Strike 3 Holdings, LLC v. John Doe

United States Court of Appeals
FOR THE DISTRICT OF COLUMBIA CIRCUIT

Argued March 6, 2020 Decided July 14, 2020

No. 18-7188

STRIKE 3 HOLDINGS, LLC,
APPELLANT

v.

JOHN DOE, SUBSCRIBER ASSIGNED IP ADDRESS 73.180.154.14,
APPELLEE

Appeal from the United States District Court
for the District of Columbia
(No. 1:18-cv-01425)

Lincoln D. Bandlow argued the cause for appellant.
With him on the briefs were Jessica Fernandez and Emilie
Kennedy.

Seth W. Lloyd, appointed by the court, argued the cause
as amicus curiae in support of the District Court’s order. With
him on the brief were Samuel B. Goldstein, Deanne E.
Maynard, and James R. Sigel.

Before: MILLETT and RAO, Circuit Judges, and
EDWARDS, Senior Circuit Judge.

Opinion for the Court filed by Circuit Judge RAO.
 2
RAO, Circuit Judge: Strike 3 Holdings, LLC, is a producer
and distributor of adult films. This case arises out of a copyright
infringement suit filed by Strike 3 against a “John Doe”
defendant alleged to have illegally downloaded and distributed
Strike 3’s films. In its complaint, Strike 3 was able to identify
the defendant only with an Internet Protocol (IP) address linked
to numerous acts of online piracy. Strike 3 could not identify
the defendant and effectuate service without subpoenaing the
defendant’s Internet service provider (ISP) and so moved the
district court for a court order under Federal Rule of Civil
Procedure 26(d)(1). The district court denied the Rule 26(d)(1)
motion and dismissed Strike 3’s complaint without prejudice.

While district courts enjoy substantial discretion with
respect to discovery matters, that discretion is not unbounded.
We hold that the district court abused its discretion by
assigning improper weight to what it viewed as the “aberrantly
salacious nature” of Strike 3’s films, by concluding that Strike
3 could not state a plausible claim for infringement against the
IP address subscriber, and by drawing unsupported, negative
inferences against Strike 3 regarding its litigation tactics. We
therefore reverse the district court’s order and remand for
further proceedings consistent with this opinion.

I.

Strike 3 distributes pornographic films through various
adult websites and owns the copyrights to the motion pictures
distributed under four different brand names. According to
Strike 3, its websites receive more than 20 million unique
monthly visitors and tens of thousands of subscribers pay for
its content.

Like many film distributors, however, the company also
faces rampant online piracy. In an effort to combat illegal
downloading and distribution of its films, Strike 3 has filed
 3
thousands of copyright infringement lawsuits in district courts
around the country.1 These lawsuits follow a consistent pattern:
Strike 3 works with a third-party forensic investigator to
monitor the BitTorrent peer-to-peer file sharing network for the
presence of Strike 3’s copyrighted works. See AF Holdings,
LLC v. Does 1–1058, 752 F.3d 990, 998 (D.C. Cir. 2014)
(describing online piracy via the BitTorrent protocol). Utilizing
forensic software, the investigator can record a specific IP
address engaging in a specific act of infringement at a specific
time. Once an infringing IP address is identified, Strike 3 uses
geolocation technology to determine where the registered
subscriber of that IP address is located. Armed with this
information, Strike 3 files a copyright infringement lawsuit in
the appropriate federal district court, naming the “John Doe”
subscriber of the offending IP address as the defendant.

