Amicus Curiae Brief — John Doe, et al., Petitioners v. X Corp., fka Twitter, Inc.
Supreme Court briefMar 12, 2026
Ask Donna
What actually matters in this document.
Text
No. 25-949
In the
Supreme Court of the United States
JOHN DOE 1 AND JOHN DOE 2,
v.
X CORP., F/K/A TWITTER, INC.,
Petitioners,
Respondent.
On Petition for a Writ of Certiorari to the
U.S. Court of Appeals for the Ninth Circuit
Amicus Brief of Senator Josh Hawley
in Support of Petitioners
JOSH HAWLEY
U.S. SENATOR
Counsel of Record
WILLIAM SIMPSON
DANIEL PARK
AARON YURATOVICH
115 Russell Senate Office Bldg.
Washington, D.C. 20510
(202) 224-6154
William_Simpson@hawley.senate.gov
i
TABLE OF CONTENTS
Interest of Amicus Curiae ................................................... 1
Summary of Argument........................................................ 2
Argument ............................................................................. 3
I.
In the absence of clarity from this Court, lower
courts are scattered in attempts to define the
scope of immunity. ........................................................ 5
II. Congress has never protected platforms that
knowingly publish or distribute child
pornography. ................................................................. 7
A. Section 230 immunity for knowing criminal
conduct would effectively nullify federal
private rights of action for crimes enabled by
interactive computer services. ............................... 8
B. Platforms cannot simultaneously claim
editorial discretion under the First
Amendment and passive-conduit status under
Section 230. ........................................................... 10
C. In the vacuum of this Court’s refusal to weigh
in, lower courts have distorted Section 230’s
original meaning over time, rendering
subsection (c)(2) surplusage. ................................ 13
D. No court has used Section 230 to immunize
the type of actual knowledge present here. ........ 14
III. Congress adopted Section 230 to codify the
distinction between distributors and publishers,
not to create a blanket shield for corporations that
knowingly distribute child pornography. .................. 17
ii
A. At common law, publisher liability and
distributor liability were distinct doctrines
with different knowledge requirements. ............. 17
B. Section 230(c)(1) removed publisher liability;
it did not abolish distributor liability. ................. 18
C. Congress underscored the limited scope of
Section 230 by labeling it a “Good Samaritan”
provision. ............................................................... 20
IV. Millions of images of child sexual abuse material
will continue proliferating unchecked online if the
Court does not clarify Section 230’s scope. ................ 21
Conclusion.......................................................................... 23
iii
TABLE OF AUTHORITIES
Cases
Ashcroft v. Iqbal, 556 U.S. 662 (2009) .............................. 10
Corley v. United States, 556 U.S. 303 (2009) ................... 14
Doe 1 v. Twitter, Inc., 148 F.4th 635 (9th Cir. 2025) ... 8, 9,
14
Doe ex rel. Roe v. Snap, Inc., 144 S. Ct. 2493 (2024) 11, 12,
16, 22
Doe v. Facebook, Inc., 142 S. Ct. 1087 (2022) ......... 7, 15, 22
Doe v. Twitter, Inc., No. 21-CV-00485-JCS, 2023 WL
8568911 (N.D. Cal. Dec. 11, 2023) ................................ 14
Epic Sys. Corp. v. Lewis, 584 U.S. 497 (2018).............. 9, 10
Fair Hous. Council of San Fernando Valley v.
Roommates.Com, LLC, 521 F.3d 1157 (9th Cir. 2008) 14
Force v. Facebook, Inc., 934 F.3d 53 (2d Cir. 2019) ..... 4, 15
Gonzalez v. Google LLC, 598 U.S. 617 (2023) .............. 5, 15
Hibbs v. Winn, 542 U.S. 88 (2004) .................................... 19
Isbrandtsen Co. v. Johnson, 343 U.S. 779 (1952) ............ 19
M.H. On behalf of C.H. v. Omegle.com LLC, 122 F.4th
1266, 1270 (11th Cir. 2024) ............................................. 7
Malwarebytes, Inc. v. Enigma Software Grp. USA, LLC,
141 S. Ct. 13 (2020) ............................................ 13, 20, 23
Moody v. NetChoice, LLC, 603 U.S. 707 (2024) ....... 3, 5, 11
NetChoice v. Griffin, No. 5:25-CV-5140, 2025 WL
3634088, at *14 (W.D. Ark. Dec. 15, 2025) ..................... 6
Russello v. United States, 464 U.S. 16 (1983) .................. 20
Smith v. California, 361 U.S. 147 (1959) ......................... 18
Stratton Oakmont, Inc. v. Prodigy Servs. Co., No.
31063/94, 1995 WL 323710 (N.Y. Sup. Ct. May 24,
1995) ..................................................................... 4, 21, 22
Twitter, Inc. v. Taamneh, 598 U.S. 471 (2023) ............ 2, 15
Yates v. United States, 574 U.S. 528 (2015). .................... 20
iv
Zeran v. America Online, Inc., 129 F.3d 327 (4th Cir.
