Woodhull Freedom Foundation v. United States of America

UNITED STATES DISTRICT COURT
FOR THE DISTRICT OF COLUMBIA
WOODHULL FREEDOM )
FOUNDATION, et al., )
)
Plaintiffs, )
)
V. ) Civil Case No. 18-1552 (RJL)
)
UNITED STATES OF AMERICA, )
et al., )
)
Defendants. )
MEMONANDUM OPINION
(March 29 _, 2022) [Dkt. ## 34, 35]
Plaintiffs—comprised of “advocacy and human rights organizations, two
individuals and the leading archival collection of Internet content,” Compl. § 2 [Dkt.
# 1]'—have brought this action challenging the constitutionality of the Allow States and
Victims to Fight Online Sex Trafficking Act of 2017, Pub. L. No. 115-164, 132 Stat. 1253(2018) (‘FOSTA” or “the Act’). Plaintiffs raise a number of constitutional claims,
including that FOSTA violates the First and Fifth Amendments to the U.S. Constitution,
as well as the Ex Post Facto clause of Article I, Section 9. See id. FJ 126-174. More
specifically, plaintiffs assert that FOSTA is overbroad and vague in multiple ways; that it
discriminates against certain speech based on its content and viewpoint; and that it
' The plaintiffs are: the Woodhull Freedom Foundation (an advocacy and lobbying organization for sexual
freedom); Human Rights Watch (a generalized human rights monitoring and advocacy organization); Jesse Maley
(also known as Alex Andrews, who founded and manages a website, “Rate that Rescue,” which provides resources
and information to sex workers); Eric Koszyk (a licensed massage therapist who relies on Craigslist for advertising);
and the Internet Archive (an internet service that, unsurprisingly, focuses on preserving websites to prevent their
erasure or disappearance with the passage of time). See Compl. ff 15-19.
\ 1
impermissibly imposes liability on conduct that was lawful at the time it occurred. Jd.
Plaintiffs seek declaratory relief stating that FOSTA is unconstitutional as well as an
injunction prohibiting enforcement of the Act. Defendants, the United States and the
Attorney General (“defendants” or “the Government”), disagree and argue that the statute
suffers from none of the claimed constitutional infirmities and thus plaintiffs are not
entitled to the relief they seek. Before me are the parties’ cross-motions for summary
judgment. Upon consideration of the pleadings, the record, and the relevant law, and for
the reasons stated below, I agree with the Government that plaintiffs’ challenges to FOSTA
are without merit. As a result, the defendants’ cross-motion for summary judgment is
GRANTED, plaintiffs’ motion for summary judgment is DENIED, and this case is hereby
dismissed.
BACKGROUND
I. Statutory Scheme
FOSTA was enacted in April 2018 and took immediate effect. Pub. L. No. 115-
164, § 4(b), 132 Stat. at 1253. It begins by memorializing the “sense of Congress” that the
Communications Decency Act of 1996, as codified at 47 U.S.C. § 230 (“Section 230”),
“was never intended to provide legal protection to websites that unlawfully promote and
facilitate prostitution and websites that facilitate traffickers in advertising the sale of
unlawful sex acts with sex trafficking victims.” FOSTA § 2(1), 132 Stat. at 1253.2 Further,
Congress found that “websites that promote and facilitate prostitution have been reckless
* The 1996 Communications Decency Act itself amended the Communications Act of 1934.
2
in allowing the sale of sex trafficking victims and have done nothing to prevent the
trafficking of children and victims of force, fraud, and coercion.” Jd. § 2(2). As a result,
the Act notes, Congress concluded that “clarification of [Section 230] is warranted” to
ensure that it does not shield “such websites” from appropriate liability. Jd. § 2(3).
To this end, the Act adds one section to the U.S. Code while amending three others.
First, FOSTA enacted 18 U.S.C. § 2421A, which creates a federal criminal offense for
owning, managing, or operating “an interactive computer service . . . with the intent to
promote or facilitate the prostitution of another person,” or attempting or conspiring to do
so. 18 U.S.C. § 2421A(a). This offense is punishable by fine or imprisonment for a term
of up to ten years. Jd. FOSTA further allows that any defendant facing this charge may
raise an affirmative defense that “the promotion or facilitation of prostitution is legal in the
Jurisdiction where the promotion or facilitation was targeted.” Jd. §2421A(e). The
defendant bears the burden of establishing this affirmative defense by a preponderance of
the evidence. /d.
FOSTA also creates an “aggravated” version of this offense, punishable by a fine or
a term of imprisonment of up to twenty-five years. Jd. §2421A(b). This aggravated
offense requires proof of an additional element on top of those already required to be
convicted for the base offense contained in § 2421A(a). Therefore, liability under
§ 2421A(b) may attach only to a defendant who owns, manages, or operates an interactive
computer service with the intent to promote or facilitate the prostitution of another
person—the base offense—and either (1) “promotes or facilitates the prostitution of five
or more persons,” see id. § 2421A(b)(1), or (2) “acts in reckless disregard of the fact that
such conduct contributed to sex trafficking[] in violation of 18 U.S.C. § 1591(a),” see id.§ 2421A(b)(2). Section 1591 (a) is a preexisting provision of the criminal law that prohibits
sex trafficking. See id. § 1591(a). FOSTA also directs that restitution be imposed by the
court for any violation of subsection (b)(2). Lastly, § 2421A(c) provides that victims of
violations of § 2421A(b) may bring civil suits in federal court to “recover damages and
reasonable attorneys’ fees.” Jd. § 2421A(c).
Section 4 of FOSTA proceeds to explicate and amend the scope of Section 230.°
Mote specifically, it clarifies the preemptive effect of Section 230, stating that “nothing in”
Section 230(c)(1)—a provision that immunizes providers of interactive computer services
from liability for the speech of third parties—‘shall be construed to impair or limit” three
categories of civil claims and criminal prosecutions. Jd. § 230(e)(5). Those claims include
federal civil claims brought pursuant to 18 U.S.C. § 1591; criminal prosecutions brought
pursuant to state law where the underlying conduct would be a violation of 18 U.S.C.