This appeal concerns a typical infringement lawsuit filed
by Strike 3. Using the technology described above, Strike 3’s
investigators recorded IP address 73.180.154.14 illegally
distributing Strike 3’s films via the BitTorrent network on
twenty-two separate occasions over the course of
approximately one year. Strike 3 determined that this IP

1
Strike 3 has also sought to reduce piracy through other means,
including by utilizing the procedures set out in the Digital
Millennium Copyright Act (DMCA). Under the DMCA, a copyright
owner who believes material posted online infringes its copyright
may file a takedown notice with the relevant ISP, requesting that the
ISP remove the allegedly infringing material from its web server. See
generally 17 U.S.C. § 512. Strike 3 states that it “send[s] on average
50,000 Digital Millennium Copyright Act notices a month” to ISPs
alerting them of infringement by their customers. Lansky Decl. ¶ 26,
App. 16. According to Strike 3, these efforts have proven futile, as
have “similar efforts by both the motion picture and recording
industries” to utilize DMCA takedown notices to combat piracy.
Strike 3 Br. 1.
 4
address is registered to a subscriber located in the District of
Columbia. In June 2018, the company filed a complaint against
the IP address subscriber in the U.S. District Court for the
District of Columbia. Because Internet service providers are
the only entities that can link an IP address to its subscriber,
Strike 3 could not serve its complaint without first subpoenaing
the subscriber’s ISP, Comcast, for information identifying the
anonymous defendant. Accordingly, the company also filed a
Rule 26(d)(1) motion seeking leave to subpoena Comcast for
records identifying the John Doe IP address subscriber.

In its Rule 26(d)(1) motion, Strike 3 argued it has good
cause to issue this subpoena because its lawsuit cannot proceed
without the requested discovery and because it would be
improper for the court to dismiss the claims before Strike 3 has
the opportunity to engage in discovery to learn the defendant’s
identity. Strike 3 further averred that it limits infringement
lawsuits to “strong cases against extreme infringers” who “not
only engage in illegal downloading, but are also large scale
unauthorized distributors of Strike 3’s content.” App. 34.
Finally, the company stated it is mindful of the sensitive nature
of the litigation and encouraged the district court to issue a
protective order allowing the defendant to proceed
anonymously.

The district court denied Strike 3’s discovery motion. See
Strike 3 Holdings, LLC v. Doe, 351 F. Supp. 3d 160 (D.D.C.
2018). Applying a multi-factor balancing test adopted by the
Second Circuit in Arista Records, LLC v. Doe 3, 604 F.3d 110
(2d Cir. 2010), the district court found that Strike 3’s need for
the subpoenaed information was outweighed by the
“potentially-noninfringing defendant’s right to be
anonymous”—a privacy interest the court found especially
weighty given the “particularly prurient pornography” at issue.
Strike 3, 351 F. Supp. 3d at 163, 164–65. The court
 5
acknowledged Strike 3’s lawsuit could not proceed without the
requested discovery, but explained that “Strike 3’s flawed
identification method” of relying on geolocation technology
and IP address tracking to identify infringers “fail[ed] to give
the court adequate confidence this defendant actually did the
infringing.” Id. at 164. In particular, the court emphasized the
risk that someone other than the IP address subscriber with
access to the IP address—such as a family member or
roommate—may have been responsible for the alleged
infringement. Id. at 162.

While the district court recognized that “honest copyright
holder[s] might balk” at its decision, which could curtail efforts
to combat online piracy through direct infringement claims, it
expressly cabined its holding to the “exceptional”
circumstances of this case. Id. at 165–66. The court pointed to
two factors limiting its holding: First, the “aberrantly salacious
nature” of Strike 3’s films, and second, the “legion pitfalls
associated with Strike 3’s tracking and identification of
infringers.” Id. at 165. “Given these high stakes,” the court
refused to “accept the risk of misidentification” and denied
Strike 3’s Rule 26(d)(1) motion. Id. But the court left open the
possibility that a future copyright holder could obtain a
different result by “show[ing] the Court a method to identify
infringers with sufficiently less risk of false accusations.” Id.
The district court then explained that Strike 3’s resulting
“inability to identify [the] defendant [made] effectuating
service or prosecuting the case impossible.” Id. Accordingly, it
dismissed the complaint without prejudice.

Throughout its opinion, the district court expressed
skepticism towards Strike 3’s motivation in suing the John Doe
defendant. The court characterized Strike 3 as a “copyright
troll” that has “flood[ed]” district courts around the country
with thousands of lawsuits “smacking of extortion,” and
 6
declared that it would not indulge Strike 3’s “feigned desire for
legal process” by “oversee[ing] a high-tech shakedown.” Id. at
161–62, 166.

