ORDER GRANTING MOTION TO DISMISS
WILLIAM H. ORRICK, United States District JudgeINTRODUCTION
In November 2015, Lloyd “Carl” Fields, Jr. and James Damon Creach were shot and killed while working as United States government contractors at a law enforcement training center in Amman, Jordan. The shooter was a Jordanian police officer who had been studying at the center. In subsequent statements, the Islamic State of Iraq and Syria (“ISIS”) claimed responsibility for the attack, describing the gunman as a “lone wolf.” Plaintiffs, the wife of Fields and the wife and children of Creach, seek to hold defendant Twitter, Inc. (“Twitter”) liable under 18 U.S.C. § 2333(a), part of the Anti-Terrorism Act (“ATA”), on the theory that Twitter provided material support to ISIS by allowing ISIS to sign up for and use Twitter accounts, and that this material support was a proximate cause of the November 2015 shooting.
Twitter moves to dismiss on several grounds, including that plaintiffs’ claims are barred by the Communications Decency Act (“CDA”), 47 U.S.C. § 230(c). As horrific as these deaths were, under the CDA Twitter cannot be treated as a publisher or speaker of ISIS’s hateful rhetoric and is not liable under the facts alleged. Twitter’s motion to dismiss is GRANTED with leave to amend.
BACKGROUND
In 2015, Fields and Creach travelled to Jordan through their work as government contractors. First Amended Complaint ¶¶ 71-72 (“FAC”) (Dkt. No. 21). Both had served as law enforcement officers in the United States, and both were assigned to the International Police Training Center (“IPTC”), a facility in Amman run by the United States Department of State. Id. ¶ 73.
One of the men studying at the IPTC was Anwar Abu Zaid, a Jordanian police captain. Id. ¶ 76. On November 9, 2015, Abu Zaid smuggled an assault rifle and two handguns into the IPTC and shot and killed Fields, Creach, and three other individuals. Id. ¶ 78. ISIS subsequently “claimed responsibility” for the attack, describing Abu Zaid as a “lone wolf’ and stating,
*967Do not provoke the Muslims more than this, especially recruited and supporters of the Islamic State. The more your aggression against the Muslims, the more our determination and revenge ... [T]ime will turn thousands of supporters of the caliphate on Twitter and others to wolves.
Id. ¶ 80.
Plaintiffs do not allege that ISIS recruited or communicated with Abu Zaid over Twitter, that ISIS or Abu Zaid used Twitter to plan, carry out, or raise funds for the attack, or that Abu Zaid ever viewed ISIS-related content on Twitter or even had a Twitter account. The only arguable connection between Abu Zaid and Twitter alleged in the FAC is that Abu Zaid’s brother told reporters that Abu Zaid had been very moved by ISIS’s execution of Jordanian pilot Maaz al-Kassasbeh in February 2015. Id. ¶ 84. After capturing al-Kassasbeh, ISIS launched a Twitter campaign “to crowd source ideas for his method of execution.” Id. ISIS subsequently used a Twitter account to distribute a 22-minute video of al-Kassasbeh’s horrific killing. Id. Plaintiffs do not allege that Abu Zaid ever viewed the video, either on Twitter or by any other means.
Plaintiffs accuse Twitter of violating 18 U.S.C. § 2333(a), part of the ATA, by knowingly providing material support to ISIS, in violation of 18 U.S.C. § 2339A and 18 U.S.C. § 2339B. FAC ¶¶ 87-90 (Count 1, section 2339A), 91-94 (count 2, section 2339B). Section 2333(a) provides:
Any national of the United States injured in his or her1 person, property, or business by reason of an act of international terrorism, or his or her estate, survivors, or heirs, may sue therefor in any appropriate district court of the United States and shall recover threefold the damages he or she sustains and the cost of the suit, including attorney’s fees.
18 U.S.C. § 2333(a). Sections 2339A and 2339B prohibit the knowing provision of “material support or resources” for terrorist activities or foreign terrorist organizations. 18 U.S.C. §§ 2339A(a), 2339B(a)(1). The term “material support or resources” is defined to include “any property, tangible or intangible, or service,” including “communications equipment.” 18 U.S.C. §§ 2339A(b)(1), 2339B(g)(4).