§ 1591; and criminal prosecutions brought pursuant to state laws where the underlying
conduct would be a violation of 18 U.S.C. § 2421A (the newly added criminal provision
3 As I described in my previous opinion in this case, see Woodhull Freedom Fdn. v. United States (Woodhull
J), 334 F. Supp. 3d 185, 190-91 (D.D.C. 2018), Section 230 has two principal functions. First, the statute immunizes
interactive computer services from liability for content created by third parties. See 47 U.S.C. 230(c)(1) (providing
that “[n]o 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”); id. § 230(e)(3) (preempting conflicting state and
local law). At the same time, Section 230 further: encourages such services to undertake “Good Samaritan” content
screening by precluding civil liability for “any action voluntarily taken in good faith to restrict access to or availability
of material that the provider or user considers to be obscene, lewd, lascivious, filthy, excessively violent, harassing,
or otherwise objectionable, whether or not such material is constitutionally protected,” as well as for actions made to
enable or make available technical means of restricting access to such materials. See id. § 230(c)(2); see also Bennett
v. Google, LLC, 882 F.3d 1163, 1165 (D.C. Cir. 2018) (noting that Section 230 “encourage[s] service providers to
self-regulate the dissemination of offensive material over their services”),
4
discussed above), and where the affirmative defense contained in § 2421A(e) does not
apply. See id.; see also FOSTA § 4(a), 132 Stat. at 1254. Section 4 of the Act lastly
provides that its amendments “shall take effect on the date of the enactment of this Act,”
and that the above-described changes to Section 230 preemption “shall apply regardless of
whether the conduct alleged occurred, or is alleged to have occurred, before, on, or after
such date of enactment.” FOSTA § 4(b), 132 Stat. at 1264-65.
Next, FOSTA amends 18 U.S.C. § 1591, the provision of the U.S. Code that
prohibits sex trafficking, by inserting a new definition. The new definition clarifies that
the term “participation in a venture” as used in that section means “knowingly assisting,
supporting, or facilitating” sex trafficking. See id. §5, 132 Stat. at 1265; 18 U.S.C.
§ 1591(e)(4). The term “participation in a venture” appears earlier in the same section but
previously was left undefined. See 18 U.S.C. § 1591(a)(2) (criminalizing the knowing
“participation in a venture” to cause sex trafficking of an adult by “force, fraud, or
coercion” or of a minor). Lastly, FOSTA amends 18 U.S.C. § 1595 by authorizing state
attorneys general to bring civil actions on behalf of residents of the state who have been
“threatened or adversely affected by any person who violates” 18 U.S.C. § 1591. See 18
US.C. § 1595(d).4
* FOSTA also includes a so-called savings clause (clarifying that the Act should not be construed to limit or
preempt prosecutions or civil actions that were previously not limited or preempted by Section 230), as well as a
requirement that the Government Accountability Office produce a study on cases brought pursuant to FOSTA’s
amendments, see FOSTA 8§ 7, 8, 132 Stat. at 1255-56, but these provisions are not directly at issue in this case.
5
Il. Procedural History
Shortly after FOSTA’s enactment, plaintiffs filed this lawsuit against the United
States and the U.S. Attorney General, seeking declaratory relief and to enjoin enforcement
of FOSTA due to its alleged facial constitutional defects. I described the plaintiffs and
their stated interests at some length in my previous opinion in this case, see Woodhull I,
334 F. Supp. 3d 185, 192-94, but for reference they are briefly recounted here. They
include three organizations: the Woodhull Freedom Foundation (an advocacy and lobbying
organization for sexual freedom), Human Rights Watch (a generalized human rights
monitoring and advocacy organization), and the Internet Archive (an internet service that,
unsurprisingly, focuses on preserving websites to prevent their erasure or disappearance
with the passage of time). There are also two individual plaintiffs: Jesse Maley (also known
as Alex Andrews), who founded and manages a website, “Rate that Rescue,” which
provides resources and information to sex workers and which, she alleges, could face
prosecution under her reading of FOSTA’s criminal provisions; and Eric Koszyk, a
licensed massage therapist who states that he relies on Craigslist for a large portion of his
advertising and had been harmed by Craigslist removing massage advertisements in the
aftermath of FOSTA’s enactment. See generally id.; see also Compl. {¥ 15-19.
Plaintiffs moved for a preliminary injunction concurrent with the filing of their
complaint. See Pls.’ Mot. for Prelim. Inj. [Dkt. #5]. The Government thereafter opposed
that motion and moved to dismiss the case for lack of subject matter jurisdiction,
contending that each of the plaintiffs lacked standing to bring a facial, pre-enforcement
constitutional challenge to the Act because they could not demonstrate a credible threat of
prosecution under the act, and thus lacked the injury requisite for Article III standing. See
Defs.’ Mem. in Opp. to Pls.’ Mot. for Prelim. Inj. and Mot. to Dismiss [Dkt. # 15]. I agreed
with the defendants and granted the motion to dismiss. See Woodhull I, 334 F. Supp. 3d at
198-203.
Plaintiffs appealed from that decision, and our Court of Appeals reversed, finding
that at least the two individual plaintiffs, Andrews and Koszyk, had sufficiently established
their Article III standing such that this case could proceed. See Woodhull Freedom Fadn. v.
United States (Woodhull II), 948 F.3d 363, 371-73 (D.C. Cir. 2020). Because both parties
contend that the reasoning behind this holding on standing bears on the merits of plaintiffs’
claims on remand, I provide a brief description of the opinion of the panel majority and the
partial concurrence here:
The majority held that Andrews had established Article III standing to mount a pre-
enforcement challenge to FOSTA under the standard clarified by the Supreme Court in
Susan B. Anthony List v. Driehaus (“SBA”), 573 U.S. 149 (2014). In SBA, the Court held
that pre-enforcement standing in this context required a prospective plaintiff to show that
they intended to engage in conduct “arguably affected with a constitutional interest,” that
the conduct was “arguably proscribed by [the] statute,” and that “the threat of future
enforcement of the [statute] was substantial.” Woodhull IT, 948 F.3d at 371 (quoting SBA,
573 U.S. at 164). As relevant here, the panel concluded that Andrews, by alleging an intent
to continue operating Rate that Rescue, had met this bar: more specifically, the operation
of Rate that Rescue was “arguably” proscribed by FOSTA under either plaintiffs’ broad
reading of 18 U.S.C. § 2421A or the narrower reading endorsed by the defendants. Jd. at
120-21. Additionally, the panel held that Koszyk had made a sufficient showing as to
redressability, concluding that he had established that a favorable decision could prompt
Craigslist to reinstate advertising it removed following FOSTA’s enactment. Jd. at 122.