Strike 3 filed a timely appeal. In the absence of a named
defendant, we appointed amicus to present arguments in
support of the district court’s order. Amicus has fulfilled that
role admirably.

II.

Federal Rule of Civil Procedure 26(d)(1) provides in
relevant part that “[a] party may not seek discovery from any
source before the parties have conferred as required by Rule
26(f), except … when authorized … by court order.” FED. R.
CIV. P. 26(d)(1). In cases involving as-yet-unknown
defendants, in which the plaintiff cannot serve its complaint—
much less confer with the defendant—without obtaining
identifying information from a third party, “the only potential
avenue for discovery is [a court order under] Rule 26(d)(1).”
AF Holdings, 752 F.3d at 995. A district court’s discretion to
order discovery, whether before or after the parties have
conferred, is cabined by Rule 26(b)’s general limitations on the
scope of discovery. Under Rule 26(b), a party may obtain
discovery only of nonprivileged materials that are “relevant to
[the] party’s claim or defense and proportional to the needs of
the case.” FED. R. CIV. P. 26(b)(1). Courts are directed to assess
proportionality by reference to several factors, including “the
parties’ relative access to relevant information” and “the
importance of the discovery in resolving the issues [at stake in
the action].” Id.2

2
Prior to 2015, Rule 26(b) included a “good cause” standard for
court-ordered discovery. See FED. R. CIV. P. 26(b)(1) (2015) (“For
 7
As we have repeatedly recognized, district courts have
broad discretion over the structure, timing, and scope of
discovery. See, e.g., Hussain v. Nicholson, 435 F.3d 359, 363–
64 (D.C. Cir. 2006). Accordingly, we will overturn a district
court’s discovery ruling only in “unusual circumstances”
amounting to an abuse of discretion. SafeCard Servs., Inc. v.
SEC, 926 F.2d 1197, 1200 (D.C. Cir. 1991). At the same time,
we must take care to ensure that “discretionary choices are not
left to a court’s inclination, but to its judgment; and its
judgment is to be guided by sound legal principles.” United
States v. Taylor, 487 U.S. 326, 336 (1988) (citation and
quotation marks omitted). Where a district court rests its
discovery ruling “on an improper factor,” Radtke v. Lifecare
Mgmt. Partners, 795 F.3d 159, 164 (D.C. Cir. 2015) (citation
omitted), or an error of law, AF Holdings, 752 F.3d at 994, it
necessarily falls short of this standard.

In this case, Strike 3 contends that the district court abused
its discretion by relying heavily on the company’s litigation
history and the content of its films rather than the relevant legal

good cause, the court may order discovery of any matter relevant to
the subject matter involved in the action.”). The 2015 amendments
to the Rules eliminated Rule 26(b)’s good cause requirement and
replaced it with the overarching relevance and proportionality
standard discussed above. The district court nonetheless assumed
court-ordered discovery remains subject to a good cause requirement
and went on to analyze good cause through the lens of the Second
Circuit’s Arista Records test. Strike 3, 351 F. Supp. 3d at 162–63
(citing Arista Records, 604 F.3d at 119).
We find the district court’s denial of Strike 3’s Rule 26(d)(1)
motion was an abuse of discretion under any applicable standard.
Therefore we do not apply the Arista Records test, nor do we address
whether Arista Records sets out an appropriate framework for
analyzing such motions or whether the “good cause” standard
continues to apply under the current version of Rule 26.
 8
standards under Rule 26 and Rule 12(b)(6). We agree, and find
that three aspects of the district court’s analysis require us to
reverse its decision: First, the district court “relied on an
improper factor” by treating the content of Strike 3’s
copyrighted works as relevant to its entitlement to early
discovery. Radtke, 795 F.3d at 164. Second, the court
committed legal error when it concluded Strike 3 could not
state a plausible claim of infringement against the IP address
subscriber even if the requested discovery was granted. And
third, the court drew unsupported, negative inferences against
Strike 3 regarding its litigation tactics and motivation in
seeking the requested discovery. We take each issue in turn.