1997) ................................................................... 16, 17, 20
Statutes
18 U.S.C. § 1591 .................................................................. 7
18 U.S.C. § 1595 .................................................................. 7
18 U.S.C. § 2252A ................................................................ 6
18 U.S.C. § 2252A(a) ......................................................... 17
18 U.S.C. § 2255 .................................................................. 7
47 U.S.C. § 230 .................................................................... 3
47 U.S.C. § 230(c)(1) ...................................................... 4, 19
47 U.S.C. § 230(c)(2) .............................................. 12, 19, 22
47 U.S.C. § 230(e)(1) ............................................................ 7
47 U.S.C. § 230(e)(5) ............................................................ 7
Other Authorities
141 CONG. REC. S8088 (daily ed. June 9, 1995) (statement
of Senator Exon)............................................................. 13
141 CONG. REC. S8089 (daily ed. June 9, 1995) (statement
of Senator Exon)............................................................. 14
FOSTA, Pub. L. No. 115-164, 132 Stat. 1253 .................... 9
Gregory M. Dickinson, Section 230: A Juridical History,
28 STAN. TECH. L. REV. 1, 13 (2024) .............................. 13
National Center for Missing & Exploited Children, 2019
CyberTipline Reports by Electronic Service Providers
(ESPs) (2020),
https://www.missingkids.org/content/dam/missingkids/p
dfs/2019-reports-by-esp.pdf ........................................... 23
National Center for Missing & Exploited Children, 2024
CyberTipline Report (2025),
https://www.missingkids.org/content/dam/missingkids/p
v
dfs/cybertiplinedata2024/2024-CyberTipline-Report.pdf.
........................................................................................ 22
National Center for Missing & Exploited Children, 2024
CyberTipline Reports by Electronic Service Providers
(ESPs) (2025),
https://www.missingkids.org/content/dam/missingkids/p
dfs/cybertiplinedata2024/2024-reports-by-esp.pdf . 22, 23
Petition for Writ of Certiorari, NetChoice, LLC v. Paxton
(No. 22-555) .................................................................... 11
Restatement (Second) of Torts (Am. L. Inst. 1977) ......... 18
The Missing and Exploited Children’s (MEC) Program:
Background and Policies, CONG. RSCH. SERV. (Jul. 22,
2019),
https://www.congress.gov/crs_external_products/RL/PD
F/RL34050/RL34050.45.pdf........................................... 22
William E. Buehlow III, Re-Establishing Distributor
Liability on the Internet: Recognizing the Applicability
of Traditional Defamation Law to Section 230 of the
Communications Decency Act of 1996, 116 W. Va. L.
Rev. 313 (2013)................................................................. 4
Treatises
W. Page Keeton et al., Prosser and Keeton on the Law of
Torts § 113 (5th ed. 1984) .............................................. 18
1
No. 25-949
In the
Supreme Court of the United States
JOHN DOE 1 AND JOHN DOE 2,
v.
X CORP., F/K/A TWITTER, INC.,
Petitioners,
Respondent.
On Petition for a Writ of Certiorari to the
U.S. Court of Appeals for the Ninth Circuit
Amicus Brief of Senator Josh Hawley
in Support of Petitioners
INTEREST OF AMICUS CURIAE1
As a Member of the U.S. Senate, Chairman of the Senate Judiciary Committee’s Subcommittee on Crime and
Counterterrorism, and a former state attorney general
with a long record of engagement on technology policy, Senator Josh Hawley has a strong interest in upholding Congress’s proper purposes regarding Section 230’s liability
shield. If this Court does not correct errant interpretations
1 Amicus provided timely notice of this brief. No counsel for a party
authored any portion of this brief. No person or entity other than the
amicus curiae signing this brief contributed money to fund the preparation or submission of this brief.
2
of this law, policymakers will continue to face challenges to
exercising their Article I authority to effectively regulate
some of the nation’s most lucrative interstate commerce.
SUMMARY OF ARGUMENT
Nothing in the text, structure, or history of the Communications Decency Act’s Section 230 shields internet
platforms that knowingly possess and distribute child pornography. Congress drafted Section 230 to avoid publisher
liability without upending distributor liability, which had
its own rich common law backdrop. The Good Samaritan
carveout in Section 230(c) is narrow and unrelated to any
facts in this case. The Ninth Circuit’s holding to the contrary represents the most extreme extension yet of Section
230 immunity in the statute’s history, and is a perfect vehicle for this Court to rein in the statute’s purview to the
four corners of its text.
First, after declining to reach the Section 230 issue in
previous terms, this Court has created a vacuum on the
scope of immunity under Section 230 that has produced
disarray in lower courts.
Second, nothing in Section 230 immunizes platforms
that have actual, verified knowledge that they are hosting
and distributing criminal content—least of all child sexual
abuse material. The Ninth Circuit’s contrary holding cannot be reconciled with the statute’s text, which addresses
only “publisher or speaker” liability; with this Court’s characterization of platforms in Twitter, Inc. v. Taamneh, 598
U.S. 471 (2023); or with Congress’s repeated legislative
judgments. Congress has specifically imposed liability
against digital platforms, including in the Allow States and
Victims to Fight Online Sex Trafficking Act (FOSTA)
amendments to Section 230 itself. This case also exposes a
conceptual contradiction at the heart of platforms’ legal
3
strategy: they cannot simultaneously claim that their content-moderation decisions are protected editorial judgments, see Moody v. NetChoice, LLC, 603 U.S. 707 (2024),
while also claiming that those same decisions are passive
publishing acts immunized by Section 230.
Third, Congress adopted Section 230 to codify the common-law distinction between publishers and distributors,
not to create blanket immunity for corporations that knowingly distribute criminal material. Section 230(c)(1) removes strict publisher liability for unknowing hosts; it does
not eliminate distributor liability for platforms with actual
knowledge. In fact, by interpreting the “publisher or
speaker” provision of subsection (c)(1) as a blanket immunity clause, courts have rendered Congress’s language in
subsection (c)(2) superfluous. In reality, Section 230(c)’s
heading—“Protection for ‘Good Samaritan’ Blocking and
Screening of Offensive Material”—confirms that Congress
designed the statute to encourage platforms to remove
harmful content, not to immunize platforms that deliberately refuse to do so.