Judge Katsas, meanwhile, concurred in part and with the judgment, clarifying the
scope of the panel’s holding as to the scope of FOSTA and specifically 18 U.S.C. § 2421A.
See id. at 122-24 (Katsas, J., concurring in part and concurring in the judgment). In so
doing, Judge Katsas rejected plaintiffs’ reading of § 2421A, “which,” he stated, “ignores
or overreads all the key statutory terms.” Jd. at 123. He noted that the majority “neither
adopt[ed] this construction of FOSTA nor follow[ed] it to its logical conclusion,” but that
instead the majority had “identiflied] this construction as at least one possible reading of
FOSTA.” Jd. Despite rejecting the plaintiffs’ reading of the statute and instead adopting
a reading consistent with that endorsed in my earlier opinion, see Woodhull I, 334 F. Supp.
3d at 199-201, Judge Katsas concluded that at least some of the alleged activities
undertaken on Andrews’s platform—such as the provision of specific information about
payment processors—arguably could be proscribed by FOSTA even under the narrower
reading, and thus Andrews had standing. Jd.
On remand, the parties filed cross-motions for summary judgment, see Pls.’ Mot.
for Summ. J. [Dkt. #34]; Defs.’ Mot. for Summ. J. [Dkt. #35], which are now pending
before the Court.
LEGAL STANDARD
The Court may enter summary judgment for a moving party “if the movant shows
that there is no genuine dispute as to any material fact and the movant is entitled to
judgment as a matter of law.” Fed. R. Civ. P. 56(a). The parties agree that there are no
facts in dispute material to the merits of plaintiffs’ facial constitutional claims, see Defs.’
Mem. in Opp. to Pls.’ Mot. for Summ. J. at 44 [Dkt. # 37]; Pls.’ Reply to Defs.’ Opp. at 24
[Dkt. #39], and thus the Court may enter judgment based solely on FOSTA’s statutory
language.” See, e.g., Ezell v. City of Chi., 651 F.3d 684, 698 (7th Cir. 2011).
ANALYSIS
Plaintiffs muster six distinct bases on which to conclude that FOSTA is facially
unconstitutional: first, that the statute is “overbroad” in violation of the First Amendment,
chilling constitutionally protected speech undertaken by the plaintiffs and others; second,
that FOSTA is impermissibly vague such that it does not provide fair notice of what it
prohibits, violating the due process clause of the Fifth Amendment; third, that FOSTA
violates the First Amendment by selectively removing Section § 230 immunity for the
hosting of content while leaving intact immunity for removal of content; fourth, that
FOSTA violates the First Amendment by discriminating against certain types of speech
and lacking a sufficient justification for such discrimination to survive strict scrutiny; fifth,
> In conjunction with their motion for summary judgment, plaintiffs did submit a statement of facts
accompanied with a number of supporting affidavits. See Pls.’ Statement of Facts & Ex. A-I [Dkt. # 34-2 to -15].
However, these facts and affidavits are material only to establishing plaintiffs’ ongoing standing—which defendants
do not challenge—and the entitlement of plaintiffs to injunctive relief should they prevail on the merits. Because, as
explained below, I find that plaintiffs’ facial constitutional claims are without merit, there is no need to address the
facts underpinning plaintiffs’ request for injunctive relief.
9
that FOSTA lacks a sufficient scienter requirement in either § 2421A or the amendments
to § 1591, leading to an impermissibly broad sweep of liability; and sixth and finally, that
FOSTA’s amendments to Section 230 permitting liability for conduct occurring prior to
the statute’s enactment violate the Ex Post Facto Clause of the Constitution. As discussed
in detail below, I find that plaintiffs have failed to establish that FOSTA is unconstitutional
for any of the first five reasons listed. Moreover, on plaintiffs’ sixth claim, the Court cannot
provide any meaningful remedy to the plaintiffs notwithstanding any merit to their claim,
as no named defendant could possibly be in a position to enforce the challenged provision
of FOSTA. Plaintiffs’ sixth claim therefore must be dismissed as well.
I. Overbreadth
Plaintiffs’ primary challenge to FOSTA is that it is facially overbroad, in TiOlaiGh
of the First Amendment. More specifically, plaintiffs argue that the statute’s language—
in particular the newly added § 2421 A—is broad enough to encompass, and therefore chill,
a large swath of First Amendment-—protected speech undertaken by the plaintiffs and others.
For example, plaintiffs contend that their advocacy work or the discussions conducted in
forums hosted on websites like Rate That Rescue constitute protected speech while at the
same time coming within the Act’s prohibition on “promot[ing]” or “facilitat[ing]’”
prostitution. See generally Pls.” Mem. in Support of Mot. for Summ. J. at 19-22 [Dkt.
# 34-1].
The doctrine of overbreadth is unique to the First Amendment context, allowing
plaintiffs to bring a facial challenge to government regulation of “spoken words” as well
10
as other “expressive or communicative conduct” based on the chilling effect of that
regulation on others. Broadrick v. Oklahoma, 413 U.S. 601, 612-13 (1973). Though it
has approved of such challenges over time, the Supreme Court has simultaneously been
careful to note that granting a claim of overbreadth is “‘strong medicine’ [that should be]
employed . . . with hesitation, and then ‘only as a last resort.’” L.A. Police Dep’t v. United
Reporting Pub. Corp., 528 U.S. 32, 39 (1999) (quoting New York v. Ferber, 458 U.S. 747,
769 (1982)). In particular, the Court has noted that “invalidating a law that in some of its
applications is perfectly constitutional—particularly a law directed at conduct so antisocial
that it has been made criminal—has obvious harmful effects.” United States v. Williams,
553 U.S. 285, 292 (2008). For this reason, the Supreme Court has “vigorously enforced
the requirement that a statute’s overbreadth be substantial, not only in an absolute sense,
but also relative to the statute’s plainly legitimate sweep,” for the statute to be declared
facially unconstitutional on this basis. Jd.
