The Fourth Circuit Remands Wikimedia’s Suit Against the NSA Back to District Court

By Jordan Brunner, Quinta Jurecic, Yishai Schwartz
Friday, May 26, 2017, 9:38 AM

In March of 2015, the Wikimedia foundation joined together with eight other non-profits in a challenge to NSA’s “mass surveillance” program. This week, the U.S. Court of Appeals for the Fourth Circuit permitted the suit to move forward, but in the process, peeled off all plaintiffs other than Wikimedia itself.

The case comes to the Fourth Circuit as an appeal from the the U.S. District Court of Maryland. In October 2015, the District Court dismissed the case for lack of standing, basing its decision on Supreme Court’s 2013 ruling in Clapper v. Amnesty International USA. In Clapper, the Court upheld the dismissal of a similar challenge from human rights groups because their complaint was speculative, alleging only a reasonable likelihood that their communications would be intercepted, not actual or imminent injury. Lacking concrete information about government programs, the Clapper plaintiffs failed to properly plead an injury in fact, and so lacked Article III standing. The same logic, held Judge Ellis of the Maryland District Court, should apply to Wikimedia’s challenge. This week, however, the Fourth Circuit rejected this comparison. Unlike Clapper, the court explained, the injuries alleged by the plaintiffs in this case are not speculative, but current, actual, and based on particularized allegations.

But the Fourth Circuit didn’t stop there and just vacate the entirety of Judge Ellis’ judgement. Instead, it distinguished between Wikimedia’s claim of injury and those of the other plaintiffs: Because Wikimedia’s claim of injury by NSA’s “Upstream” collection is based on particularized facts and sound inferences--construed in the light most favorable to Wikimedia for the purpose of the motion--it survives a motion to dismiss. By contrast, because the other plaintiffs’ allegation of injury (based on the allegation that NSA is “‘intercepting, copying, and reviewing substantially all’ textbased communications entering and leaving the United States, including their own”) is unsupported by “enough well-pleaded facts,” the District Court’s dismissal was proper.

Below, we first summarize the Fourth Circuit’s reasoning with respect to Wikimedia, and then turn to its analysis of the other eight plaintiffs’ claim of injury.

The Wikimedia Allegation: Not Speculative, and Well-pleaded

The factual assertions in the Wikimedia Allegation are based on a combination of public information about the operation of Section 702, PCLOB reports about what NSA is doing, as well as technical analysis how NSA must be accomplishing what PCLOB reports.

Under Section 702 of the Foreign Intelligence Surveillance Act, the government is permitted to target for surveillance non-US persons reasonably believe to be outside the United States. The procedures for making such determinations are reviewed and approved by the FISC. According to the plaintiffs, NSA conducts this surveillance by installing surveillance devices on at least some of the 49 international submarine cables that carry communication in and out of the United States which comprise the internet “backbone.” NSA, the plaintiffs suggest, lacks the technical capability to sift the communication prior to collection. It therefore must collect and copy “substantially all international text-based communications—and many domestic ones” as they flow across this ‘backbone’ in the United States”. Only after this initial collection, can NSA “attempt[] to filter out and discard some wholly domestic communications,” and then review and retain only those copied communications for that contain “targeted selectors” (such as specific IP or email addresses).

The essence of the Wikimedia allegation is that because of the technical realities of this “Upstream collection,” and because of the sheer volume of Wikimedia’s communications, the NSA has almost certainly collected at least some of the organization’s communications. Wikimedia reasons that because Upstream surveillance requires the NSA to copy even wholly domestic communications before filtering them out and discarding them, and because Wikimedia “engages in more than one trillion international communications each year, with individuals who are located in virtually every country on earth,” the NSA must necessarily be “intercepting, copying, and reviewing” some of Wikimedia’s communications.

Declassified documents show that a single service provider facilitates upstream surveillance at “seven major international chokepoints in the United States.” But given the quantity of its communication, even if the NSA is only collecting communication from a single Internet backbone link, Wikimedia asserts its communications must have been intercepted. Thus, Wikimedia’s “acute privacy interest in its communications,” are implicated by NSA programs.