Fourth, child sexual abuse material continues to proliferate online, to the tune of tens of millions of illegal images.
Without curtailing the excesses of lower courts, victims will
be effectively left without recourse that Congress has provided under FOSTA.
ARGUMENT
This case is an ideal vehicle for this Court to clarify the
scope of a statute that has grown larger than its own terms.
In 1996, Congress passed 47 U.S.C. § 230 (“Section 230”)
for a specific purpose: to help the nascent internet grow and
thrive without being demolished by litigation. As websites
like online message boards started hosting user-generated
speech, courts began treating those sites as the publishers
4
of their users’ speech if they engaged in any content moderation at all. See Stratton Oakmont, Inc. v. Prodigy Servs.
Co., No. 31063/94, 1995 WL 323710 at *4 (N.Y. Sup. Ct.
May 24, 1995). That early doctrine effectively punished
website owners for moderating offensive content in any
way. See William E. Buehlow III, Re-Establishing Distributor Liability on the Internet: Recognizing the Applicability
of Traditional Defamation Law to Section 230 of the Communications Decency Act of 1996, 116 W. Va. L. Rev. 313,
332 (2013) (tracing the implications of the Stratton Oakmont ruling). If the internet was going to continue to develop, Congress needed to clarify that websites passively
hosting user-generated content wouldn’t be buried in personal-injury lawsuits.
Section 230 was the congressional solution. In relevant
part, that statute states 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.” 47 U.S.C. § 230(c)(1). So, for
instance, when a user posts comments to social media, the
platform will not be treated as the publisher of the
speech—meaning that even if the speech is defamatory or
otherwise unlawful, the platform is not automatically liable.
In the years since, interactive computer service providers have argued for an expansive construction of Section
230. In their view, Section 230 provides a blank check for
any content-management decisions platforms make, extending even to content that their platforms affirmatively
recommend. See, e.g., Force v. Facebook, Inc., 934 F.3d 53
(2d Cir. 2019) (extending Section 230 immunity to protect
algorithmic recommendations). The Ninth Circuit now
goes one step further, applying this sweeping immunity to
5
platforms’ decisions to decline moderation, even in the most
extreme cases of confirmed child pornography.
But the plain text of Section 230, along with its governing logic, says nothing about a catch-all liability shield of
this magnitude. Rather, Section 230 states only that interactive computer service providers will not be treated as
speakers or publishers of other users’ speech. Treating Section 230 as a totalizing shield collapses an important distinction: cases where internet platforms distribute, but do
not publish, content that they know or should have known
is illegal. The Ninth Circuit’s decision wrongly engulfed
these distinctions.
I.
In the absence of clarity from this Court, lower
courts are scattered in attempts to define the
scope of immunity.
Disarray in lower courts has only increased since this
Court’s decision in Moody v. NetChoice, LLC, 144 S. Ct.
2383 (2024). There, this Court examined the First Amendment dimensions of state laws regulating content moderation but did not address the scope of Section 230 immunity.
The decision’s characterization of content moderation as an
exercise of editorial discretion has complicated, rather than
clarified, the immunity analysis. If a platform’s decisions
about what content to retain or remove constitute protected editorial judgment, lower courts struggle to determine whether Section 230 immunizes those same decisions
even when the platform has actual knowledge that the content depicts the sexual abuse of a child. Moody supplied a
First Amendment vocabulary for describing platform conduct but offered no limiting principle for statutory immunity.
This is not the first time the Court has left that question unanswered. In Gonzalez v. Google, 598 U.S. 617
6
(2023), the Court granted certiorari to determine whether
Section 230 immunized algorithmic recommendations of
terrorist content, then vacated the decision below without
reaching the merits. Together, Moody and Gonzalez represent two consecutive terms in which this Court had the opportunity to define the outer boundaries of Section 230 immunity and declined to do so. The consequence has been
predictable: lower courts have been left to construct those
boundaries on their own, applying a statute enacted in
1996 to the modern internet with varied results, lacking
guidance on the most basic question of whether “publisher”
immunity has any limiting principle at all.
In this Court’s silence on the scope of publisher or
speaker liability in subsection (c)(1), lower courts find it
“hard to say” when speech is “properly attributable to the
interactive computer service on which it is posted.”
NetChoice v. Griffin, No. 5:25-CV-5140, 2025 WL 3634088,
at *14 (W.D. Ark. Dec. 15, 2025).
Most consequentially, the Ninth Circuit’s decision in
this case stands in tension with the Eleventh Circuit on
whether actual knowledge of sex trafficking on a platform's
servers is sufficient to overcome Section 230 immunity under the FOSTA amendments, or whether a plaintiff must
additionally show that the platform engaged in some affirmative act beyond knowing retention. According to the
Ninth Circuit, the FOSTA exception requires the platform’s “own conduct” to constitute an independent violation
of § 1591, meaning “affirmative conduct” beyond merely
“turning a blind eye to illegal revenue-generating content.”
Doe 1 v. Twitter, Inc., 148 F.4th 635, 644 (9th Cir. 2025).
But the Eleventh Circuit has alluded to a different standard, requiring only that defendant platforms “had actual
knowledge of . . . the underlying incident,” without imposing an additional affirmative conduct requirement. M.H.
