Amicus Curiae Brief — Reynaldo Gonzalez, et al., Petitioners v. Google LLC
Supreme Court briefJan 19, 2023
Ask Donna
What actually matters in this document.
Text
No. 21-1333
In the Supreme Court of the United States
REYNALDO GONZALEZ, ET AL., Petitioners,
v.
GOOGLE LLC.
On Writ of Certiorari to
the United States Court of Appeals
for the Ninth Circuit
BRIEF FOR REASON FOUNDATION
AS AMICUS CURIAE
SUPPORTING RESPONDENT
MANUEL S. KLAUSNER
ERIK S. JAFFE
LAW OFFICES OF
Counsel of Record
JOSHUA J. PRINCE
MANUEL S. KLAUSNER
5538 Red Oak Drive
SCHAERR|JAFFE LLP
Los Angeles, CA 90068
1717 K Street NW
(213) 675-1776
Suite 900
mklausner@klausnerlaw.us Washington, DC 20006
(202) 787-1060
ejaffe@schaerr-jaffe.com
Counsel for Amicus Curiae
TABLE OF CONTENTS
Page
TABLE OF AUTHORITIES ....................................... iii
INTRODUCTION AND INTEREST OF
AMICUS CURIAE ............................................... 1
SUMMARY OF ARGUMENT ..................................... 3
ARGUMENT ................................................................ 8
I.
Congress intended and correctly
predicted that Section 230 would
generate enormous benefits for free
speech, technological innovation, and
the economy. ................................................... 8
A.
Congress’s enacted legislative
judgment regarding the purposes
and benefits of Section 230 take
priority over any competing policy
claims or concerns raised by third
parties or even the Court itself. ............. 9
B.
Congress accurately predicted the
benefits of Section 230, which
allows providers and users of
interactive computer services to
innovate, collectively expand the
reach of protected speech, and
generate economic growth without
the crushing transaction cost of
liability based on the speech of
others. .................................................... 14
ii
II. Algorithms or other actions that merely
suggest information content provided
by another content provider do not
generally convert that content into
“information
provided
by”
the
recommending user or provider of an
interactive computer service. ....................... 23
A.
The plain text of Section 230
draws a distinction between
redistribution
of
“information
provided by another information
content provider” and information
content provided by a user or
interactive computer service itself. ...... 24
B.
Algorithms that point users of
interactive computer services to
the content of others generally do
not themselves create or develop
content from the recommending
person or entity. .................................... 26
CONCLUSION .......................................................... 31
iii
TABLE OF AUTHORITIES
Cases
Page(s)
Batzel v. Smith,
333 F.3d 1018 (9th Cir. 2003)................................ 25
BNSF Ry. Co. v. Loos,
139 S. Ct. 893 (2019) ................................................ 9
Epic Sys. Corp. v. Lewis,
138 S. Ct. 1612 (2018) ............................................ 10
Flynn v. Cable News Network, Inc., No. 1:21CV-2587-GHW, 2022 WL 3334716 (S.D.N.Y.