Plaintiffs assert that' Twitter’s “provision of material support to ISIS was a proximate cause of [their] injuries].” FAC ¶¶ 89, 93. They allege that Twitter “has knowingly permitted... ISIS to use its social network as a tool for spreading extremist propaganda, raising funds and attracting new recruits,” and that “[t]his material support has been instrumental to the rise of ISIS and has enabled it to carry out numerous terrorist attacks, including the November 9, 2015 shooting attack in Amman, Jordan in which [Fields and Creach] were killed.” Id. ¶ 1.
Specifically, plaintiffs contend that ISIS uses Twitter to disseminate its. official media publications and other content, thereby “spread[ing] propaganda and incit[ing] fear [through] graphic photos and videos of its terrorist feats.” Id. ¶¶ 35-36. ISIS also uses Twitter “to raise funds for its terrorist activities,” id. ¶ 30, and to “post instructional guidelines and promotional videos,” id. ¶ 23.
In addition, ISIS uses Twitter as a recruitment platform, “reach[ing] potential recruits by maintaining accounts on Twitter so that individuals across the globe can reach out to [ISIS] directly.” Id. ¶ 20. “After first contact, potential recruits and ISIS recruiters often communicate via Twitter’s Direct Messaging capabilities.”1 *968Id. Plaintiffs allege that “[t]hrough its use of Twitter, ISIS has recruited more than 30,000 foreign recruits over the last year,” Id. ¶ 29.
Plaintiffs cite a number of media reports from between 2011 and 2014 concerning ISIS’s use of Twitter and Twitter’s “refusal to take any meaningful action to stop it.” Id. ¶¶ 48-56. They also describe several attempts by members of the public and United States government to persuade Twitter to crack down on ISIS’s use of its services. Id. ¶¶ 57-62. They allege that, while Twitter has now instituted a rule prohibiting threats of violence and the promotion of terrorism, “many ISIS-themed accounts are still,easily, found on Twitter.” Id. ¶ 70.
LEGAL STANDARD
Federal Rule of Civil Procedure 8(a)(2) requires a complaint to contain “a short and plain statement of the claim showing that the pleader is entitled to relief,” Fed. R. Civ. P. 8(a)(2), in, order to “give the defendant fair notice of what the claim is and the grounds upon which' it rests,” Bell Atl. Corp. v. Twombly, 550 U.S. 544, 555, 127 S.Ct. 1955, 167 L.Ed.2d 929 (2007) (internal quotation marks and alterations omitted).
A motion to dismiss for failure to state a claim under Federal Rule of Civil Procedure 12(b)(6) tests the legal sufficiency of a complaint. Navarro v. Block, 250 F.3d 729, 732 (9th Cir.2001). “Dismissal under Rule 12(b)(6) is appropriate only where the complaint lacks a cognizable legal theory or sufficient facts to support a cognizable legal theory.” Mendiondo v. Centinela Hosp. Med. Ctr., 521 F.3d 1097, 1104 (9th Cir.2008). While a complaint “need not contain detailed factual allegations” to survive a Rule 12(b)(6) motion, “it must plead enough facts to state .a claim to relief that is plausible on its face.” Cousins v. Lockyer, 568 F.3d 1063, 1067-68 (9th Cir.2009) (internal quotation marks and citations omitted). A claim is facially plausible when it “allows the court to draw the reasonable inference that the defendant is liable for the misconduct alleged.” Ashcroft v. Iqbal, 556 U.S. 662, 678, 129 S.Ct. 1937, 173 L.Ed.2d 868 (2009) (internal quotation marks omitted).
In considering whether a claim satisfies this standard, the court must “accept factual allegations in the complaint as true and construe the pleadings in the light most favorable to the nonmoving party.” Manzarek v. St. Paul Fire & Marine Ins. Co., 519 F.3d 1025, 1031 (9th Cir.2008). However, “conclusory allegations of law and unwarranted inferences are insufficient to avoid a Rule 12(b)(6) dismissal.” Cousins, 568 F.3d at 1067 (internal quotation marks omitted), A court may “reject, as implausible, allegations that are too speculative to warrant further factual development.” Dahlia v. Rodriguez, 735 F.3d 1060, 1076 (9th Cir.2013).