7
On behalf of C.H. v. Omegle.com LLC, 122 F.4th 1266, 1270
(11th Cir. 2024).
The Ninth Circuit itself acknowledged this tension, describing the Eleventh Circuit’s analysis as addressing “a
slightly different question.” Doe 1, 148 F.4th at 643 n.3. But
that undefined distinction offers no clarity to other courts.
Congress did not strip Section 230 immunity for sex-trafficking claims only to have it reimposed through a judicially created affirmative-conduct requirement that appears nowhere in the statutory text. To resolve the disarray, this is an “appropriate case” to “address the proper
scope of immunity under § 230.” Doe v. Facebook, Inc., 142
S. Ct. 1087, 1089 (2022) (mem.) (Thomas, J., respecting the
denial of certiorari).
II. Congress has never protected platforms that
knowingly publish or distribute child pornography.
This case presents an unprecedented factual record, far
from anything contemplated by Congress during passage
of the Communications Decency Act. Granting certiorari
would provide this Court with an ideal opportunity to clarify the limits of Section 230 immunity. No recorded Section
230 case has involved a platform that (1) received a direct
report identifying minor victims by name, (2) required the
victim himself to submit government-issued identification,
(3) had human agents for the platform review the content
at issue, (4) confirmed that the subjects were minors, and
(5) affirmatively decided to leave the illegal content online.
The record, in other words, does not present a case of
passive or unwitting hosting. It presents a case of knowing,
verified, and deliberate retention of criminal contraband—
after human review, after confirmation that the subjects
were minors, and after an affirmative decision to leave the
8
material available for public consumption. The question is
whether Section 230(c)(1) immunizes that conduct.
The Ninth Circuit held that it does. In the court’s view,
Section 230 immunized all of Twitter’s conduct because the
decision to retain the material was a “quintessential publishing activity.” Doe 1 v. Twitter, Inc., 148 F.4th 635, 643
(9th Cir. 2025). That holding cannot be sustained. It is irreconcilable with the statute’s text, with Congress’s companion enactments addressing child sexual exploitation,
with the conceptual framework that platforms themselves
have advanced in other litigation, and with how this Court
has characterized platform conduct.
A. Section 230 immunity for knowing criminal
conduct would effectively nullify federal private rights of action for crimes enabled by interactive computer services.
The Ninth Circuit’s blanket rule immunizing any refusal to take down illegal content would effectively nullify
other statutes undergirding federal child sexual abuse protections. Congress has enacted a comprehensive framework to combat the sexual exploitation of children online.
18 U.S.C. § 2252A criminalizes the knowing possession and
distribution of child pornography, including by means of a
computer. Section 2255 authorizes civil remedies for victims of such offenses, permitting recovery of actual damages, attorneys’ fees, and other relief. 18 U.S.C. § 2255. Sections 1591 and 1595 criminalize sex trafficking and authorize victims to pursue civil remedies against those who
knowingly benefit from participation in a venture that they
know involves trafficking. 18 U.S.C. §§ 1591, 1595.
Congress has also acted specifically to ensure that Section 230 does not override these protections. Subsection
(e)(1) provides that “[n]othing in this section shall be
9
construed to impair the enforcement of” federal criminal
law, including laws “relating to sexual exploitation of children.” 47 U.S.C. § 230(e)(1). And in 2018, Congress enacted
FOSTA, Pub. L. No. 115-164, 132 Stat. 1253, adding Section 230(e)(5), which confirms that Section 230 does not
“impair or limit” civil actions under 18 U.S.C. § 1595 where
the underlying conduct violates 18 U.S.C. § 1591.
The Ninth Circuit’s holding effectively nullifies all
these provisions as applied to platforms. If a platform with
actual, verified knowledge of child pornography on its servers is nevertheless “perforce immune” under Section 230,
then the civil remedies Congress codified in § 2255 and
§ 1595 are unavailable in the very circumstances Congress
designed them to address. Doe 1, 148 F.4th at 642. A victim
of child sexual exploitation who identifies the platform
hosting the material, reports it, submits his own identification to prove his claims, and still cannot obtain relief has
been left without the remedy Congress provided him. That
puts Section 230 “at war” with itself and with the broader
federal code, and flies in the face of courts’ “duty to interpret Congress’s statutes as a harmonious whole.” Epic Sys.
Corp. v. Lewis, 584 U.S. 497, 502 (2018).
Sidestepping the fact that Section 230(c)(2) only grants
immunity for actions that “restrict access” to—i.e., take
down—inappropriate content, the Ninth Circuit effectively
immunized interactive computer services from even
broader claims of civil liability for knowing criminal conduct that Congress legislated to prevent. In fact, the Ninth
Circuit’s decision as to Count 2 (knowingly benefiting from
a sex-trafficking venture) subverted an explicit congressional exception to immunity for the very statute that enabled the private right of action. See Doe 1, 148 F.4th at 64344.
10
Plaintiffs plausibly alleged, and the Ninth Circuit did
not question, that Twitter reviewed the videos at issue,
knew that the videos contained child pornography, and
nonetheless kept the videos on its platform. As such, there
is at least a question of fact as to whether Twitter knew
that keeping the content online was facilitating a sex-trafficking venture. See generally Ashcroft v. Iqbal, 556 U.S.
662, 678 (2009) (“To survive a motion to dismiss, a complaint must contain sufficient factual matter, accepted as
true, to state a claim to relief that is plausible on its face.”