Aug. 12, 2022) ........................................................ 28
Force v. Facebook, Inc.,
934 F.3d 53 (2d Cir. 2019) ............. 12, 15, 26, 27, 29
Hall v. United States,
566 U.S. 506 (2012) ................................................ 11
Jane Doe No. 1 v. Backpage.com, LLC,
817 F.3d 12 (1st Cir. 2016) .................................... 12
Jones v. Dirty World Ent. Recordings LLC,
755 F.3d 398 (6th Cir. 2014).................................. 25
Kisor v. Wilkie,
139 S. Ct. 2400 (2019) ............................................ 12
Little Sisters of the Poor Saints Peter & Paul
Home v. Pa., 140 S. Ct. 2367 (2020) ...................... 11
Miami Herald Pub. Co. v. Tornillo,
418 U.S. 241 (1974) ................................................ 19
Rosenberger v. Rector & Visitors of the Univ. of
Va., 515 U.S. 819 (1995) ........................................ 30
iv
Cases—continued:
Zedner v. United States,
547 U.S. 489 (2006) ................................................ 10
Zeran v. America Online, Inc.,
129 F.3d 327 (4th Cir. 1997).............................. 5, 18
Statutes
47 U.S.C. § 230(a)(1) .................................................. 10
47 U.S.C. § 230(a)(3) .................................................... 1
47 U.S.C. § 230(a)(4) .............................................. 1, 11
47 U.S.C. § 230(a)(5) .................................................. 11
47 U.S.C. § 230(b)(1) .............................................. 4, 11
47 U.S.C. § 230(b)(2) .............................................. 4, 11
47 U.S.C. § 230(b)(3) .............................................. 4, 11
47 U.S.C. § 230(c)(1) .... 1–3, 6, 8, 10, 12, 23, 24, 27, 31
47 U.S.C. § 230(f)(2) .............................................. 6, 24
47 U.S.C. § 230(f)(3) .................................................. 24
47 U.S.C. § 230(f)(4)(B) ................................... 6, 25, 29
47 U.S.C. § 230(f)(4)(C) ............................. 6, 25, 29, 31
Bills
Biased Algorithm Deterrence Act of 2019, H.R.
492, 116th Cong. (2019) ......................................... 13
Justice Against Malicious Algorithms Act of
2021, H.R. 5596, 117th Cong. (2021) .................... 13
Platform Integrity Act, H.R. 9695, 117th Cong.
(2022) ...................................................................... 13
v
Bills—continued:
Protecting Americans from Dangerous
Algorithms Act, H.R. 8636, 116th Cong.
(2020) ...................................................................... 13
Other Authorities
@HawleyMO, Twitter (Nov. 27, 2018, 11:22
AM), https://tinyurl.com/33fa3fxm ........................ 19
Elizabeth Banker, Understanding Section 230
& the Impact of Litigation on Small
Providers (2022),
https://tinyurl.com/2nn66m87 ............................... 22
Susan Benkelman, The Law that made the
Internet what it is today, The Washington
Post (Apr. 26, 2019),
https://tinyurl.com/2rkc9b5x ................................. 17
Matt Binder, Truth Social already censoring
content, bans user who made fun of Trump
Media CEO, Mashable (Feb. 22, 2022),
https://tinyurl.com/yc87hunc ................................. 20
Elizabeth Nolan Brown, 4 Cases That Show the
Scope of Services, Speech, and Conduct
Protected by Section 230, Reason (Jan. 25,
2019), https://tinyurl.com/3v63bvb5 ..................... 14
Elizabeth Nolan Brown, Expect More
Conservative Purges on Social Media If
Republicans Target Section 230, Reason
(Nov. 28, 2018),
https://tinyurl.com/bddn2mde ............................... 19
vi
Other Authorities—continued:
Elizabeth Nolan Brown, Section 230 Is the
Internet's First Amendment. Now Both
Republicans and Democrats Want To Take It
Away., Reason (Jul. 29, 2019),
https://tinyurl.com/2x5zh9vd....................... 5, 18, 19
Jonathan Cannon & Adam Thierer, Quite a
Fall for Digital Tech, Discourse (Dec. 7,
2022), https://tinyurl.com/yd43h3ua ..................... 14
David N. Cicilline, Cicilline, Buck Release Bill
to Hold Online Platforms Accountable for
Promoting Extremism (Dec. 27, 2022),
https://tinyurl.com/mnhsutz .................................. 13
Daniel A. Crane, Enacted Legislative Findings
and the Deference Problem, 102 Geo. L.J. 637
(2014) ........................................................................ 9
Internet Association, Best Of The Internet:
Consumers Rely On User-Generated Content
(2019), https://tinyurl.com/yjtde9tc ....................... 22
Ashley Johnson & Daniel Castro, Overview of
Section 230: What It Is, Why It Was Created,
and What It Has Achieved, Information
Technology & Innovation Foundation (Feb.
