require a court to also hold that any reckless
contribution to a terrorist group or affiliate, no matter its
attenuation, must result in civil liability. See, e.g., Rothstein,
708 F.3d at 96 (rejecting theory that proximate cause was
established per se by violation of antiterrorism law because
that theory would impose strict liability inconsistent with
Congress’s intent in using “by reason of” language); In re
Terrorist Attacks on Sept. 11, 2001, 714 F.3d 118, 125 (2d
Cir. 2013) (“Although Congress clearly intended to create
FIELDS V. TWITTER 21
impediments to terrorism by ‘the imposition of liability at
any point along the causal chain of terrorism,’ the ‘by reason
of’ language of the statute restricts the imposition of such
liability to situations where plaintiffs plausibly allege that
defendants[’] actions proximately caused their injuries.”
(quoting S. Rep. No. 102-342, at 22 (1992)); Gill v. Arab
Bank, PLC, 893 F. Supp. 2d 474, 522 (E.D.N.Y. 2012)
(rejecting “the contention that any reckless contribution to a
terrorist group or its affiliate, no matter how attenuated, will
result in civil liability, without the demonstration of a
proximate causal relationship to the plaintiff’s injury”).
Thus, the fact of fungibility does not modify the causal
requirement imposed by the ATA’s “by reason of” element.
A plaintiff must show at least some direct relationship
between the injuries that he or she suffered and the
defendant’s acts to bring a successful ATA claim.
II. Plaintiffs-Appellants Fail to Plead Adequately
Proximate Causation.
Here, Plaintiffs-Appellants have not pleaded that
Twitter’s provision of communication equipment to ISIS, in
the form of Twitter accounts and direct messaging services,
had any direct relationship with the injuries that Plaintiffs-
Appellants suffered. At most, the SAC establishes that
Twitter’s alleged provision of material support to ISIS
facilitated the organization’s growth and ability to plan and
execute terrorist acts. But the SAC does not articulate any
connection between Twitter’s provision of this aid and
Plaintiffs-Appellants’ injuries. Rather, as the district court
noted,
the allegations in the SAC do not support a
plausible inference of proximate causation
between Twitter’s provision of accounts to
22 FIELDS V. TWITTER
ISIS and the deaths of Fields and Creach.
Plaintiffs allege no connection between the
shooter, Abu Zaid, and Twitter. There 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.
Thus, proximate causation was not shown in this case.
Though we do not diminish the tragedy of the events that led
to this lawsuit, we hold that Plaintiffs-Appellants have not
pleaded that Twitter’s provision of accounts and messaging
services to ISIS had any direct relation to the injuries
Plaintiffs-Appellants suffered.
CONCLUSION
For the foregoing reasons, we affirm the district court’s
dismissal of Plaintiffs-Appellants’ SAC. Because we have
decided the case based on the insufficiency of Plaintiffs-
Appellants’ pleading alone, we decline to reach the second
question presented: whether Section 230 of
Communications Decency Act of 1996 protects Twitter from
liability. Plaintiffs-Appellants shall bear the costs on appeal.
AFFIRMED.