(cleaned up and citation omitted)).
The canon of harmonization requires a reading of Section 230 that preserves the operation of these companion
statutes. Where two federal provisions appear to conflict, it
is this Court’s “duty to interpret Congress’s statutes as a
harmonious whole rather than at war with one another.”
Epic Sys. Corp., 584 U.S. at 502. Section 230(c)(1) can be
given full effect without nullifying §§ 2252A, 2255, 1591,
and 1595. Section 230 provides that platforms will not be
treated as publishers or speakers of user content, thus
eliminating strict liability for unknowing hosts, while leaving intact the knowledge-based liability that Congress enacted in its child-exploitation and trafficking statutes. The
Ninth Circuit’s reading, by contrast, treats Section 230 as
an absolute bar on all civil claims against platforms, regardless of the platform’s knowledge—a reading that can
be maintained only by judicially vetoing § 230(e)(1) and the
FOSTA amendments altogether.
B. Platforms cannot simultaneously claim editorial discretion under the First Amendment
and passive-conduit status under Section 230.
The Ninth Circuit’s holding also exposes a conceptual
contradiction in the legal positions that major internet and
11
social media platforms have adopted before this Court.
Such platforms have a well-trodden history of demanding
First Amendment protection for so-called expressive, editorial judgments. E.g., Petition for Writ of Certiorari at 13,
NetChoice, LLC v. Paxton, 2023 WL 8437869 (No. 22-555)
(“[T]he First Amendment protects private parties’ editorial
choices to choose whether and how to disseminate speech,”
which “applies to social media websites, which both publish
and disseminate speech.”). But to find haven in Section 230’s “Good Samaritan” immunity, they have also argued that their platforms act as mere passive hosts of user
expression. E.g., Doe ex rel. Roe v. Snap, Inc., 144 S. Ct.
2493, 2494 (2024) (Thomas, J., dissenting from the denial
of certiorari) (“Platforms claim that since they are not
speakers under §230, they cannot be subject to any suit implicating users’ content, even if the suit revolves around
the platform’s alleged misconduct.” (emphasis in original)).
These platforms cannot have their cake and eat it too.
They are either publishers of speech or passive conduits for
users’ speech—but they’ve conceded time and again that
they cannot be neither; and they surely aren’t both.
Indeed, the platforms successfully argued before this
Court that their content-moderation decisions were “expressive” choices constituting the exercise of protected “editorial control.” Moody, 603 U.S. at 738. Justice Kagan’s
majority opinion accepted this characterization: “The individual messages may originate with third parties, but the
larger offering is the platform’s. . . . And in the aggregate
they give the feed a particular expressive quality.” Id. (emphases added). Yet in this case, Twitter seeks blanket immunity under Section 230 on the theory that its decision to
keep child pornography on its platform was a “quintessential publishing activity” entitled to automatic statutory
protection. 148 F.4th at 643. The tides have apparently
12
turned, and Twitter’s editorial decision making is no longer
expressive. Here and now, the platform was merely monitoring and disseminating third-party content, Twitter
claims.
It is, of course, in the platforms’ interest to maintain
this contradiction. Equivocating on their role in “publication” allows the platforms to invoke Section 230 to shield
their behavior from private suits while simultaneously invoking the First Amendment to shield their behavior from
nettlesome state laws. See Snap, Inc., 144 S. Ct. at 2494
(Thomas, J., dissenting from the denial of certiorari) (“In
the platforms’ world, they are fully responsible for their
websites when it results in constitutional protections, but
the moment that responsibility could lead to liability, they
can disclaim any obligations and enjoy greater protections
from suit than nearly any other industry.”).
These positions are logically incompatible. If a platform’s decision to retain content after human review is an
exercise of editorial control sufficient to invoke the First
Amendment, it cannot simultaneously be a passive publishing act that automatically triggers blanket Section 230
immunity. The platform must choose: either its contentmoderation decisions are active, editorial, and expressive—
in which case it is exercising judgment, not passively distributing—or they are passive, automatic, and non-expressive—in which case it cannot invoke the First Amendment
to resist regulation. The Court should not permit platforms
to toggle between these irreconcilable positions depending
on which liability they wish to avoid. And until the Court
reconciles this incompatibility, platforms will continue to
use Section 230 “as a get-out-of-jail free card.” Snap, Inc.,
144 S. Ct. at 2494 (Thomas, J., dissenting from the denial
of certiorari).
13
C. In the vacuum of this Court’s refusal to weigh
in, lower courts have distorted Section 230’s
original meaning over time, rendering subsection (c)(2) surplusage.
Section 230(c)—titled the “Protection for ‘Good Samaritan’ blocking and screening of offensive material”—has
two parts, but courts have read subsection (c)(1)’s definitional statement so broadly as to read the operative immunity provision in Section 230(c)(2) out of existence. See
Malwarebytes, Inc. v. Enigma Software Grp. USA, LLC,
141 S. Ct. 13, 16-17 (2020) (Thomas, J., respecting the denial of certiorari). Subsection (c)(2) defines the outer
bounds of the immunity conferred by Section 230.
Put plainly, subsection (c)(2) protects interactive computer services against claims that they took down too much
content they deemed inappropriate. See 47 U.S.C. §
230(c)(2) (“restrict access”). And critically, the statute confers immunity for civil liability to “Good Samaritan” providers only when those takedown decisions are made “in
good faith.” Id.; see generally infra Section III(c). Thus, any
immunity that subsection (c)(1) confers on interactive computer services must be read in harmony with the immunity
conferred by subsection (c)(2). Failing to do so renders subsection (c)(2) surplusage and risks vitiating a core purpose
of Section 230: the protection of children.