A.

We first consider Strike 3’s argument that the district court
gave improper weight to the content of Strike 3’s films in
denying its Rule 26(d)(1) motion. From the outset, the district
court made clear that its assessment of the Rule 26(d)(1)
motion was informed by the pornographic content of Strike 3’s
films. The district court analyzed Strike 3’s discovery motion
by balancing the parties’ relative interests. The court assigned
“great weight” to the John Doe’s privacy expectations in this
case, reasoning that the potential impact of a false accusation
on the defendant’s personal and professional life would be dire
in a case involving “particularly prurient pornography.” Strike
3, 351 F. Supp. 3d at 164–65. Given these purportedly “high
stakes,” which the court viewed as “unlikely to appear in more
typical [copyright] cases,” the court held that “the risk of
misidentification” was too great to justify granting Strike 3 the
requested discovery. Id. at 165. The court emphasized,
however, that its holding was limited by the fact that this case
involved “aberrantly salacious” content. Id. (“[T]he typical
 9
case does not involve pornography, nor is this even run-of-the-
mill porn.”).3

The district court suggested in effect that Strike 3’s ability
to defend its copyright turned on the content of its films. This
reasoning is not supported by the relevant legal standards. Even
assuming without deciding that privacy interests have
continuing relevance to the Rule 26(d)(1) analysis, see supra
note 2, none of the cases cited by the district court suggest that
privacy interests vary depending on the content of a particular
copyrighted work. Nor has amicus identified any authority for
the proposition that content-based distinctions bear on a
copyright plaintiff’s entitlement to discovery in general. To the
contrary, as the district court correctly recognized, “defendants
have little expectation of privacy in downloading and
distributing copyrighted [content] without permission.” Strike
3, 351 F. Supp. 3d at 164 (citation omitted). The mere fact that
a defendant may be embarrassed to have his name connected
to pornographic websites is not a proper basis on which to
diminish a copyright holder’s otherwise enforceable property
rights. Indeed, as one district court aptly observed, “while there
may be some social stigma attached to consuming pornography
... it is [nonetheless] the rare civil lawsuit in which a defendant
is not accused of behavior of which others may disapprove.”
Strike 3 Holdings, LLC v. Doe, 2019 WL 4745360, at *6
(D.N.J. Sept. 30, 2019) (citation and quotation marks omitted).

Amicus offers little substantive defense for the district
court’s focus on the content of Strike 3’s films, choosing
instead to downplay the impact the films’ content had on the
court’s analysis. See Amicus Br. 39 (“The district court applied
the same standard that courts apply for other content.”).

3
The district court did not provide any support for its conclusion that
Strike 3’s films are “aberrantly salacious.”
 10
Beyond that, amicus merely observes—incorrectly—that
“[t]here is nothing improper in recognizing that privacy
interests differ depending on whether the accusations involve
content from [pornographic websites] rather than from some
other source.” Id.

Basic copyright principles establish that a plaintiff’s
ability to defend its copyrights cannot turn on a court’s
subjective view of the copyrighted material. The fundamental
premise of copyright law is that the owner of a valid copyright
has a protectable property interest in its creative works. See,
e.g., Dowling v. United States, 473 U.S. 207, 216–17 (1985).
Once a copyright is registered, the owner “may seek judicial
enforcement of his property rights against subsequent
infringers.” Metro. Reg’l Info. Sys., Inc. v. Am. Home Realty
Network, Inc., 722 F.3d 591, 596 (4th Cir. 2013). And a
copyright holder’s ability to defend this property interest does
not turn on the subjective value judgments of a particular judge.
Cf. Bleistein v. Donaldson Lithographing Co., 188 U.S. 239,
251 (1903) (Holmes, J.) (“It would be a dangerous undertaking
for persons trained only [in] the law to constitute themselves
final judges of the worth of pictorial illustrations.”); Situation
Mgmt. Sys., Inc. v. ASP. Consulting LLC, 560 F.3d 53, 60 (1st
Cir. 2009).