The Fourth Circuit largely accepted Wikimedia’s arguments for the purpose of the motion. In order to establish Article III standing, a plaintiff must show “an injury in fact.” And in order to survive a motion to dismiss, a complaint must have “sufficient factual matter” in the complaint, such that the contents were accepted as true, it would “state a claim of relief that is plausible on its face.” Following the Third Circuit’s decision in Schuchardt v. President of the United States, the Fourth Circuit analyzed plaintiffs’ claims in two steps: first, it analyzed whether the allegations were “sufficiently particularized to satisfy the injury-in-fact requirement” and second, it analyzed whether the allegations contain sufficient detail to be “credited… as true for the purpose of resolving” a facial challenge to a complaint. Wikimedia’s claims, ruled the Court, met both criteria.

The court lays out what it sees as three presumptively key facts from Wikimedia’s allegation:

  1. Given the relatively small number of international chokepoints, the volume of its communications, and the geographical diversity of its users, Wikimedia almost certainly uses every international backbone link;
  2. To reliably obtain communications under Section 702 “about” collection, the NSA must to copy and review all text-based communications as a result of the technical functioning of the internet; and
  3. The NSA has confirmed that it conducts Upstream surveillance at more than one point along the internet backbone.

Taken together, these three points indicate that “Wikimedia has plausibly alleged that its communications travel all of the roads that a communication can take, and that the NSA seizes all the communications along at least one of those roads,” and therefore may have violated Wikimedia’s Fourth Amendment rights. Its allegations are thus predicated on specific asserted facts and directly implicate the potential interests of Wikimedia in a concrete and particularized manner.

Unlike the speculation and guesswork of Clapper, Wikimedia’s combination of technical assertions and government documents amounted to a properly pleaded complaint. And unlike Clapper, the Wikimedia case concerns a “motion to dismiss” rather than “a motion for summary judgment,” dictating greater deference to complainant's’ account of facts.

While acknowledging that Wikimedia’s probability calculation (“even if one assumes a 0.00000001% chance . . . of the NSA copying and reviewing any particular communication, the odds of the government copying and reviewing” one of Wikimedia’s communication in a one year period “would be greater than 99.999999999%”), was “incomplete and riddled with assumptions,” the court concluded that it wasn’t relevant for standing purposes. Importantly, given the “motion to dismiss” stage, the court also declined to consider the government’s evidence (in the form of expert affidavits) disputing plaintiff’s technical statements about NSA must operate. Without the opportunity to dispute the factual basis of the expert dispute, the government had little leg to stand on.

The Dragnet Allegation: Speculative, and Not Well-Pleaded

The other eight plaintiffs joined with Wikimedia in making the second allegation, which the court terms the “Dragnet Allegation.” The plaintiffs alleged that “in the course of conducting Upstream surveillance the NSA is ‘intercepting, copying, and reviewing substantially all’ text-based

communications entering and leaving the United States, including their own.” The core of the allegation is the assertion that due to the technical functionality of the internet, the NSA must be “intercepting, copying, and reviewing” information from most backbone chokepoints in order to engage effectively in Upstream collection. And if NSA is surveilling most backbone chokepoints, then it is likely that the plaintiffs’ communications have been examined. The plaintiffs allege that they have had to take “burdensome and sometimes costly measures” to protect themselves from this surveillance.

As with the Wikimedia Allegation, the district court dismissed the Dragnet Allegation under Clapper. The circuit court affirmed the dismissal, but on a different rationale, departing from the district court’s reliance on Clapper for the reasons expressed above.