In undertaking this analysis of whether the challenged statute sweeps substantially
too far, the “first step” for this Court is “to construe the challenged statute” because “‘it is
impossible to determine whether a statute reaches too far without first knowing what the
statute covers.” Id. at 293; see also United States v. Stevens, 559 U.S. 460, 473 (2010)
(noting, in overbreadth context, that “the constitutionality of [the challenged statute] hinges
on how broadly it is construed”); United States v. Hillie, 14 F.4th 677, 685 (D.C. Cir. 2021)
(noting that proper interpretations of challenged statutes “were necessary antecedents to
determining whether the statutes at issue . . . were overbroad”). Unfortunately for the
11
plaintiffs, it is at this first stage that their overbreadth claim runs into trouble, because that
claim is premised on a reading of FOSTA that is untethered from the language of the statute
or its context. As explained in more detail below, proper application of the tools of
statutory construction to FOSTA’s language instead leads to the conclusion that FOSTA’s
“plainly legitimate sweep”—the range of its hypothetical applications covering only
unprotected activity—comprises all, or at worst the vast majority, of potential applications
of the statute. As a result, when correctly interpreted, FOSTA cannot reasonably be labeled
overbroad, and plaintiffs’ claim to the contrary must be rejected. How so?
Proper interpretation of FOSTA begins with its text. And because plaintiffs’
overbreadth challenge centers on the newly enacted § 2421A, the analysis starts there. As
stated above, that section imposes liability on anyone who “owns, manages, or operates an
interactive computer service . . . with the intent to promote or facilitate the prostitution of
another person.” See 18 U.S.C. § 2421A(a); see also FOSTA § 3(a), 132 Stat. at 1253.
Plaintiffs argue, in essence, that the intent element in § 2421A is vaguely worded and
accordingly sweeps far too wide. With neither “promote” nor “facilitate” expressly defined
in the Act, plaintiffs turn immediately to those verbs’ bare dictionary definitions, i.e., “to
make easier,” see Woodhull IT, 948 F.3d at 372, to claim that liability under FOSTA could
be imposed on any internet service that would make prostitution easier in any way.
However, as explained in my prior opinion in this case concerning plaintiffs’ standing, as
well as in Judge Katsas’s concurrence on the same issue, “promote” and “facilitate” are not
verbs inserted into the U.S. Code without context. Indeed, it is that context—both from
12
within FOSTA itself and from background criminal-law interpretive principles—that serve
to narrow these terms such that the Act itself is not overbroad. Cf United States v.
Abuelhawa, 556 U.S. 816, 819-20 (2009) (“[B]ecause statutes are not read as a collection
of isolated phrases . . . [a] word in a statute may or may not extend to the outer limits of its
definitional possibilities.” (citations omitted)).
First, the relevant promotion and facilitation that is encompassed by the statute is
not that of the general concepts of sex work or prostitution, as plaintiffs contend. Instead,
as I noted in my previous opinion, FOSTA contains textual indications that make it quite
clear that the Act targets the promotion and facilitation only of specific instances of
prostitution: it is not a crime to operate an internet service with the intent to “promote...
prostitution” as a general matter, as plaintiffs would read it, but instead to promote (or
facilitate) the “prostitution of another person.” See Woodhull I, 334 F. Supp. 3d at 200;
see also Woodhull IT, 948 F.3d at 375 (Katsas, J., concurring in part and concurring in the
judgment) (“FOSTA focuses not on prostitution as an abstract legal or policy matter, but
on ‘the prostitution of another person’—a widely criminalized act involving the exchange
of sex for money... .”).°
6 This conclusion is only strengthened by the existence of FOSTA’s affirmative defense, which allows a
defendant to avoid liability on the basis that “promotion or facilitation of prostitution” is legal in the jurisdiction at
which that conduct was targeted; because the forms of generalized advocacy, harm-reduction, and other discussion
that plaintiffs allege are encompassed by the language of § 2421A could not be illegal in any jurisdiction—and
plaintiffs certainly have not tried to contend as much—-it is either the case that this affirmative defense swallows the
entire statute, or that “promotion or facilitation of prostitution” must be a narrower category of activity than plaintiffs
contend.
13
Moreover, in this statutory context, the verbs in question, “promote” and
“facilitate,” are not necessarily to be read as coterminous with their dictionary definitions.
See Abuelhawa, 556 U.S. at 819-20. As Judge Katsas noted, “promotion” of prostitution
is an established term in the criminal law, meaning “to pander or pimp” such as by
“recruiting a prostitute or soliciting prospective customers.” Woodhull IT, 948 F.3d at 375
(Katsas, J., concurring in part and concurring in the judgment) (citing Pandering, BLACK'S
LAW DICTIONARY, (10th ed. 2014) (“The act or offense of recruiting a prostitute, finding
a place of business for a prostitute, or soliciting customers for a prostitute —Also termed
promoting prostitution.”)). And the Supreme Court has previously interpreted the verb
“facilitate” in the criminal law context, noting that in such statutes, it is synonymous to
aiding or abetting. See Abuelhawa, 556 U.S. at 819; cf Hillie, 14 F.4th at 686 (“A judicial
construction of a statute is an authoritative statement of what the statute meant before as
well as after the decision of the case giving rise to that construction.”). Indeed, and again
as Judge Katsas noted, Congress’s use in FOSTA of the specific phrasing of “promote or
facilitate” mirrors exactly “the canonical formulation for the offense of aiding and
abetting,” Woodhull IT, 948 F.3d at 375 (Katsas, J., concurring in part and concurring in
the judgment), that is, to provide assistance “with the intent thereby to promote or facilitate
commission of the crime,” see Rosemond vy. United States, 572 U.S. 65, 74 (2014)
(emphasis added).