Here, the district court found that “Strike 3’s alleged
ownership of an infringed copyright sets forth a prima facie
claim.” Strike 3, 351 F. Supp. 3d at 164. Having accepted Strike
3’s allegations of copyright ownership, the district court could
not weaken the property rights attached to that ownership by
 11
imposing a content-based restriction on Strike 3’s access to
discovery.4

For the reasons stated, we hold that the content of a
copyrighted work is per se irrelevant to a Rule 26(d)(1) motion
seeking discovery to identify an anonymous infringer. The
protections afforded by copyright law do not turn on a
copyright holder’s popularity or perceived respectability.
Accordingly, we find the district court abused its discretion by
factoring the pornographic content of Strike 3’s films into its
decision to deny Strike 3’s discovery request.

B.

Strike 3 also contends that the district court erred when it
found that Strike 3’s subpoena would not “identify a copyright
infringer who can be sued.” Strike 3, 351 F. Supp. 3d at 164.
By the district court’s account, merely identifying the IP
address subscriber would not enable Strike 3 to state a plausible
claim against that subscriber—rather, Strike 3 would have to
“resort[] to far more intensive discovery machinations
sufficiently establishing defendant did the infringing.” Id.
Because the district court believed Strike 3 would have no
prospect of surviving a Rule 12(b)(6) motion even if it learned
the subscriber’s identity, the court concluded that Strike 3
should not be permitted to discover the subscriber’s identity at
all.

This reasoning is flawed in two critical respects. As an
initial matter, it misstates the relevant inquiry when a plaintiff
seeks immediate, court-ordered discovery to identify an
4
Although the district court intimated that pornography may not be
entitled to copyright protection, Strike 3, 351 F. Supp. 3d at 165 n.5,
that question is not before us as the district court found that Strike 3
stated a valid claim of copyright ownership.
 12
anonymous defendant. At this stage, the court is not asked to
pass judgment on the strength of the plaintiff’s allegations
against the defendant, but to determine whether the plaintiff
should have the opportunity to name that defendant in the first
place. It is well established that plaintiffs are permitted to
proceed against John Doe defendants so long as discovery can
be expected to uncover the defendant’s identity. See Newdow
v. Roberts, 603 F.3d 1002, 1010–11 (D.C. Cir. 2010) (citation
omitted). That is precisely what Strike 3 seeks here: the
discovery necessary to reveal the defendant’s identity and to
begin litigating its infringement claims.

This is not to suggest that the quality of a plaintiff’s
allegations has no bearing on a Rule 26(d)(1) motion. As with
all discovery requests, courts must look carefully to the
complaint’s allegations to determine if the requested discovery
is relevant and proportional to the needs of the case. See FED.
R. CIV. P. 26(b)(1). In AF Holdings, for example, we rejected a
plaintiff’s attempt to rely on Rule 26(d)(1) to uncover the
identities of hundreds of John Doe defendants with no apparent
connection to the district in which the plaintiff had sued. 752
F.3d at 995–96. Given that the plaintiff could not make even a
threshold showing of personal jurisdiction, “there [was] little
reason to believe that the information sought [would] be
‘relevant to the subject matter involved in the action.’” Id. at
995 (quoting FED. R. CIV. P. 26(b)(1) (2015)). In other words,
if a plaintiff plainly has no “realistic chance of successfully
suing” the defendant, we will not allow the plaintiff to “abuse[]
the discovery process” by seeking irrelevant information. Id. at
997. But the mere possibility that an unnamed defendant may
defeat a complaint at a later stage is not a legitimate basis to
deny a Rule 26(d)(1) motion that otherwise satisfies Rule 26’s
discovery standards.
 13
The district court’s analysis also misconstrued the
pleading burden at the Rule 12(b)(6) phase. The court
maintained that Strike 3 would need to engage in “far more
intensive discovery … sufficiently establishing defendant did
the infringing” in order to state a plausible claim. Strike 3, 351
F. Supp. 3d at 164. At this stage, however, a plaintiff “need not
set out all of the precise facts on which the claim is based.” EIG
Energy Fund XIV, L.P. v. Petroleo Brasileiro, S.A., 894 F.3d
339, 345 (D.C. Cir. 2018) (citation omitted). Nor must the
plaintiff prove its ultimate claim. See Herron v. Fannie Mae,
861 F.3d 160, 173 (D.C. Cir. 2017) (“A Rule 12(b)(6) motion
to dismiss … does not require a court to assess the truth of what
is asserted or determine whether a plaintiff has any evidence to
back up what is in the complaint.” (citation and quotation
marks omitted)). To the contrary, “‘even if it strikes a savvy
judge that actual proof of [the alleged] facts is improbable, and
that a recovery is very remote and unlikely,’” the judge must
accept the complaint’s well-pleaded allegations as true and
“draw all reasonable inferences from those allegations in the
plaintiff’s favor.” Banneker Ventures, LLC v. Graham, 798
F.3d 1119, 1129 (D.C. Cir. 2015) (quoting Bell Atlantic Corp.
v. Twombly, 550 U.S. 544, 556 (2007)).