22, 2021), https://tinyurl.com/2kfhs7wh ............... 17
Jeff Kosseff, The Twenty-Six Words That
Created the Internet (2019) ...................................... 1
Leighanna Mixter, Three principles in CDA
230 that make Wikipedia possible, Diff (Nov.
9, 2017), https://tinyurl.com/yc5d4w55 ................. 17
vii
Other Authorities—continued:
David Post, Content Moderation, Social Media,
and the Constitution, Reason (Nov. 1, 2022),
https://tinyurl.com/ms2tcwru ................................ 16
Russell Rothstein, The gig economy and
customer reviews: Trust in the digital era,
The Future of Commerce and Customer
Engagement (2016),
https://tinyurl.com/yc3pupju ................................. 22
Scott Shackford, Supreme Court To Hear 2
Cases About Social Media Moderation and
Liability for Terrorism, Reason (Oct. 3,
2022), https://tinyurl.com/mt24k3bm ................... 16
Ethan Wham, An Economic Case for Section
230, Disruptive Competition Project (Sept. 6,
2019), https://tinyurl.com/2m6kvam5 ............. 21, 22
INTRODUCTION AND INTEREST
OF AMICUS CURIAE1
When Congress adopted Section 230, it made legislative findings articulating the purposes and benefits it perceived from that provision. Congress found
that the Internet and other interactive computer services, taken as a whole, could serve as “a forum for a
true diversity of political discourse,” “offer * * *
unique opportunities for cultural development,” and
provide “myriad avenues for intellectual activity.”
47 U.S.C. § 230(a)(3). It also recognized the benefits
that flowed to the American people when the Internet
was allowed to grow “with a minimum of government
regulation.” Id. § 230(a)(4). To implement such limits
on government regulation, it adopted Section 230, the
key provision of which 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.”
Id. § 230(c)(1).
The 26 words of Section 230(c)(1) have “come to
mean that, with few exceptions, websites and Internet service providers are not liable for the comments,
pictures, and videos that their users and subscribers
post, no matter how vile or damaging.”2 Insofar as
there is any uncertainty in interpreting those words,
1 Pursuant to Supreme Court Rule 37.6, no counsel for any
party authored this brief in whole or in part and no entity or
person, aside from Amicus Curiae, its members, and its counsel,
made any monetary contribution toward the brief’s preparation
or submission.
2 Jeff Kosseff, The Twenty-Six Words That Created the Internet 8 (2019).
2
Congress’s expansive legislatively declared purposes
favor a broad reading of Section 230’s protections for
providers and users of interactive computer services,
not a cramped one.
Because an incorrect reading of Section 230(c)(1)
would be enormously harmful to the development of
new technology, the furthering of free speech, and the
online economy, this case interests Amicus Reason
Foundation (Reason). Reason is a national, nonpartisan, and nonprofit public policy think tank, founded
in 1978. Reason’s mission is to advance a free society
by applying and promoting libertarian principles and
policies—including free markets, individual liberty,
and the rule of law. Reason supports dynamic market-based public policies that allow and encourage
individuals and voluntary institutions to flourish.
Reason advances its mission by publishing Reason
magazine, as well as commentary on its websites,
and by issuing policy research reports. To further
Reason’s commitment to “Free Minds and Free Markets” and equality before the law, Reason selectively
participates as amicus curiae in cases raising significant issues.