When Congress enacted the Communications Decency
Act, Congress, led by Senator Exon of Nebraska, drafted it
with a “fundamental purpose … to provide much-needed
protection for children.” 141 CONG. REC. S8088 (daily ed.
June 9, 1995) (statement of Senator Exon); see also Gregory
M. Dickinson, Section 230: A Juridical History, 28 STAN.
TECH. L. REV. 1, 13 (2024). And specifically, the prevention
of “erotic” pictures of “children” from circulating online was
14
an explicit priority for the legislation. 141 CONG. REC.
S8089 (daily ed. June 9, 1995) (statement of Senator Exon).
But like so many courts, the Ninth Circuit here has
read Section 230(c)(1) so broadly as to render “any activity
that can be boiled down to deciding whether to exclude material that third parties seek to post online . . . perforce immune under section 230.” Doe 1, 148 F.4th at 643 (quoting
Fair Hous. Council of San Fernando Valley v. Roommates.Com, LLC, 521 F.3d 1157 (9th Cir. 2008)). In doing
so, the Ninth Circuit fabricates a strict non-liability rule of
subsection (c)(1) for a failure to remove known criminal
content it voluntarily reviewed after repeated requests for
removal.
Tellingly, neither the Ninth Circuit nor the district
court cites subsection (c)(2) even once—apparently deeming the express congressional mandate to limit civil immunity to actions to “restrict access” irrelevant. See 148
F.4th 635; Doe v. Twitter, Inc., No. 21-CV-00485-JCS, 2023
WL 8568911, at *2 (N.D. Cal. Dec. 11, 2023). That’s not how
this Court reads statutes. Congress’s words here “should be
construed so that effect is given to all its provisions, so that
no part will be inoperative or superfluous, void or insignificant.” Corley v. United States, 556 U.S. 303, 314 (2009)
(cleaned up and citation omitted). Restoring full force to the
text of Section 230(c)(2) would require reining in the Ninth
Circuit’s decision to ignore it.
D. No court has used Section 230 to immunize
the type of actual knowledge present here.
The Ninth Circuit’s holding is not only textually unsupported, it is without precedent. No court—including this
one—has applied Section 230 to immunize a platform with
actual knowledge of child sexual abuse material.
15
When this Court declined to reach the Section 230 issue in Gonzalez, 598 U.S. 617, the platform’s knowledge
was constructive at best. It derived from the algorithmic
processing of ISIS propaganda—a context in which the
platform’s awareness of the specific content could only be
imputed from the operation of its recommendation systems. Similarly, in Force v. Facebook, Inc., 934 F.3d 53 (2d
Cir. 2019), and Doe v. Facebook, Inc., 142 S. Ct. 1087 (2022)
(mem.), the platforms’ knowledge had not been established
through the platforms’ own internal verification processes.
Here, by contrast, Twitter’s knowledge was not constructive or inferred. It was actual, verified, and the product of a deliberate internal review process. Twitter required the victim to submit identification. Twitter’s own
agents reviewed the explicit videos. They confirmed that
the subjects depicted were minors. And Twitter decided, in
a written communication, that it would take “no action.”
This is not a case about a platform that failed to detect unlawful content amid billions of posts. It is a case about a
platform that detected, verified, and then chose to retain
criminal content. No past Section 230 case presented such
stark facts.
This Court’s analysis in Twitter, Inc. v. Taamneh reinforces the point. 598 U.S. 471 (2023). In Taamneh, the
Court rejected aiding-and-abetting liability against socialmedia platforms for hosting ISIS content, in part because
the plaintiffs had not alleged that the platforms “consciously and culpably” chose to keep the specific terrorist
content at issue online. Id. at 497. The Court emphasized
that the platforms’ algorithms “appear agnostic as to the
nature of the content,” and that the relationship between
the platforms and ISIS was “highly attenuated.” Id. at 499500.
16
Taamneh’s logic implies that where a platform does
have specific knowledge and does make a conscious decision to retain criminal content, the analysis changes. This
case presents exactly those facts. Twitter did not unknowingly host child pornography through an agnostic algorithm. Its human agents reviewed the material, confirmed
the subjects were minors, and affirmatively decided to take
“no action.” As Justice Thomas has observed, the courts
have extended Section 230 far beyond anything that plausibly follows from the text. See Snap, 144 S. Ct. at 2494
(Thomas, J., dissenting from the denial of certiorari). This
case presents the strongest occasion to correct that error.
Moreover, because this case addresses a degree of
knowledge not contemplated by previous decisions, it does
not implicate the policy concerns motivating those decisions. Many lower courts have focused as much attention
on mitigating policy concerns as they have on statutory interpretation, worrying that imposing liability on platforms
for user-generated content would impose “an impossible
burden in the Internet context.” Zeran v. America Online,
Inc., 129 F.3d 327 (4th Cir. 1997).
But Petitioners do not ask this Court to impose a duty
to monitor. They do not seek to hold Twitter liable for content it never reviewed, or for content it failed to detect amid
the volume of material posted to its platform. They merely
ask the Court to hold that a platform cannot claim blanket
statutory immunity after it has already monitored, already
verified, and already decided to retain content it knows to
be criminal. The distinction between a duty to monitor and
accountability for knowing retention of child sexual abuse
material is not a fine one. It is not for the Ninth Circuit to
grant platforms greater protection than the legislature has
chosen to provide.