Taking these two interpretive steps together, then, it becomes clear that the scope of
§ 2421A is in fact far more limited than under the reading propounded by the plaintiffs.
14
Rather than target advocacy, debate, or discussion, FOSTA is narrowly tailored to those
services that are owned, operated, or managed with the intent to aid, abet, or assist specific
instances of prostitution. Far from encompassing a substantial portion of protected speech,
then, FOSTA’s scope is limited to legitimately criminal activity. In light of this conclusion,
plaintiffs’ claim of overbreadth must fail.
Plaintiffs contend that I should instead follow the path recently taken by the Fourth
Circuit in United States v. Miselis, 972 F.3d 518, 535-37 (4th Cir. 2020), in which that
Court agreed in part with an overbreadth challenge to the federal Anti-Riot Act, 18 U.S.C.
§ 2101. That statute criminalizes, among other things, travel or use of a facility in interstate
or foreign commerce with “intent . . . to organize, promote, [or] encourage... ariot.” 18
U.S.C. § 2101(a)(2). The Fourth Circuit concluded that in the context of the Anti-Riot Act,
criminalization of an intent to “promote” or to “encourage” a riot rendered the statute
overbroad. The Fourth Circuit commenced its analysis by noting that “mere
encouragement is quintessential protected advocacy,” citing the Supreme Court’s past
holding that “[t]he mere tendency of speech to encourage unlawful acts is not a sufficient
reason for banning it.” Miselis, 972 F.3d at 536 (citing Ashcroft v. Free Speech Coalition,
535 U.S. 234, 253 (2002)). Turning to “promote,” the Fourth Circuit noted its “wide range
of meanings depending on context” but concluded that “in the context of an enterprise of a
riot,” the verb was “best understood” to take on a definition akin to “encourage,” and thus
its use in the Anti-Riot Act rendered that statute overbroad. /d. In reaching that conclusion,
the court rejected the Government’s contention that “promote” was “readily susceptible”
15
to a limiting construction as the Supreme Court had found in United States v. Williams
(and, as discussed above, I do here). Jd. at 537 (citing Williams, 553 U.S. at 294-95). In
Williams, a challenge to a child pornography statute, 18 U.S.C. § 2252A, the Supreme
Court had rejected an overbreadth claim in part due to the “transactional connotation” in
which the verb “promote” appeared. Williams, 553 U.S. at 294-95. In contrast, in Miselis
the Fourth Circuit found that similar reasoning would be “inapposite in the context of the
Anti-Riot Act, where the object of the promotional speech—the ‘riot’ defined under [18
U.S.C.] § 2102(a)—is wholly non-transactional, and can’t materialize until a sufficient
number of people are persuaded to show up at a certain future time and place and engage
in lawless conduct.” Miselis, 972 F.3d at 537. Instead, the court concluded that “promote”
as used in the context of the Anti-Riot Act “refers to abstract advocacy.” Id.
Even to the extent it is correct, however, the Fourth Circuit’s reasoning does not
lend itself to wholesale application to the merits of plaintiffs’ present challenge to FOSTA.
Whereas the statutory context of the Anti-Riot Act—including the verbs included alongside
“promote” and the very nature of the criminal conduct being so promoted—was found by
the Fourth Circuit to lead to an interpretation of promote that impinged on protected speech,
the statutory context of FOSTA points in the opposite direction. As noted above, FOSTA’s
prohibition is tailored toward specific instances of illegal prostitution, and “promote or
facilitate,” the exact phrase used in FOSTA, is best read as synonymous with traditional
principles of aiding-and-abetting liability. A limiting construction of § 2421A, to
encompass only such aiding and abetting of specified instances of criminal conduct, is thus
16
readily available to the Court, and for that reason I find unpersuasive plaintiffs’ reliance on
Miselis as a basis to hold FOSTA unconstitutional.’
Plaintiffs also contend that the prior decision of our Court of Appeals as to plaintiffs’
standing in this case precludes my holding that § 2421A is susceptible to the narrowing
construction I endorse above. I disagree. Indeed, I find that plaintiffs’ argument not only
overreads the majority’s opinion, but also ignores Judge Katsas’s concurrence. More
specifically, while the majority did point out that FOSTA’s language, including the
“promote or facilitate” elements discussed above, could be read to sweep broadly “when
considered in isolation,” Woodhull I, 948 F.3d at 372, the panel did so in the context of its
analysis of plaintiffs’ standing to bring a pre-enforcement challenge. That standing
analysis merely requires considering whether plaintiffs have established that they engage
in activities “arguably” within the scope of the challenged statute, see SBA, 573 U.S. at
164, not that the statute does in fact prohibit the alleged activities. As such, the majority
was not determining the precise scope of what FOSTA proscribes, but rather whether
plaintiffs’ broad reading of FOSTA was “arguably” a valid one. In short, the majority did
not decide how FOSTA should be construed, only how it could be construed. To that end,
the narrowing construction of the law discussed above was neither endorsed, nor rejected,
7 For much the same reason, my conclusion on the constitutionality of FOSTA is not affected by the recent
decision of the Ninth Circuit in United States v. Hansen, 25 F.4th 1103, 1107-09 (9th Cir. 2022), in which that court
found unconstitutional a statute that imposes criminal liability on anyone who “encouraged or induced an alien to
‘enter[] or reside in the United States .. . in violation of law.’” Jd. at 1108 (quoting 8 U.S.C. § 1324 (a)(1)(A)(iv)).
Like the Fourth Circuit in Miselis, the Ninth Circuit concluded that the relevant immigration law lacked the types of
textual and contextual indicators that would provide for a narrowing construction sufficient to avoid constitutional
difficulty. Id. Indeed, the Ninth Circuit specifically noted that a separate provision in the statute provided for aiding
and abetting liability such that the challenged “encourage” and “induce” language should not be read to take on that
meaning. Jd. In contrast, here J find that “promote or facilitate” as used in § 2421A are best understood as equivalents
to aiding and abetting.