Amicus agrees with Respondent on the merits of
the question presented. The plain text of Section 230
precludes claims that would treat YouTube as the
publisher of videos created by third parties solely because YouTube uses an algorithm that organizes and
presents the videos of its users to others who might
be interested in them. They write separately to address how, consistent with Congress’s legislative findings, the economy and the public indeed have benefited from Section 230. And even if there is debate or
disagreement about the policy pros and cons of Sec-
3
tion 230, it is the congressional findings and purposes
that take priority and favor broad protection of providers and users should it become necessary to resolve any ambiguity in the text. Amicus also writes to
address a key textual distinction in the statute—
between information provided by a “provider” or “user” of an interactive computer service themselves
versus information provided by “another information
content provider.”
SUMMARY OF ARGUMENT
I. For nearly three decades, Section 230 has
served as the backbone of the Internet, precisely as
Congress correctly anticipated and intended. The legislatively enacted congressional findings and purpose
favor an expansive reading of Section 230’s protections in the event of any uncertainty or perceived
ambiguity in the language of Section 230(c)(1).
A. Section 230’s benefits were by design, even if
Congress could not have predicted every detail—or
challenge—of a growing Internet. What Congress did
know is that, for the Internet to grow, it had to be left
alone without fear of the “litigation minefield,” Resp.
Br. 19, that would cripple its expansion in its infancy
if the providers and users of interactive computer
services could be found liable for the content created
by others. Congress thus enacted Section 230 with a
list of policy statements that show what it intended
and expected the statute to do: protect platforms and
users from liability for the speech of others and promote the growth and use of interactive computer services.
Congress explained that the goal of Section 230 is
to “promote the continued development of the Inter-
4
net” by, among other things, “encourag[ing] the development of technologies which maximize user control over what information is received by” those “who
use the Internet and other interactive computer services.” Id. (b)(1), (3). Section 230 has done that. Congress also expressed the importance of “preserv[ing]
the vibrant and competitive free market that presently exists for the Internet and other interactive computer services, unfettered by Federal or State regulation.” Id. (b)(2) (emphasis added). Section 230 has
created that world, too.
Those policy statements are not mere pieces of legislative history entered into the Congressional Record
by opportunistic politicians or their staffers—to the
contrary, they are the product of bicameralism and
presentment just like any other duly enacted legislation. And such statements are entitled controlling
weight regarding what policy considerations might
potentially influence the interpretation of Section
230. Whether Section 230 creates good policy is not a
question for this Court to decide. That question remains where it was in 1996—with Congress.
B. Even years after Congress’s legislative findings
and purpose, Section 230 has overwhelmingly fulfilled such legislative predictions and goals. By
providing immunity from liability for the content
posted by others, it has allowed for the development
of new technologies that make it easier for everyone
to find information online, to organize and to let others help organize the information they receive, and to
associate both directly and indirectly with people
around the world sharing common interests. These
advances in technology have also led to the development of all manner of social-media sites, including
5
video-based platforms, dating apps, and even improved traditional chatrooms providing users many of
the same organization tools as providers themselves.
The improved ability to find and organize information online is only one of the many benefits of Section 230. It also has led to an exponential growth in
the amount of speech online. As providers have innovated and users have enthusiastically participated in
online speech free from “the “specter of liability,” Zeran v. America Online, Inc., 129 F.3d 327, 331 (4th
Cir. 1997), interactive computer services have made
it easier for ideas to spread than ever before in human history. Through retweets and other user engagements, the views and content created by even
the poorest Americans can spread around the country
and world in a way that wouldn’t have been possible
just twenty years ago.3
Other benefits from Section 230 abound. The economic benefits to innovators, providers, users, and
the economy as a whole have been tremendous. It has
facilitated the gig economy by allowing individuals
and small businesses to flourish on websites provided
by bigger platforms. It has also allowed consumers to
directly review products and other services, make
those reviews readily available online for the next
consumer, and pass along or comment upon reviews
by others, thus democratizing the marketplace of
products and services as well as the marketplace of
3 E.g., Elizabeth Nolan Brown, Section 230 Is the Internet's
First Amendment. Now Both Republicans and Democrats Want
To
Take
It
Away.,
Reason
(Jul.