In this case, Strike 3 asserts that it has used a combination
of forensic and geolocation technology to tie a single IP
address, registered to a user in the District of Columbia, to
twenty-two acts of infringement on specified dates over the
course of a year. Based on these allegations, a court could
reasonably infer that someone with prolonged, continuous
access to this IP address was responsible for the alleged
infringement. Viewing the allegations in the light most
favorable to Strike 3, we think it at least plausible that the
registered IP address subscriber “actually did the infringing.”
Strike 3, 351 F. Supp. 3d. at 164.
 14
Amicus contends that the district court’s reasoning was
correct because there is an “obvious alternative explanation”
for the John Doe’s alleged conduct: that someone else with
access to the IP address in question committed the alleged
infringement. Amicus Br. 25–26 (quoting Twombly, 550 U.S.
at 567–68). It is undoubtedly true that individuals other than
the IP address subscriber may have been responsible for the
infringement at issue. On these facts, however, we do not find
this alternative explanation so obvious as to render Strike 3’s
claim against the subscriber facially implausible. Cf. In re U.S.
Office of Personnel Mgmt. Data Sec. Breach Litig., 928 F.3d
42, 57 (D.C. Cir. 2019) (distinguishing cases where a
defendant’s “obvious alternative explanations were necessarily
incompatible with the plaintiffs’ versions of events”).

For similar reasons, amicus’s reliance on the Ninth
Circuit’s decision in Cobbler Nevada, LLC v. Gonzales, 901
F.3d 1142 (9th Cir. 2018), is misplaced. In Cobbler Nevada, a
copyright holder’s complaint against an IP address subscriber
was dismissed under Rule 12(b)(6)—but only after the plaintiff
was permitted to subpoena the defendant’s ISP. Id. at 1145.
Indeed, the district court in Cobbler Nevada granted the
plaintiff leave not only to issue its subpoena, but to depose the
subscriber once he was identified. Id. This preliminary
discovery revealed that the IP address was registered to an
individual who operated an adult foster care home where
numerous residents and visitors had access to the same Internet
service and IP address. Id. at 1144–45. After these facts came
to light and the plaintiff failed to plead any additional facts
linking the defendant to the alleged infringement, the case was
dismissed. Under those circumstances, the Ninth Circuit found,
the defendant’s “status as the registered subscriber of an
infringing IP address, standing alone, does not create a
reasonable inference that he is also the infringer.” Id. at 1145.
 15
Here, by contrast, Strike 3 has not even been able to gather
such threshold facts about the alleged infringer. To be sure, the
requested discovery may ultimately cast doubt on Strike 3’s
claim against the IP address subscriber. For example, the
subpoenaed records could reveal that the IP address is
associated with a coffee shop, a public library, or, as in Cobbler
Nevada, a residential facility where numerous users access the
same Internet service. See id. at 1146. At that point, if Strike 3
is unable to plead additional facts tying the registered
subscriber to the alleged infringement, the scales may tilt
towards implausibility.5 At this stage, however, we cannot
know what Strike 3’s subpoena will uncover. The mere fact
that discovery may demonstrate that the subscriber is not the
proper defendant is no basis to close the courthouse doors
before Strike 3 can step inside.