DISCUSSION
Twitter moves to dismiss on multiple grounds, but its principal argument is that plaintiffs’ claims are barred by section 230(c), the “protection for ‘Good Samaritan’ blocking and screening of offensive material” provision of the CDA, 47 U.S.C. § 230(c). Section 230(c), contains two subsections, only the first of which, section 230(c)(1), is relevant here:
(1) Treatment of publisher or speaker
No provider or user of an interactive computer service shall be treated as the publisher or speaker of any information provided by another information content provider.
47 U.S.C. § 230(c)(1).
While the Ninth Circuit has described the reach of section 230(c)(1) in broad *969terms, stating that it “immunizes providers of interactive computer services against liability arising from content created by third parties,” the statute does not “create a lawless no-man’s-land on the internet.” Fair Hous. Council of San Fernando Valley v. Roommates.Com, LLC, 521 F.3d 1157, 1162, 1164 (9th Cir.2008); see also Doe v. Internet Brands, Inc., 824 F.3d 846, 852 (9th Cir.2016) (noting that “the CDA does not declare a general immunity from liability deriving from third-party content”) (internal quotation marks omitted). Rather, separated into its elements, section 230(c)(1) protects from liability only (a) a provider or user of an interactive computer service (b) that the plaintiff seeks to treat as a publisher or speaker .(c) of information provided by another information content provider. Barnes v. Yahoo!, Inc., 570 F.3d 1096, 1100-01 (9th Cir.2009).
Plaintiffs do not dispute that Twitter is an interactive computer service provider, or that the offending content highlighted in the FAC was provided by another information content provider. They dispute only the second element of Twitter’s section 230(c)(1) defense, i.e., whether-they seek to treat Twitter as a publisher or speaker.
The prototypical cause of action seeking to treat an interactive computer service provider as a publisher or speaker is defamation. See, e.g., Internet Brands, Inc., 824 F.3d at 851; Barnes, 570 F.3d at 1101.2 However, “the language of the statute does not limit its application to defamation cases.” Barnes, 570 F.3d at 1101. Courts have applied section 230(c)(1) against a variety of claims, including negligent undertaking, id. at 1102-03, intentional assault, Klayman v. Zuckerberg, 753 F.3d 1354, 1357-60 (D.C.Cir.2014), and violation of anti-sex-trafficking laws, Jane Doe No. 1 v. Backpage.com, LLC, 817 F.3d 12, 18-24 (1st Cir.2016). “[W]hat matters is not the name of the 'cause of action— defamation versus negligence versus intentional- infliction of emotional distress— [but] whether the cause of action inherently requires the court to treat the defendant as the publisher or speaker of content provided by another.” Barnes, 570 F.3d at 1101-02 (internal quotation marks omitted). “[C]ourts must ask whether the duty that the plaintiff alleges the defendant violated derives from the defendant’s status or conduct as a publisher or speaker. If it does, section 230(c)(1) precludes liability.” Id. (internal quotation marks omitted).
Twitter contends that plaintiffs seek to hold it liable as the publisher of content created by ISIS. Mot. at 14-16 (Dkt. No. 27). It highlights the opening paragraph of the FAC—i.e., that Twitter “knowingly permitted [ISIS] to use its social network as a tool for spreading extremist propaganda, raising funds and attracting new recruits,” FAC ¶ 1—and the numerous descriptions in the FAC of content created ánd disseminated by ISIS through the Twitter platform. According to Twitter, plaintiffs’ claims are based on Twitter’s alleged failure to exclude this third-party content, a quintessential responsibility of a publisher. See Mot. at 14-16; see also Klayman, 753 F.3d at 1359 (“the very essence of publishing is making the decision whether to print or retract a given piece of content”); Doe v. MySpace, Inc., 528 F.3d 413, 420 (5th Cir.2008) (“decisions relating to the monitoring, screening, and deletion of content [are] actions quintessentially related to a publisher’s role”) (internal quotation marks omitted); Roommates, 521 *970F.3d at 1170-71 (noting that section 230(c)(1) applies to “any activity that can be boiled down to deciding whether to exclude material .that third parties seek to post online,” and that “determining] whether or not to prevent [the] posting” of material by third parties is “precisely the kind of activity” covered by the statute); Batzel v. Smith, 333 F.3d 1018, 1031 (9th Cir.2003) (“the exclusion of ‘publisher’ liability necessarily precludes liability for exercising the usual prerogative of publishers to choose among proffered material”).