29,
2019),
https://tinyurl.com/2x5zh9vd.
6
ideas. Thus, insofar as such practical considerations
matter to the interpretation of Section 230(c)(1), the
findings and purposes of Congress are not only controlling, they are right.
II. The language of Section 230 both reflects such
Congressional policy and confirms that Respondent
should prevail in this case.
A. An “interactive computer service” “provides or
enables computer access by multiple users to a computer server.” 47 U.S.C. § 230(f)(2). “Interactive computer services” expressly include “access software
providers,” which—as relevant here—are providers of
software or tools that can “pick, choose, analyze, or
digest,” “transmit, receive, display, forward, cache,
search, subset, organize, reorganize, or translate content.” Id. (f)(2), (4)(B), (C). The providers of such services and their users can both create their own information content and can organize, transmit, and
provide access to information content provided by
others.
B. YouTube’s algorithm, which organizes and reorganizes the content uploaded to YouTube by others,
thus performs a function which Congress expressly
included in the definition of an interactive computer
service. Indeed, as both a provider and user of such
software, Respondent falls squarely within the class
protected by Section 230(c)(1). Insofar as Petitioners
are seeking to hold Google liable for the consequences
of having presented or organized the “information
provided by another,” rather than for creating and
publishing Google’s own information content, Section
230(c)(1) bars such liability.
7
To the extent any given algorithm or other organizational policy or choice might be said to create
Google’s own “content,” the further question becomes
the precise parameters of such content as distinguished from the content of others. That distinction
helps clarify that even where an algorithm or other
organizational action or policy itself might create
some information content (appending a warning label
for example), a user or provider may only be held responsible for that information alone, and not the underlying information “provided by another.” Alternatively, if YouTube or any other user of its service
were to expressly adopt or endorse the information
content of another as its own, such adopted content
may well fall outside of Section 230’s protection.
But merely identifying, organizing, or even recommending the content of another is a far cry from
adopting it as your own. YouTube’s algorithm, for example, analyzes different users’ activity and viewing
behavior to predict what that user might find interesting and to organize further information content
provided by others according to such predictions.
Though the algorithm’s analysis and predictions are
more automated and sophisticated than manual efforts to organize or recommend content in a manner
appealing to users, it remains fundamentally the
same as the manual choices exercised by chatroom
moderators, bloggers, and indeed, any individual user
who selects, reposts, “likes,” or otherwise passes
along the information content of others in a way such
user believes might be interesting or appealing to her
followers and potential followers. Such organizational
effort by both providers and users of interactive computer services is precisely what Congress anticipated
8
and intended to encourage via Section 230, and the
text provides broad protection reflecting that purpose.
ARGUMENT
I. Congress intended and correctly predicted
that Section 230 would generate enormous
benefits for free speech, technological innovation, and the economy.
While Amicus agrees with Respondent that the
plain language of Section 230(c)(1) protects Google
against claims based on its algorithms organizing the
content of others, to the extent that this Court perceives any uncertainty in such language, that uncertainty should be resolved according to the findings
and purposes expressly adopted by Congress, and not
based on any competing policy concerns expressed by
Petitioners, their Amici, or even members of this
Court. Congress favored broad protection for providers and users of interactive computer services regarding their organization and passing along of the information content of others. And it was right to do so.
The issues presented by this case risk seriously undermining the legislative judgments of Congress and
undermining the many benefits that Section 230 provides to the public and interactive computer services
alike. To the extent that time and technological developments have created new challenges or concerns,
addressing such issues, weighing them against the
original predictions and policies of Section 230, and,
if necessary, adjusting the protections afforded by
that section are tasks for Congress, not this Court.
Any uncertainty regarding the language of Section
230(c)(1) should be resolved according to Congress’s
9
expansive and protective express intent adopt in conjunction with that provision.
A. Congress’s enacted legislative judgment
regarding the purposes and benefits of
Section 230 take priority over any competing policy claims or concerns raised by
third parties or even the Court itself.