While the court writes that its analysis of standing for the Wikimedia Allegation applies also to the Dragnet Allegation as far as the presence of a “particularized and cognizable ongoing injury,” traceability, and redressability go (hence why Clapper is not controlling), it ultimately finds that the plaintiffs lack standing due to their failure to plausibly state a claim. The court considers the Wikimedia Allegation plausible, but the broader claims made in the Dragnet Allegation are harder for it to swallow:

In the Dragnet Allegation, Plaintiffs must plausibly establish that the NSA is intercepting “substantially all” text-based communications entering and leaving the United States, whereas it’s sufficient for purposes of the Wikimedia Allegation to show that the NSA is conducting Upstream surveillance on a single backbone link.

The plaintiffs pointed to the same evidence for the Dragnet Allegation as they do for the Wikimedia Allegation, with the addition of one New York Times article on Upstream surveillance: in other words, the mechanical details of how the internet functions and the NSA’s stated goals of using Upstream collection to acquire information to, from, and about targets. But in the Wikimedia Allegation, the plaintiffs used that information to speculate about “the way the NSA does what we know it to be doing” (that is, engaging in Upstream collection). In contrast, in the Dragnet Allegation, the plaintiffs are using technical information and guesswork about NSA’s incentives to speculate about the scope of NSA activities. Furthermore, the allegations “fall short of the level of detail in Schuchardt,” in which the Third Circuit found that the plaintiffs had demonstrated the sheer scale of the collection suggested a dragnet

This is a bridge too far for the court, which finds this claim implausible and therefore holds that the plaintiffs lack standing on Fourth Amendment grounds. Following this logic, the court also dismisses the plaintiffs’ First Amendment claims of chilled speech and their effort to establish standing on the grounds of their “burdensome” efforts to avoid surveillance, finding that in the absence of a plausible claim, these concerns constitute “fears of hypothetical future harm” such as are inadequate to provide standing under Clapper.

The plaintiffs creatively cite Fourth Circuit precedent in the form of a two-year-old antitrust case, SD3, LLC v. Black & Decker, as evidence that motive is an important factor in establishing standing—pointing to NSA’s alleged “incentive to establish a dragnet.” The court dismisses this assertion, saying that while it “should come as no surprise that motive is an important factor in establishing an antitrust conspiracy,” that the court had never intended to have the case “stand for the broad proposition that motivation is always of special significance in plausibly pleading an injury.” The court also distinguishes SD3 based on the level of detail provided by the plaintiffs on the existence of the boycott, which the court concludes were “by and large absent,” from the plaintiffs’ complaint here.

The court concludes by addressing Judge Andre Davis’s dissent in part, specifically his assertion that the court need not have separately considered the non-Wikimedia plaintiffs’ standing. Given that “the complaint rests upon the premise that the NSA is seizing each Plaintiff’s unique communications,” the questions of standing and relief for the Wikimedia and non-Wikimedia plaintiffs are also individualized and must be considered separately.

Judge Andre Davis’s Dissent-in-Part

Judge Davis, while concurring with court’s finding that Clapper is not controlling and that Wikimedia has standing, dissented on the grounds that the non-Wikimedia plaintiffs do as well.

Davis explains that while he agrees with the majority’s decision to accept as plausible Wikimedia’s factual allegation, he disagrees with the majority’s assertion that the other plaintiffs have not plausibly alleged in the Dragnet Allegation that the NSA is surveilling most backbone links. He gives greater credence to the plaintiffs’ citation of the New York Times report to bolster their allegation, which the majority dismissed as essentially a restatement of the original allegation. More importantly, he argues that because of the technical functionality of the internet to which the plaintiffs point, NSA “cannot know which link the communications it targets will traverse when they enter or leave the United States,” and therefore “the only way it can comprehensively acquire its targets’ communications is by surveilling virtually every backbone link.” In his view, this allegation is a “logical extension” of the Wikimedia Allegation, and is therefore plausible as well.

In a footnote, Davis also criticizes the majority’s decision to assess the standing of the non-Wikimedia plaintiffs separate from that of Wikimedia. Quoting the Supreme Court’s decision in Horne v. Flores, he argues that “in all standing inquiries, the critical question is whether at least one petitioner has alleged such a personal stake in the outcome of the controversy as to warrant his invocation of federal-court jurisdiction.”