Plaintiffs make two arguments in response. First, they contend that their claims are not based on “the contents of tweets, the issuing of tweets, or the failure to remove tweets,” but rather on Twitter’s “provision of Twitter accounts to ISIS in the first place.” Oppo. at 3 (Dkt. No. 31). In other words, “[b]ecause the creation of a Twitter account necessarily occurs before the issuing of tweets from that account, and separately from the creation of published content, [Twitter’s] violations of the ATA cannot be accurately characterized as publishing activity, but rather as the provision of the means through which ISIS spreads its poison.” Id. at 4. Plaintiffs further explain that
[Twitter’s] liability does not arise out of its .publishing conduct, but rather its separate legal duty under the ATA not to provide terrorists with material support. . .Indeed, plaintiffs’ claims do not require publishing or speaking as a critical element. In no sense was [Twitter] acting as a publisher when it permitted ISIS members to sign up for and create these accounts. It was not reviewing, revising, or editing content. Nor was it deciding whether content should be publicly disseminated or withdrawn ’ from the internet. Plaintiffs’ claims are not based on a theory that any particular tweets from ISIS members should have been altered or restricted in any way. Plaintiffs’ claims under the ATA are not based on offending content at all. They are based on Twitter’s provision of material support to ISIS, which is separate and apart from—and antecedent to—the publication of any content. Even if ISIS had never issued a single tweet, [Twitter’s] provision of material support to ISIS in the form of Twitter accounts would constitute a violation of the ATA.
Id. at 7-8 (internal quotation marks, citations, and alterations omitted).
Second, plaintiffs highlight their allegations regarding Twitter’s Direct Messaging capabilities and assert that “[b]ecause these private messages are not published..., a lawsuit based on their content is not barred by the CDA.” Id. at 2. Plaintiffs assert that publishing under the CDA “necessarily involves the dissemination of information to the public” and thus does not encompass the transmission of private messages through Direct Messaging. Id. at 9-11.
I. PROVISION OF ACCOUNTS THEORY
There are at least three problems with plaintiffs’ provision of accounts theory. The first is that it does not align with the allegations in the FAC. Those allegations describe a theory of liability based on Twitter’s knowing failure to prevent ISIS from disseminating content through the Twitter platform, not its mere provision of accounts to ISIS.
To be sure, there are some allegations in the FAC concerning Twitter’s provision of accounts to ISIS. For example, plaintiffs highlight their allegations that: (1) “[s]ince first appearing on Twitter in 2010, ISIS accounts on Twitter have grown at an astonishing rate,” FAC ¶ 3; (2) as of December 2014, ISIS had approximately 70,000 Twitter accounts and posted at least 90 tweets per minute, id. ¶ 6; (3) Al-Furqan, ISIS’s official media wing, “maintained a *971... Twitter page where it posted messages from ISIS leadership as well as videos and images of beheadings and other brutal ... executions to 19,000 followers,” id. ¶ 3; (4) Al-Hayat Media Center, ISIS’s official public relations group, maintained “at least a half dozen Twitter accounts emphasizing the recruitment of Westerners” and had nearly 20,000 followers as of June 2014, id. ¶ 4; (5) ISIS “reaches potential recruits by maintaining accounts on Twitter so that individuals across the globe may reach out to [it] directly,” id. ¶ 20; (6) “[e]ven when Twitter shuts down an ISIS-linked account, it does nothing to stop it from springing right back up” with a different but nearly identical name, id. ¶ 69; and (7) while Twitter has now instituted a rule prohibiting threats of violence and the promotion of terrorism, “many ISIS-themed accounts are still easily found on Twitter.com,” id. ¶ 70.