17
by the majority’s opinion. Indeed, in his concurrence, Judge Katsas expressly stated that
the majority did not purport to construe the statute for anything other than the standing
analysis, noting instead that the plaintiffs’ preferred reading was only “identiffied] .. . as
at least one possible reading of FOSTA.” Woodhull I, 948 F.3d at 375 (Katsas, J.,
concurring in part and concurring in the judgment). Judge Katsas wrote separately
specifically to indicate that he viewed the plaintiffs’ reading ultimately as untenable, even
if he did also agree that the plaintiffs had standing under his narrower reading. I therefore
find that plaintiffs are incorrect in arguing that I am precluded from reading FOSTA so
narrowly: our Court of Appeals did not take any position on that reading of FOSTA, and
indeed Judge Katsas expressly adopted it.
As such, a proper construal of FOSTA leads to the conclusion that it is narrowly
tailored toward prohibiting activity that effectively aids or abets specific instances of
prostitution. I therefore have no trouble finding that its legitimate sweep, encompassing
only conduct or unprotected speech integral to criminal activity, predominates any sweep
into protected speech—indeed, under the narrow construal above, I do not read FOSTA to
possibly prohibit any such protected speech, much less a sufficient amount so as to render
the Act overbroad.* Plaintiffs’ claim to the contrary must be dismissed.
8 And, of course, my decision as to the merits of a facial, pre-enforcement challenge to the entirety of FOSTA
does not preclude a future, as-applied challenge to the statute were the Government to reach beyond the statute’s focus
and attempt to prosecute protected speech. I hold today only that the statute’s language, properly construed, is not
facially overbroad as defined by the Supreme Court.
18
II. = Plaintiffs’ Additional] First and Fifth Amendment Challenges
Though overbreadth is the primary basis on which plaintiffs challenge FOSTA,
plaintiffs have also raised a number of other theories rooted in the First and Fifth
Amendments as to why the Act is facially unconstitutional. However, these theories too
are unpersuasive: they depend in large part on the same overbroad reading of FOSTA that
I rejected above. As such, I will briefly explain my specific disagreement with each of
these claims below.
a. Void for Vagueness
First, plaintiffs allege that FOSTA is void for vagueness under the Due Process
Clause of the Fifth Amendment. “[A]n enactment is void for vagueness if its prohibitions
are not clearly defined.” Grayned v. City of Rockford, 408 U.S. 104, 108 (1972). The
doctrine applies to any statute that “fails to provide a person of ordinary intelligence fair
notice of what is prohibited, or is so standardless that it authorizes or encourages seriously
discriminatory enforcement.” Williams, 553 U.S. at 304. Though these concerns are
heightened “where a vague statute abuts upon sensitive areas of basic First Amendment
freedoms,” Grayned, 408 U.S. at 109, the Supreme Court has at the same time cautioned
that “perfect clarity and precise guidance have never been required even of regulations that
restrict expressive activity,” Ward v. Rock Against Racism, 491 U.S. 781, 794 (1989).
Here, plaintiffs’ vagueness claims against FOSTA rest in substantial part on their
construction of § 2421A, which, I have already concluded, misreads the key terms of that
section. Properly construed, § 2421A provides clear notice of the type of conduct
9
prohibited—or, more precisely, clear notice of the intent necessary to come within the
statute’s scope. Read within the criminal law context in which they appear, the statute’s
requirement of intent to “promote or facilitate” a specific instance of prostitution is
sufficiently clear for purposes of providing notice to the ordinary person; moreover, there
is no additional and independent obligation on Congress to define these or other similarly
well-established statutory terms.’
Beyond the trouble with plaintiffs’ reading of the challenged provisions, plaintiffs’
vagueness concerns are further undercut by the presence of the heightened scienter
requirements throughout FOSTA’s prohibitions. See Nat’l Ass’n of Manufacturers v.
Taylor, 582 F.3d 1, 27 (D.C. Cir. 2009) (“A scienter requirement may mitigate a law's
vagueness, especially with respect to the adequacy of notice to the complainant that his
conduct is proscribed.”). Each restriction contained in FOSTA requires proof of at least
knowledge, see 18 U.S.C. § 1591(e)(4) (requiring proof of “knowingly” assisting,
supporting, or facilitating a violation of § 1591(a)), with the main substantive prohibitions
requiring the even higher bar of proof of specific intent, see id. § 2421A(a), (b) (requiring
proof of “intent to promote or facilitate the prostitution of another person”); id.
§ 2421 A(b)(2) (requiring proof of “intent to promote or facilitate the prostitution of another
° Plaintiffs also suggest that the phrases “contribute to sex trafficking,” as used in § 2421A(b)(2) (the
aggravated offense), and “participation in a venture,” as used in both § 1591(a)(2) and § 1595, are unlawfully vague.
Neither contention has merit. “Sex trafficking” for purposes of § 2421A(b)(2) is necessarily narrowed by that
provision’s cross-reference to § 1591(a)—where the violation in question is defined—while, again read properly in
context, “contribute” is not so vague as to fail to provide notice. And second, “participation in a venture” is a defined
term under § 1591(e)(4) and § 1591(a)(1) by cross-reference: the statutory definition requires knowing assistance or
support of the well-defined criminal activity prohibited by § 1591(a)(1). As explained infra, section II.d, the terms
contained in FOSTA’s definition of participation in a venture are established synonyms in the criminal law for material
support. In sum, neither of these additional identified phrases are vague so as to render FOSTA facially defective.