Finally, the district court’s opinion implied that its
perception of Strike 3’s films may have colored not only its
assessment of the Rule 26(d)(1) motion, but the merits of Strike
3’s claims. The district court explicitly stated that it “w[ould]
not accept the risk of misidentification” in this particular
context. Strike 3, 351 F. Supp. 3d at 165. This suggests the court
may have taken a different view of the strength of Strike 3’s
allegations but for its perception of the “aberrantly salacious
nature” of Strike 3’s films. Id. For the same reasons set out in
Part II.A, supra, the plausibility of a plaintiff’s allegations
against an anonymous IP address subscriber does not vary
depending on the content of the copyrighted work at issue.

5
Counsel for Strike 3 conceded at oral argument that it would be
difficult to state a plausible claim against an IP address subscriber
associated with a public facility, and further maintained that the
company would “immediately dismiss” a lawsuit under those
circumstances. Oral Arg. Tr. 11:12–22.
 16
On the allegations before us, we cannot conclude there is
no “realistic chance” the discovery Strike 3 sought will yield
information relevant to its suit or that Strike 3 could not state a
plausible claim against the IP address subscriber if the suit
reaches the Rule 12(b)(6) phase. See AF Holdings, 752 F.3d at
997. Because the district court’s contrary conclusion was
predicated on an error of law, we hold the court abused its
discretion in denying Strike 3’s Rule 26(d)(1) motion.

C.

Lastly, we turn to Strike 3’s argument that the district court
failed to afford Strike 3 the benefit of all reasonable inferences,
and instead relied on extra-record sources to question Strike 3’s
motivation in seeking the requested discovery. We find the
district court abused its discretion in this regard as well.

As already explained, the district court was required to
assume the truth of Strike 3’s factual allegations and draw all
reasonable inferences in its favor. See Banneker Ventures, 798
F.3d at 1129. It failed to do so. Instead, the court went outside
the record to conduct an in-depth review of Strike 3’s publicly
available court filings, and relied on this research to conclude
that Strike 3 is engaged in a pattern of coercive litigation
designed to extract shame settlements rather than identify
actual infringers. Strike 3, 351 F. Supp. 3d at 162, 166. The
district court dubbed Strike 3 a “copyright troll,” id. at 161–62,
“[a]rmed with hundreds of cut-and-pasted complaints and
boilerplate discovery motions,” id. at 166. It accused Strike 3
of not caring “whether defendant actually did the infringing,”
and instead seeking simply to “settle as many claims as
possible [and] abandon the rest.” Id. at 162. And it found that
Strike 3’s Rule 26(d)(1) motion reflected a “feigned desire for
legal process” that “masks what it really seeks: for the Court to
oversee a high-tech shakedown.” Id. at 166.
 17
The facts before the district court did not support these
conclusions. It is true, as the district court pointed out, that
Strike 3 filed nearly 2,000 copyright lawsuits in courts around
the country in the thirteen months leading up to the district
court’s decision. Strike 3 does not dispute this figure. It argues,
however, that the volume of its litigation is dwarfed by the
scale of infringement it faces. Indeed, Strike 3 maintains that
although its films “are among the most pirated content in the
world,” Compl. ¶ 16, App. 4, it has pursued “less than 1 out of
1,000 US-based Strike 3 infringers” via litigation, Strike 3 Br.
19 n.18. Strike 3 further contends that no negative inference
can be drawn from the fact that many of these lawsuits settle or
otherwise do not proceed beyond the initial stages of litigation,
given the established principle that “parties are free to settle
their disputes on their own terms.” Frank v. Gaos, 139 S. Ct.
1041, 1046 (2019).