Plaintiffs characterize these allegations as “focus[ed] on [Twitter’s] provision of ... accounts to ISIS, not the content of the tweets.” Oppo. at 4. But with the exception of the statement that “ISIS accounts on Twitter have grown at an astonishing rate,” FAC ¶ 3, all of the allegations are accompanied by information regarding the ISIS-related content disseminated from the accounts. Plaintiffs allege not just that ISIS had approximately 70,000 Twitter accounts, but that ISIS used those accounts to post at least 90 tweets per minute, id. ¶ 6; not just that Al-Furqan maintained a Twitter page, but that it maintained one “where it posted messages from ISIS leadership as well as videos and images of beheadings and other brutal... executions to 19,000 followers,” id. ¶ 3; not just that Twitter failed to stop an ISIS-linked account from “springing right back up,” but that an inflammatory message was tweeted from this account following the shooting attack in San Bernadino, California in December 2015, id. ¶ 69.
The rest of the FAC is likewise riddled with detailed descriptions of ISIS-related messages, images, and videos disseminated through Twitter and the harms allegedly caused by the dissemination of that content. The FAC also includes a number of allegations specifically faulting Twitter for failing to detect and prevent the dissemination of ISIS-related content through the Twitter platform. See, e.g., id. ¶¶ 60 (Twitter “failed to respond to pleas to shut down clear incitements to violence”), 66 (Twitter “does not actively monitor and will not censor user content”). Indeed, the opening paragraph of the FAC could not be more clear about the content-based theory underlying plaintiffs’ claims:
Twitter has knowingly permitted... ISIS to use its social network as a tool for spreading extremist propaganda, raising funds and attracting new recruits. This material support has been instrumental to the rise of ISIS and has enabled it to carry out numerous terrorist attacks, including the November 9, 2015 shooting attack in Amman, Jordan in which [Fields and Creach] were killed.
Id. ¶ 1 (emphasis added). In the following paragraph, plaintiffs allege that ISIS “has exploited social media, most notoriously Twitter, to send its propaganda and messaging out to the world and to draw in people vulnerable to radicalization,” and that ISIS has been able to use Twitter “to exert an outsized impact on how the world perceives it, by disseminating images of graphic violence (including the beheading of Western journalists and aid workers) ... while using social media to attract new recruits and inspire lone actor attacks.” Id. ¶ 2 (emphasis added).
In short, the theory of liability alleged in the FAC is not that Twitter provides material support to ISIS by providing it with Twitter accounts, but that Twitter does so *972by “knowingly permitting] [ISIS] to. use [those accounts] as a tool for spreading extremist propaganda, raising funds, and attracting new recruits.” Id. ¶ 1. Plaintiffs do not. dispute that this theory seeks to treat Twitter as a publisher and is barred by section 230(c)(1). See, e.g., Oppo. at 2.
The second problem with plaintiffs’ provision of accounts theory is that, even if it were alleged in the FAC, it would be just as barred by section 230(c)(1) as the theory plaintiffs’ actually have alleged. As noted above, courts have repeatedly described publishing activity under section 230(c)(1) as including decisions about what third-party content may be posted online. See, e.g., Klayman, 753 F.3d at 1359 (“the very essence of publishing is making the decision whether to print or retract a given piece of content”); MySpace, 528 F.3d at 420 (“decisions relating to the monitoring, screening, and deletion of content [are] actions quintessentially related to a publisher’s role”); Roommates.Com, 521 F.3d at 1170-71 (“determining] whether or not to prevent [the] posting” of third-party material online is “precisely the kind of activity” covered by the CDA); Batzel, 333 F.3d at 1031 (“the exclusion of ‘publisher’ liability necessarily precludes liability for exercising the usual prerogative of publishers to choose among proffered material”). Plaintiffs’ provision of accounts theory is slightly different, in that it is based on Twitter’s decisions about whether particular third parties may have Twitter accounts, as opposed to what particular third-party content may be posted. But it is not clear to me why this difference matters for the purposes of section 230(c)(1). Under either theory, the alleged wrongdoing is the decision to permit third parties to post content—it is just that under plaintiffs’ provision of accounts theory, Twitter would be liable for granting permission to post (through the provision, of Twitter accounts) instead of for allowing postings that have already occurred. Plaintiffs do not explain why this difference means that the provision of accounts theory seeks to treat Twitter as something other than a publisher of third-party content, and I am not convinced that it does. Despite being based on Twitter accounts instead of tweets, the theory is still based on- Twitter’s alleged violation of a “duty ... derive[d] from [its] status or conduct as a publisher.” Barnes, 570 F.3d at 1102.