17
III. Congress adopted Section 230 to codify the distinction between distributors and publishers,
not to create a blanket shield for corporations
that knowingly distribute child pornography.
Granting certiorari and reversing the Ninth Circuit’s
decision in this case would restore the original meaning of
Section 230. Subsection (c)(1) codified the common-law distinction between publishers and distributors. It removed
publisher-level strict liability so that platforms would not
be deterred from moderating content. It did not remove distributor liability—that is, liability predicated on actual
knowledge—because distributor liability was the mechanism by which the common law held intermediaries accountable for content they knew to be unlawful. The Ninth
Circuit’s contrary reading rests on Zeran v. America
Online, Inc., 129 F.3d 327 (4th Cir. 1997), which collapsed
the publisher/distributor distinction by treating all dissemination liability as “publisher” liability. That reading was
wrong when it was decided and this Court should not amplify it.
A. At common law, publisher liability and distributor liability were distinct doctrines with
different knowledge requirements.
At common law, publishers and distributors were subject to different standards of liability, reflecting different
degrees of involvement with the content at issue. Publishers, those who originate or select content for publication,
were strictly liable for the harmful material they disseminated. A newspaper that printed a defamatory letter, for
example, was liable even if it had no knowledge of the letter’s falsity. See W. Page Keeton et al., Prosser and Keeton
on the Law of Torts § 113 (5th ed. 1984).
18
Distributors—bookstores, libraries, newsstands, and
similar intermediaries—occupied a different position. Because they handled large volumes of material without the
capacity to review each item, the common law imposed a
lesser standard: a distributor was liable only if it knew or
had reason to know that the material was unlawful. See
Smith v. California, 361 U.S. 147, 153 (1959); see also Restatement (Second) of Torts §§ 578, 581 (Am. L. Inst. 1977).
The distinction was not a technicality. Its rationale was
both practical and moral.
Holding a bookstore strictly liable for every defamatory
passage in every volume on its shelves would impose an
impossible burden and restrict the flow of information. But
a bookstore that learns that a particular volume contains
defamatory content, and chooses to keep selling it, stands
in a different position. The common law recognized that
knowledge transforms the nature of an intermediary’s responsibility. That principle is no less applicable to internet
platforms than it was to their brick-and-mortar predecessors.
B. Section 230(c)(1) removed publisher liability;
it did not abolish distributor liability.
The text of subsection (c)(1) provides 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.” 47 U.S.C. §
230(c)(1) (emphasis added). The provision says nothing
about “distributors.” Had Congress intended to foreclose
distributor liability, it could have written “publisher,
speaker, or distributor,” or used the broader formulation
found in subsection (c)(2) (“shall be held liable on account
of”). It did neither.
Three canons of construction confirm this reading.
19
First, the canon against surplusage. If “publisher” encompasses all disseminators, including distributors, then
the word “speaker” in the same provision is superfluous—
a result this Court avoids. See Hibbs v. Winn, 542 U.S. 88,
101 (2004) (“A statute should be construed so that effect is
given to all its provisions, so that no part will be inoperative or superfluous, void or insignificant.”). The most natural reading is that “publisher” and “speaker” refer to the
two categories of persons who were strictly liable at common law—publishers and the original speakers of defamatory content—while distributors, who were liable only upon
knowledge, occupy a separate category that subsection
(c)(1) does not address.
Second, the common law background canon. Statutes
in derogation of common law are construed narrowly, in favor of retaining established principles. Isbrandtsen Co. v.
Johnson, 343 U.S. 779, 783 (1952) (“Statutes which invade
the common law . . . are to be read with a presumption favoring the retention of long-established and familiar principles, except when a statutory purpose to the contrary is
evident.”). The common law distinction between publisher
and distributor liability was a well-established principle at
the time Section 230 was enacted. A narrow construction
of subsection (c)(1), removing publisher liability but preserving distributor liability, is more consistent with this
canon than the expansive reading adopted by the Ninth
Circuit.
Third, the meaningful variation canon. Subsection
(c)(2) uses broader immunity language than (c)(1): it provides that no provider “shall be held liable on account of”
good-faith content-moderation actions. The phrase “held liable on account of” contemplates a materially broader set
of content than “treated as the publisher or speaker.” This
difference in language should be presumed meaningful. See
20
Russello v. United States, 464 U.S. 16, 23 (1983) (stressing
the significance of when “Congress includes particular language in one section of a statute but omits it in another
section of the same Act”); see also Malwarebytes, Inc., 141
S. Ct. at 18 (Thomas, J., respecting the denial of certiorari)
(observing that the face value of the statute offers a “modest understanding” that is “a far cry from what has prevailed in court”).
The Fourth Circuit’s contrary holding in Zeran v.
America Online, Inc., collapsed the publisher/distributor
distinction by holding that distributor liability is “merely a
subset, or a species, of publisher liability.” 129 F.3d at 332.
That conclusion was driven by policy concerns about the
burden that a notice-and-takedown regime would impose
on early-internet platforms—not by the statutory text. See
id. at 333. Whatever force those concerns had in 1997, they
have none in a case where the platform already received
notice, already investigated, and already decided to take no
action. Zeran’s reasoning does not extend to these facts,
and the Court should decline to extend it further.
C. Congress underscored the limited scope of
Section 230 by labeling it a “Good Samaritan”
provision.