As will be discussed more fully below, Section 230
has in fact been enormously beneficial to interactive
computer services and the public alike. This was by
design. Indeed, although Congress could not have
understood all of the specific benefits Section 230
would have for the economy and the public, what it
did conclude was that interactive computer services
needed to be protected from ruinous litigation if the
Internet was to thrive.
1. One need not guess about what Congress intended to do when it passed Section 230—Congress
told us. The statute includes both congressional findings and general policy statements. To be sure, several members of this Court have reflected on the folly
of looking to generalized and often cherry-picked legislative history to divine the will of Congress as a
whole. But in response to such concerns “Congress
increasingly makes such findings in the course of enacting statutes.”4 Enacted findings provide “the sort
of textual evidence everyone agrees can sometimes
shed light on meaning.” See BNSF Ry. Co. v. Loos,
139 S. Ct. 893, 906 (2019) (Gorsuch, J., dissenting).
Daniel A. Crane, Enacted Legislative Findings and the
Deference Problem, 102 Geo. L.J. 637, 639 (2014).
4
10
Enacted findings and policy statements are worthy
of this Court’s considerations because they lack the
problems of general statements in legislative history.
Thus, while legislative history is “not the law,” Epic
Sys. Corp. v. Lewis, 138 S. Ct. 1612, 1631 (2018), enacted policy statements are. Similarly, while courts
cannot divine “the view of Congress as a whole” from
“what is said by a single person in a floor debate or
by a committee report,” Zedner v. United States, 547
U.S. 489, 510 (2006) (Scalia, J., concurring in part
and in the judgment), enacted policy statements or
factual findings are the product of the constitutional
requirements of bicameralism and presentment. Accordingly, to the extent that a court finds the operative provisions of a statute ambiguous, any legislative
policy statements or factual findings are entitled substantial weight in resolving that ambiguity, particularly as compared to conflicting policy concerns or
factual propositions offered by others either before or
after the statute was adopted.
2. Those policy statements and factual findings
enacted as part of Section 230 are particularly illuminating and show why Section 230(c)(1) is correctly
read to protect Google or any other provider or user of
an interactive computer service from liability for organizing and presenting third-party content to other
users through an algorithm or any other means of organizing such third-party content.
Starting first with the findings, Congress found
that the Internet’s rapid growth represented “an extraordinary advance in the availability of educational
and informational resources.” 47 U.S.C. § 230(a)(1).
That was true when Section 230 was enacted, and
the Internet’s ability to make information available
11
has only expanded—and exponentially so—since.
Congress also found that the Internet, when left
“with a minimum of government regulation,” offers
“myriad
avenues
for
intellectual
activity.”
47 U.S.C. § 230(a)(4), (5).
Congress’s findings about the Internet’s potential
informed the purpose that it established when it enacted Section 230. That purpose could not have been
clearer—Congress intended the government, including the judiciary, to get out of the way of the Internet’s growth. Congress explained that the goal of Section 230 is to “promote the continued development of
the Internet” by, among other things, “encourag[ing]
the development of technologies that maximize user
control over what information is received by” those
“who use the Internet and other interactive computer
services.” Id. (b)(1), (3). And Congress expressed its
goal of “preserv[ing] the vibrant and competitive free
market that presently exists for the Internet and
other interactive computer services, unfettered by
Federal or State regulation.” Id. (b)(2) (emphasis added).
These findings and policy considerations should
guide the Court’s interpretation of Section 230 even if
members of the Court disagree with the policy Congress enacted or believe that time and circumstance
have presented new and unanticipated concerns. As
this Court has explained, “[i]t is not for us to rewrite
the statute so that it covers only what we think is
necessary to achieve what we think Congress really
intended.” Little Sisters of the Poor Saints Peter &
Paul Home v. Pa., 140 S. Ct. 2367, 2381 (2020) (citation omitted); accord Hall v. United States, 566 U.S.