A recent First Circuit case, Jane Doe No. 1 v. Backpage.com, LLC, 817 F.3d 12 (1st Cir.2016), ádds further support to this conclusion. The plaintiffs, each of whom had been a victim of sex trafficking, sued the defendant website provider under the Trafficking Victims Protection Reauthori-zation Act, asserting that the defendant had violated the statute through various “choices [it] ha[cl] made about the posting standards for advertisements,” including “the lack of controls on the display of phone numbers,' the option to anonymize email addresses, [and the] acceptance of anonymous payments.” Id. at 20. The plaintiffs argued that “these choices are distinguishable from publisher functions.” Id. The First Circuit disagreed, holding that section 230(c)(1) “extends to the formulation of precisely the sort of website policies and practices [the plaintiffs] assail.” Id. The court explained that decisions regarding the “structure and operation of [a] website”—such as “permitt[ing] users to register under multiple screen names” and other decisions regarding “features that are part and parcel of the overall design and operation of the website”— “reflect choices about what content can appear on the website and in what form” and thus “fall within the purview of traditional publisher functions.” Id. at 20-21.
Likewise, here, Twitter’s decisions to structure and operate itself as a “platform ... allowing] for the freedom of expression [of] hundreds of millions of people *973around the world,” FAC ¶ 65, and to allow even ISIS to “sign up for accounts on its social network,” Oppo. at 3, “reflect choices about what [third-party] content can appear on [Twitter] and in what form,” Backpage, 817 F.3d at 21. Where such choices form the basis of a plaintiffs claim, section 230(c)(1) applies. Id.
Plaintiffs attempt to liken the provision of accounts theory to the promissory estoppel claim raised in Barnes v. Yahoo!, Inc., but the facts of that case are substantially different. The plaintiff in Barnes sued Yahoo on the ground that she had relied on its promise that it would remove explicit photographs her ex-boyfriend had posted online without her consent. 570 F.3d at 1098-99. The Ninth Circuit found that this promissory estoppel claim was not precluded by section 230(c)(1) because the plaintiff did not “seek to hold Yahoo liable as a publisher or speaker of third-party content, but rather as the counter-party to a contract, as a promisor who has breached.” Id. at 1107. In other words, the plaintiffs theory of liability was based not on Yahoo’s “publishing conduct,” but rather on its “manifest intention to be legally obligated to do something.” Id. By contrast, plaintiffs here assert no theory based on contract liability and allege no promise made or breached by Twitter. Barnes does not indicate that the conduct underlying the provision of accounts theory is beyond the scope of publishing conduct.
Plaintiffs also rely on Doe v. Internet Brands, Inc., which the Ninth Circuit decided shortly before oral argument in this case.3 Again, that case involves a substantially different set of facts from this one. The plaintiff there sued the defendant website operator for negligent failure to warn, alleging that the defendant knowingly failed to warn her that two individuals were using the website to identify and lure rape victims. 824 F.3d at 849-50. Although the plaintiff had posted information on the website, the two individuals had not. Id. In holding that the plaintiff did not seek to hold the defendant liable .as- a publisher of third-party content, the Ninth Circuit emphasized that her negligent failure to warn claim
would not require [the defendant] to remove any user content or otherwise affect how it publishes or monitors such content ... Any alleged obligation to warn could have been satisfied without changes to the content posted by the website’s users and -without conducting a detailed .investigation. [The defendant] could have given a warning to... users, perhaps by posting a notice on the website or by informing users by email what it knew about the activities of [the individuals]. Posting or emailing such a warning could be deemed an act of publishing information, but-section 230(c)(1) bars.only liability that treats a website as a publisher or speaker of content provided by somebody else ... A post or email warning that [the defendant] generated would involve only content that [the defendant] itself produced.
Id. at 851. Plaintiffs’ provision of accounts theory-, on the other hand, has nothing to do with information Twitter itself should have posted online. Moreover, it would significantly affect Twitter’s monitoring and publication of third-party content by effectively requiring Twitter to police and restrict its provision of Twitter accounts. Internet Brands, like Barnes, does not help plaintiffs’ case.