Section 230(c) is titled “Protection for ‘Good Samaritan’
Blocking and Screening of Offensive Material.” Under this
Court’s precedents, a section heading is a “telling cue” as
to the provision’s scope. Yates v. United States, 574 U.S.
528, 539-40 (2015). The heading signals Congress’s design:
Section 230 protects platforms that affirmatively act to
block and screen harmful material—i.e., Good Samaritans
who intervene. Platforms that review harmful material,
identify it as criminal, and decide to leave it in place are
21
the opposite. Twitter’s conduct in this case is the antithesis
of Good Samaritan behavior.
Good Samaritan laws, in every jurisdiction that has
adopted them, protect those who attempt to render aid.
They do not protect those who distribute and profit from
illicit content. The same logic applies to Section 230(c).
Congress created the provision so that platforms would be
free to “restrict access to” material without fear of publisher liability. 47 U.S.C. § 230(c)(2)(A), (c)(2)(B); see Stratton Oakmont, Inc., 1995 WL 323710 (N.Y. Sup. Ct. 1995)
(the decision that prompted Section 230, holding that a
platform’s efforts to moderate content made it a “publisher”
subject to strict liability). That rationale has no application
to a platform that has already reviewed criminal content
and affirmatively decided to keep it online.
IV. Millions of images of child sexual abuse material
will continue proliferating unchecked online if
the Court does not clarify Section 230’s scope.
The Ninth Circuit’s interpretation turns the Good Samaritan provision on its head. Under its holding, the more
a platform knows about criminal content on its servers—
and the more deliberately it chooses to retain that content—the more certain its immunity becomes, because
each decision to retain content is a “publishing activity”
protected by subsection (c)(1). That reading converts a statute designed to encourage the removal of harmful material
into a statute that immunizes the deliberate retention of
criminal content. It is contrary to the text, structure, and
purpose of the statute Congress enacted.
By misconstruing Congress’s guardrails in Section 230,
lower courts have smashed the floodgates that could limit
child sexual abuse material online. When courts first
waded into liability issues animating the Communications
22
Decency Act in 1996, the National Center for Missing and
Exploited Children (NCMEC) hadn’t even created its Cyber
Tipline yet. See Stratton Oakmont, Inc., 1995 WL 323710,
at *4; The Missing and Exploited Children’s (MEC) Program: Background and Policies, CONG. RSCH. SERV. (Jul.
22, 2019), https://www.congress.gov/crs_external_products/RL/PDF/RL34050/RL34050.45.pdf. As the internet at
that time was in its nascent beginnings, so too was the
online circulation of child sexual abuse material and related reporting mechanisms. NCMEC’s Cyber Tipline
would go on to become the nation’s preeminent source for
reports of suspected child sexual exploitation from the public and electronic service providers.
In 2024, NCMEC’s Cyber Tipline received more than
20 million reports from electronic service providers containing more than 62 million images, videos, and other files
related to child sexual exploitation. National Center for
Missing & Exploited Children, 2024 CyberTipline Report
(2025), https://www.missingkids.org/content/dam/missingkids/pdfs/cybertiplinedata2024/2024-CyberTipline-Report.pdf. An overwhelming majority of those 20-plus million reports came from the very platforms that regularly
invoke Section 230’s haven. National Center for Missing &
Exploited Children, 2024 CyberTipline Reports by Electronic Service Providers (ESPs) (2025), https://www.missingkids.org/content/dam/missingkids/pdfs/cybertiplinedata2024/2024-reports-by-esp.pdf.
If the Court does not grant certiorari to course-correct
the scope of Section 230, these figures will undoubtedly
rise. This Court has recognized as much. Snap, 144 S. Ct.
at 2494 (Thomas, J., dissenting from the denial of certiorari); see also Doe v. Facebook, Inc., 142 S. Ct. at 1088
(Thomas, J., respecting the denial of certiorari) (recognizing that, “[a]t the very least, before we close the door on
23
such serious charges, ‘we should be certain that is what the
law demands.’” (quoting Malwarebytes, Inc., 141 S. Ct. at
18 (Thomas, J., respecting the denial of certiorari))).
As this case unfortunately illustrates, these platforms
are a hotbed for child sexual abuse material. Twitter, for
example, reported 686,176 instances of suspected child sexual exploitation in 2024. National Center for Missing & Exploited Children, 2024 CyberTipline Reports by Electronic
Service Providers (ESPs) (2025), https://www.missingkids.org/content/dam/missingkids/pdfs/cybertiplinedata2024/2024-reports-by-esp.pdf. That figure is
more than fifteen-fold Twitter’s 2019 reporting (45,726).
National Center for Missing & Exploited Children, 2019
CyberTipline Reports by Electronic Service Providers
(ESPs)
(2020),
https://www.missingkids.org/content/dam/missingkids/pdfs/2019-reports-by-esp.pdf.
Time is of the essence for exploited children all over the
world. Fortunately, Section 230 itself provides the remedy
against violative platforms—if this Court summons the
will to say so.
CONCLUSION
The decision below should be reversed.
Respectfully submitted,
JOSH HAWLEY
U.S. SENATOR
Counsel of Record
WILLIAM SIMPSON
DANIEL PARK
AARON YURATOVICH
24
115 Russell Senate Office Bldg.
Washington, D.C. 20510
(202) 224-6154
William_Simpson@hawley.senate.gov
March 12, 2026
This is a copy of a public record, reproduced as it was published. It is not legal advice, and it may not be the version a court would rely on. Check the official source before you cite it.