20
person” plus “reckless disregard of the fact that such conduct contributed to sex
trafficking’). Requiring such evidence of mens rea serves to further protect against
discriminatory enforcement while also vastly reducing the risk of prosecution for guiltless
behavior. In short, I find FOSTA’s language to easily clear the bar of providing
constitutionally required notice of its scope, while the heightened scienter requirements
contained in the statute further mitigate any risk of the harms protected against by the void-
for-vagueness doctrine.
b. Selective Removal of Immunity
Plaintiffs next claim that FOSTA is violative of the First Amendment due to its
“selective” removal of Section 230 immunity. More specifically, plaintiffs take issue with
the fact that FOSTA removes immunity for violations of FOSTA’s substantive provisions
under § 230(c)(1)—that is, for treatment of an interactive computer service as a publisher
or speaker of content—while leaving intact those services’ § 230(c)(2) immunity for “any
action voluntarily taken in good faith to restrict access to or availability of material that the
provider or user considers to be obscene, lewd, lascivious, filthy, excessively violent,
harassing, or otherwise objectionable, whether or not such material is constitutionally
protected.” See 47 U.S.C. § 230(c). To plaintiffs, this choice encourages an overbroad
censoring and removal of content, with platforms facing liability for allowing certain
content on their sites but no reciprocal liability for removing too much content. See, e.g.,
Compl. § 163. While this claim may plausibly encapsulate the incentives created by this
section of FOSTA, plaintiffs cannot point to a First Amendment principle that, generally
21
speaking, restricts Congress’s ability to re-balance the provision of an immunity it
conferred in the first place (excepting, of course, more specific First Amendment principles
such as content-based discrimination). Plaintiffs instead appear to raise this claim in
conjunction with their broader overbreadth and vagueness claims—i.e., because FOSTA’s
scope of liability is, dosonding to the plaintiffs, so poorly defined, platforms will be
especially incentivized to cull as much material in FOSTA’s domain from their pages as
possible to avoid any chance of liability. These broader claims fail for the reasons
discussed above, and plaintiffs’ additional arguments concerning the policy implications
of removing immunity for some actions but not others cannot on their own support a facial
First Amendment challenge.
c. Viewpoint Discrimination
Plaintiffs also challenge FOSTA on more distinct First Amendment grounds than
those above, arguing that the Act discriminates against certain types of speech and fails the
strict scrutiny to which such content-based regulation is subject. However, this argument
falls short because FOSTA, on its face, is not a direct regulation of speech. As discussed
above, FOSTA’s provisions at most implicate speech by imposing restrictions on internet
services, which, among other things, may provide a forum or serve as a medium for speech.
In other words, FOSTA “serves purposes unrelated to the content of expression”—that is
the regulation of activity related to, and specifically the use of the internet in service of,
instances of illegal prostitution and sex trafficking—and thus its effect on speech is
inherently content neutral. See Ward, 491 U.S. at 791 (“Government regulation of
22
expressive activity is content neutral so long as it is ‘justified without reference to the
content of the regulated speech.’” (quoting Clark v. Community for Creative Non—
Violence, 468 U.S. 288, 293 (1984)); see also A.N.S.W.E.R. Coal. v. Basham, 845 F.3d
1199, 1208-09 (D.C. Cir. 2017). Though FOSTA may well implicate speech in achieving
its separate purpose, such an indirect effect does not provide a basis for strict scrutiny:
“even if [a law] has an incidental effect on some speakers or messages but not others,” it
is to be treated as content neutral. A.N.S.W.E.R. Coal., 845 F.3d at 1210 (quoting Ward,
491 U.S. at 791). Plaintiffs’ contention to the contrary rests, again, primarily on their
overreading of § 2421A. As discussed above, FOSTA does not purport to regulate the use
of the internet to host the generalized “promotion” of prostitution as a concept, including
via advocacy, and thus FOSTA does not enact a government regulation of speech that takes
a “pro-prostitution” stance, as plaintiffs suggest. Instead, FOSTA prohibits only a much
narrower range of activity, that which can be shown to aid or abet specific instances of
prostitution: while the former type of legislation could be viewed as targeting the source of
certain messages but not others, the latter is quite clearly justified by purposes wholly
related to any expressive content. As a result, plaintiffs’ claim that FOSTA fails strict
scrutiny must also be rejected.'®
‘0 Plaintiffs raised the argument that FOSTA fails even lesser forms of scrutiny only in their response to the
Government’s cross-motion, see Pls. Opp. to Defs.’ Mot. for Summ. J. (“Pls.’ Opp.”) [Dkt. # 38] at 33-34, and their
Complaint contains no suggestion that plaintiffs challenge the Act on this basis, and thus the Court need not address
such a claim. See, e.g., District of Columbia v. Barrie, 741 F.Supp.2d 250, 263 (D.D.C. 2010) (“It is well established
that a party may not amend its complaint or broaden its claims through summary judgment briefing.”) In any event,
plaintiffs’ claim based on intermediate scrutiny is unpersuasive, for it relies on the same misreading of § 2421A to
argue that FOSTA does not have the requisite “close fit between means and ends.” Pls. Opp. at 34. Construed
correctly, FOSTA is narrowly targeted at a band of conduct that materially advances specific instances of illegal
prostitution; it does not sweep beyond that purpose in any manner sufficient to trigger First Amendment concern.
23
d. “Deficient Scienter”
Plaintiffs’ final claim under the First or Fifth Amendments is a slightly narrower
version of their overbreadth and vagueness arguments: they argue that FOSTA does not
include a sufficient scienter requirement in § 2421A or its amendments to § 1591, leading
to an impermissibly broad sweep of liability in a First Amendment-—protected area.!! Like
plaintiffs’ other claims, however, this argument rests on a misreading of the statute.
Indeed, contrary to plaintiffs’ contention, both of the challenged sections include at least a
requirement of proof of knowledge for conviction. Specifically, in any prosecution under
§ 2421A, the Government is required to show that the defendant acted with specific intent,
that is “with the intent to promote or facilitate the prostitution of another person.”!*
Plaintiffs argue that this intent standard is undefined due to the breadth and vagueness of
“promote” or “facilitate,” but, of course, this claim amounts to little more than a reframing
of the vagueness and overbreadth arguments disposed of above. Those terms have clear
meanings that provide for a sufficiently specific intent element for FOSTA to withstand
this form of First Amendment scrutiny.