Amicus, on the other hand, argues that Strike 3’s
“litigation conduct across a large volume of cases and its
consistent pattern of never pursuing any case to judgment on
the merits” provided ample basis for the district court to infer
that Strike 3 sought discovery for a reason other than defending
its copyrights. Amicus Br. 29. According to amicus, the district
court acted appropriately in exercising its authority to take
judicial notice of Strike 3’s litigation pattern and in relying on
that pattern to infer that Strike 3 seeks hasty settlements rather
than successful litigation.

We do not dispute that district courts may properly take
judicial notice of proceedings and filings in other courts. See
Dupree v. Jefferson, 666 F.2d 606, 608 n.1 (D.C. Cir. 1981). A
court cannot, however, reasonably infer that a plaintiff lacks a
legitimate motive in pursuing discovery based solely on the
plaintiff’s litigation volume and case history. Where, as here, a
plaintiff alleges that it is the victim of copyright infringement
 18
on a massive scale, the mere fact that it has filed a significant
number of lawsuits is not a valid basis on which to impute an
improper purpose. Nor is the fact that many such lawsuits settle
or are dismissed at an early stage necessarily suggestive of
improper intent. Cf. Frank, 139 S. Ct. at 1046.

In its briefing and at oral argument, amicus relied heavily
on our decision in AF Holdings to support the district court’s
assessment of Strike 3’s litigation tactics. That case provides a
poor analogy. AF Holdings involved a law firm that multiple
courts had found to be associated with a long pattern of abusive
copyright litigation. See 752 F.3d at 992–93. The law firm in
question formed a holding company, acquired the copyrights to
several pornographic films, and “initiated massive ‘John Doe’
copyright infringement lawsuits” against hundreds of IP
address subscribers in a single district court in an effort to
minimize filing fees. Id. at 992. This particular pattern, coupled
with AF Holdings’ failure to tailor its discovery requests to
individuals who could conceivably be sued in the district, made
it reasonable for the court to conclude that the company sought
to “manipulate judicial procedures to serve [its] own improper
ends” rather than pursue a legitimate litigation interest. Id.

This case involves markedly different facts. Strike 3’s
filings aver that it produced the copyrighted materials at issue,
persistently seeks to protect its copyrights through the
procedures set out in the Digital Millennium Copyright Act,
and resorted to civil litigation only after countless DMCA
takedown notices failed to stem the tidal wave of infringement.
See Lansky Decl. ¶¶ 26–27, App. 16 (noting that DMCA
notices have “do[ne] virtually nothing to stop the rampant
copyright infringement,” making “lawsuits like this one” the
“only effective way to stop … piracy”). Unlike the scattershot
approach undertaken in AF Holdings, Strike 3 utilized
technology to target individual serial infringers in the
 19
appropriate district court. At this stage of the litigation, Strike
3 was entitled to have the court accept these allegations as true
and draw all reasonable inferences in Strike 3’s favor.

While the district court did not explicitly refer to Strike 3’s
litigation conduct when analyzing the Rule 26(d)(1) motion,
the court’s perception of Strike 3’s motives appears to have
colored its analysis. Indeed, the court took pains to emphasize
that a different result might obtain in a case involving an
“honest copyright holder,” Strike 3, 351 F. Supp. 3d at 165, and
derided Strike 3’s “feigned desire for legal process” in
summarizing its holding, id. at 166. The factual record did not
support the negative inferences drawn. We therefore hold that
the district court abused its discretion by failing to apply the
appropriate legal standards to Strike 3’s complaint.

* * *

For the foregoing reasons, the district court’s denial of
Strike 3’s Rule 26(d)(1) motion must be reversed. The district
court’s decision to dismiss Strike 3’s complaint without
prejudice was predicated entirely on its denial of the Rule
26(d)(1) motion. See id. at 165 (“Denying Strike 3’s discovery
motion dooms its claim.”). Because we find that the district
court abused its discretion in denying Strike 3’s discovery
motion, its dismissal for failure to state a claim is also reversed.

While we do not lightly intercede in a district court’s
management of the discovery process, this case represents the
rare circumstance in which the district court exceeded its
substantial discretion. We therefore reverse the district court’s
order and remand for further proceedings consistent with this
opinion.

So ordered.