The third problem with the provision of accounts theory is that. plaintiffs have not adequately alleged causation. Although the parties dispute the exact formulation of the appropriate causal test for civil liability under the ATA, they agree *974that the statute requires a showing of proximate causation. See Mot. at 20-23; Oppo. at 13-16; see also 18 U.S.C. § 2333(a) (authorizing a suit for damages by “[a]ny national of the United States injured ... by reason of an act of international terrorism”) (emphasis added); In re Terrorist Attacks on Sept. 11, 2001, 714 F.3d 118, 123-25 (2d Cir.2013) (affirming a Rule 12(b)(6) dismissal of ATA claims for failure to plausibly allege proximate causation); Rothstein v. UBS AG, 708 F.3d 82, 94-98 (2d Cir.2013) (same).
Even under plaintiffs’ proposed “substantial factor” test, see Oppo. at 13, the allegations in the FAC do not support a plausible inference of proximate causation between Twitter’s provision of accounts to ISIS and the deaths of Fields and Creach. The only arguable connection between Abu Zaid and Twitter identified in the FAC is that Abu Zaid’s brother told reporters that Abu Zaid had been very moved by ISIS’s horrific execution of al-Kassasbeh, which ISIS publicized through Twitter. See FAC ¶ 84. That connection is tenuous at best regardless of the particular theory of liability plaintiffs decide to assert. But the connection is particularly weak under the provision of accounts theory because it is based on specific content disseminated through Twitter, not the mere provision of Twitter accounts.
The rest of plaintiffs’ arguments and allegations with respect to proximate causation are similarly content-based. For example, plaintiffs assert that Twitter’s provision of material support to ISIS proximately caused the deaths of Fields and Creach because Twitter has (1) “permitted ISIS to use its social network ‘as a tool for spreading extremist propaganda, raising funds and attracting new recruits,’ ” Oppo. at 15 (quoting FAC ¶ 1); (2) “g[i]ve[n] ISIS access to its Direct Messaging capabilities, which [ISIS] has used for ‘covert signaling’ as well as ‘fundraising and operational purposes,’” id. (quoting FAC ¶ 21); (3) allowed ISIS to use Twitter “to recruit more than 30,-000 foreign fighters,” id. (citing FAC ¶ 29); (4) allowed ISIS to use .Twitter “to raise untold sums from its supporters around the world,” id. at 15-16 (citing FAC ¶¶ 21-22, 30-34); and (5) “enabled ISIS to effectuate one of the most ... effective propaganda campaigns in history,” id. at 16 (citing FAC ¶¶ 35-47). Nowhere in their opposition brief or FAC do plaintiffs explain how Twitter’s mere provision of Twitter accounts to ISIS—conduct that allegedly created liability before “the publication of any content” and would support liability “[e]ven if ISIS had never issued a single tweet,” Oppo. at 7-8—proximately caused the November 2015 shooting.
On the one hand, this underscores the conclusion stated above, i.e., that plaintiffs have not actually alleged a theory based on Twitter’s mere provision of Twitter accounts to ISIS. On the other hand, it highlights that, even assuming that plaintiffs have asserted such a theory, they have not plausibly alleged the causal connection necessary to support it.4 Plaintiffs’ claims based on the provision of accounts theory are DISMISSED WITH LEAVE TO AMEND.
II. DIRECT MESSAGING THEORY
Plaintiffs’ other attempt to evade section 230(c)(1) is based on Twitter’s Di*975rect Messaging capabilities, which allow for the sending of private messages through the Twitter platform. Oppo. at 9-11. Plaintiffs contend that publishing activity under section 230(c)(1) “necessarily involves the dissemination of information to the public” and thus does not encompass the transmission of private messages through Direct Messaging. Id. In support of their Direct Messaging theory, plaintiffs abandon all pretense of a content-less basis for liability and assert instead that ISIS has “used [Direct Messaging] to its great advantage,” specifically, to contact and communicate with potential recruits. Id. at 10; see also FAC ¶¶ 20-22.