Likewise, plaintiffs’ argument about the scienter contained in § 1591 also rests on
their concern about vague verbs. Section 5 of FOSTA adds a definition of “participation
'! This claim is based in part on obscenity cases where the Supreme Court struck down statutes imposing, for
example, strict liability for the sale of obscene books. See, e.g., Smith v. California, 361 U.S. 147, 153-54 (1959).
!2 The plaintiffs also suggest that the aggravated version of § 2421A—which imposes a heightened sentence
in certain circumstances—actually has a lower mens rea than the main offense because it can be imposed where a
defendant “acts with reckless disregard of the fact” that his conduct contributed to sex trafficking. But this argument
is deeply flawed: the cited reckless disregard element is imposed on top of the required proof of intent to promote or
facilitate prostitution that is contained in the base offense.
24
in a venture” to § 1591, defining the term as “knowingly assisting, supporting, or
facilitating” a criminal sex-trafficking violation as defined in subsection (a)(1) of § 1591.
See 18 U.S.C. § 1591(a)(1), (e)(4). Before this addition, § 1591 criminalized knowingly
benefitting from participation in a venture, and plaintiffs suggest that the additional verbs
in the new definition serve to in fact expand the scope of criminalized participation, even
as they provide greater definition. But the new verbs—including, once more, facilitating—
are analogues to aiding and abetting in this criminal law context; they do not, as plaintiffs
suggest, capture such a broad and undefined scope of conduct that a defendant could
unwittingly be violating this provision. Section 1591 thus continues to include a
sufficiently defined scienter requirement of knowledge of the illegal conduct in question;
as with § 2421A, this element is more than enough for the amended § 1591 to pass
constitutional muster.
III. Ex Post Facto Liability
Lastly, moving outside of the context of the First and Fifth Amendments, plaintiffs
ask this Court to declare the Act unconstitutional on its face as violating the Constitution’s
Ex Post Facto Clause, see U.S. Const. art. I, § 9, cl. 3 (“No Bill of Attainder or ex post
facto Law shall be passed.”’). Plaintiffs make this claim on the basis that § 4(b) notes that
FOSTA’s amendments to the grants of immunity contained in Section 230 “shall apply
regardless of whether the conduct alleged occurred, or is alleged to have occurred, before,
on, or after such date of enactment.” FOSTA § 4(b), 132 Stat. at 1254-55. As discussed
above, the Act’s amendments to Section 230 are provided for in FOSTA § 4(a), which
25
clarify that certain grants of immunity to interactive computer services in Section 230 do
not extend to civil suits brought pursuant to 18 U.S.C. § 1595 or to state criminal
prosecutions where the underlying conduct would constitute a violation of either § 2421A
or § 1591. See id. § 4(a), 132 Stat. at 1254. Plaintiffs argue, therefore, that the challenged
FOSTA provision effectively exposes interactive computer services to retroactive liability
for conduct previously immunized by Section 230, violating the Ex Post Facto Clause. See
U.S. Const. art. I, § 9, cl. 3; see also United States v. Alvaran-Velez, 914 F.3d 665, 668
(D.C. Cir. 2019) (“Any ‘law that changes the punishment, and inflicts a greater
punishment[] than the law annexed to the crime[] when committed,’ is an ex post facto
law.” (quoting Calder v. Bull, 3 U.S. (3 Dall.) 386, 390 (1798))).
To its credit, the Government does not squarely dispute the merits of this argument.
See generally Gov’t Cross-Mot. for Summ. J. at 36-37. Indeed, in a letter to Congress
during the drafting of FOSTA, the Department of Justice expressed concern that any
retroactive application of FOSTA’s amendment of Section 230 would run afoul of the Ex
Post Facto Clause: the Government noted that “insofar as [FOSTA] would ‘impose[] a
punishment for an act which was not punishable at the time it was committed’ or ‘impose[]
additional punishment to that then prescribed,’” it would offend the Ex Post Facto clause.
See 164 Cong. Rec. H1297 (daily ed. Feb. 27, 2018) (quoting Cummings v. Missouri, 71
US. (4 Wall.) 277, 325-26 (1867)). Rather than dispute the substance of plaintiffs’
constitutional concerns in this respect, the Government instead argues that plaintiffs cannot
rely on a hypothetical violation of the Ex Post Facto Clause to bring a facial pre-
26
enforcement challenge to FOSTA. The Government contends that such claims may be
brought only on an as-applied basis, once a defendant faces a specific enforcement action
that it claims is unconstitutional.
Here, however, I need not decide whether the Government is correct that facial, pre-
enforcement challenges based on the Ex Post Facto Clause are per se barred. This is so
because in this case, plaintiffs seek injunctive relief only against defendants comprising the
federal law enforcement community, asking the Court to enjoin them from enforcing
FOSTA. By the Act’s plain terms, however, any enforcement of § 4(b)—the provision
claimed to be a violation of the Ex Post Facto Clause—would be undertaken by parties not
before the Court. Indeed, any retroactive application of FOSTA would be initiated instead
either by private plaintiffs bringing civil suits under 18 U.S.C. § 1595 or by state law
enforcement bringing prosecutions pursuant to state criminal laws. See FOSTA § 4(a), 132
Stat. at 1254. “[A] federal court exercising its equitable authority may enjoin named
defendants from taking specified unlawful actions,” but “no court may ‘lawfully enjoin the
2
world at large,’” or even “challenged laws themselves.” Whole Women’s Health v.
Jackson, 145 8S. Ct. 522, 535 (2021) (quoting Alemite Mfg. Corp. v. Staff, 42 F.2d 832, 832
(2d Cir. 1930) (Learned Hand, J.)). As such, without a named defendant before me who
could even hypothetically undertake enforcement of FOSTA in a manner that would violate
the Ex Post Facto Clause, I cannot provide the plaintiffs with any relief, and this claim
must be dismissed as well.
27
CONCLUSION
For the foregoing reasons, defendants’ cross-motion for summary judgment is
GRANTED, plaintiffs’ motion for summary judgment is DENIED, and this case is
DISMISSED. A separate order consistent with this decision accompanies this
Memorandum Opinion.
A
Cebus Ayo
RICHARD Ji LBON
United States District Judge
28