I am not persuaded that publishing activity under section 230(c)(1) does not extend to Twitter’s Direct Messaging capabilities. As noted above, Congress enacted section 230(c)(1) in part to respond to a New York state court decision finding that an internet service provider could be held liable for defamation based on third-party content posted on its message boards. See, e.g., Roommates, 521 F.3d at 1163. In defamation law, the term “publication” means “communication [of the defamatory matter] intentionally or by a negligent act to one other than the person defamed.” Barnes, 570 F.3d at 1104 (internal quotation marks omitted). The Fourth Circuit has held that an internet service provider covered by the “traditional definition” of publisher is protected by section 230(c)(1). Zeran v. Am. Online, Inc., 129 F.3d 327, 332 (4th Cir.1997) (explaining that “every repetition of a defamatory statement is considered a publication,” and “every one who takes part in the publication is charged with publication”) (internal quotation marks and alterations omitted). And while the Ninth Circuit has declined to construe “the reach of section 230(c)(1) [as] fastened ... tightly to the nuances of defamation law,” the court has also indicated that the statute’s protections extend at least as far as the “treat[ment] [of] internet service providers as publishers ... for the purposes of defamation.” Barnes, 570 F.3d at 1104. Under this analysis, the private nature of Direct Messaging does not remove the transmission of such messages from the scope of publishing activity under section 230(c)(1).
Neither of the two decisions cited by plaintiffs calls for a different result. In Batzel, the Ninth Circuit addressed what it means for content to be “provided” by a third party, not what it means to “publish” third-party content online. See 333 F.3d at 1032-35. In F.T.C. v. Accusearch, Inc., No. 06-CV-00105, 2007 WL 4356786 (D.Wyo. Sept. 28, 2007), the court found that the defendants were not being treated as publishers where “the only parties that had access to the phone records at issue were [the defendants] and the particular [customer] that [purchased them].” Id. at *4. The court emphasized the “ill-gotten” nature of the phone records, and the commercial nature of the defendants’ use of them. Id. It also observed that the “argument [that defendants are being treated as publishers] might be more persuasive if the FTC sought to hold liable an internet service provider who, by virtue of the email-hosting services [it] provide[s], merely delivered an email containing ill-gotten consumer phone records.” Id. at *5 (emphasis added). That, of course, is essentially the situation here.
Meanwhile, a number of courts have applied the CDA to bar claims predicated on a defendant’s transmission of nonpublic messages, and have done so without questioning whether the CDA applies in such circumstances. See Hung Tan Phan v. Lang Van Pham, 182 Cal.App.4th 323, 324-28, 105 Cal.Rptr.3d 791 (2010); Delfino v. Agilent Techs., Inc., 145 Cal.App.4th 790, 795-96, 804-08, 52 Cal.Rptr.3d 376 (2006); Beyond Sys., Inc. v. Keynetics, *976Inc., 422 F.Supp.2d 523, 528, 536-37 (D.Md.2006).
Apart from the private'nature of Direct Messaging, plaintiffs identify no other way in which their Direct Messaging theory seeks to treat Twitter as anything other than a publisher of information provided by another information content provider. Accordingly, plaintiffs’ claims based on this theory are DISMISSED WITH LEAVE TO AMEND.
CONCLUSION
Twitter’s motion to dismiss is GRANTED. Plaintiffs shall file their second amended complaint, if any, within 20 days of the date of this Order.
IT IS SO ORDERED.
. Twitter's Direct Messaging capabilities allow Twitter users to communicate privately *968through messages that can be seen only by the people included on them, FAC ¶ 20.
. Congress enacted section 230(c)(1) in part to respond to a New York state court decision, Stratton Oakmont, Inc. v. Prodigy Servs. Co., 1995 WL 323710 (N.Y.Sup.Ct. May 24, 1995), finding that an internet service provider could be' held liable for defamation based on third-party content posted on its message boards. See Internet Brands, 2016 WL 3067995, at *5;Barnes, 570 F.3d at 1101;Roommates, 521 F.3d at 1163.
. Twitter submitted a Statement of Recent Decision regarding the opinion. Dkt. No. 34.
. While courts have not required plaintiffs bringing ATA claims based on material support theories to "trace specific dollars to specific attacks,” Strauss v. Credit Lyonnais, S.A., 925 F.Supp.2d 414, 433 (E.D.N.Y.2013); accord Linde v. Arab Bank, PLC, 97 F.Supp.3d 287, 328 (E.D.N.Y.2015) (on appeal), they have nevertheless rejected alleged causal connections that are too speculative or attenuated to raise a plausible inference of proximate causation, Terrorist Attacks, 714 F.3d at 123-25; Rothstein, 708 F.3d at 94-98.