506, 523 (2012) (“[T]here may be compelling policy
12
reasons” to reach a particular result, “[b]ut if Congress intended that result, it did not so provide in the
statute. Given the statute’s plain language, context,
and structure, it is not for us to rewrite the statute[.]”).
This Court should thus follow the path taken by
other courts around the country that have looked to
Congress’s stated objectives to conclude that “Section
230(c)(1) should be construed broadly in favor of immunity,” Force v. Facebook, Inc., 934 F.3d 53, 64 (2d
Cir. 2019), and that Section 230 “should not be construed grudgingly,” Jane Doe No. 1 v. Backpage.com,
LLC, 817 F.3d 12, 18 (1st Cir. 2016). In the process,
the Court should expressly reject the Petitioners’
suggestion, at 18–25, that it rewrite Section 230 to
conflict with Congress’s broad policy goals by “restrain[ing] the automation of interactive computer
services’ publishing activities” as a precondition to
Section 230 immunity. Force, 934 F.3d at 67.
The policy positions of Section 230, moreover,
should govern even though there are many in Congress who seemingly disagree with Section 230’s
reach and the policy considerations that led to its enactment. Congress remains free to amend the statute
through its own legislative efforts, not by some of its
members encouraging this Court to do the work for
them. Kisor v. Wilkie, 139 S. Ct. 2400, 2439 (2019)
(Gorsuch, J., concurring in judgment) (“If Congress
disagrees with how courts are interpreting an existing statute, it is free to amend the statute to establish a different rule going forward.”).
Many in Congress have been trying to adopt such
amendments. Numerous legislative options have
13
been proposed to respond to the policy concerns
raised by Petitioners here. Just a few weeks before
the filing of this brief, for example, a bipartisan group
introduced the Platform Integrity Act, H.R. 9695,
117th Cong. (2022), a bill expressly designed to supersede judicial holdings that Section 230 “bars victims of terrorist attacks from seeking relief from a social-media company for its proactive role connecting
the perpetrators through friend- and contentsuggestion algorithms.”5 Another bill, the Justice
Against Malicious Algorithms Act of 2021, H.R. 5596,
117th Cong. (2021), would have done essentially the
same thing. The Biased Algorithm Deterrence Act of
2019, H.R. 492, 116th Cong. (2019), would remove
Section 230 immunity when an algorithm reorders
user content in a way that is not chronological. The
Protecting Americans from Dangerous Algorithms
Act, H.R. 8636, 116th Cong. (2020), would strip platforms of immunity whenever an algorithm amplifies
content tied to or promoting terrorism or even violations of civil rights.
The virtues or folly of such proposals, and the details of balancing their competing policy concerns, is
a matter for the political branches, not this Court. As
it stands, Congress has established the immunityfavoring policy of the United States, and that policy
should guide this Court’s interpretation of the operative provisions of Section 230 until Congress legislatively adopts a different or modified policy or limitation on Section 230 immunity.
5 David N. Cicilline, Cicilline, Buck Release Bill to Hold
Online Platforms Accountable for Promoting Extremism (Dec.
27, 2022), https://tinyurl.com/mnhsutz.
14
B. Congress accurately predicted the benefits of Section 230, which allows providers
and users of interactive computer services to innovate, collectively expand the
reach of protected speech, and generate
economic growth without the crushing
transaction cost of liability based on the
speech of others.
In addition to the weight that should be accorded
to the legislative facts and purposes adopted by Congress, Section 230 is indeed acting precisely as Congress intended. In the nearly 30 years since Section
230 was enacted, it has played an incalculable role in
the way that Americans live in the digital age.
The key benefits from Section 230 stem from the
innovation it facilitates. As several commentors have
explained, Section 230 has served as the “permissionless innovation model that fueled America’s commanding lead in the digital information revolution,”
has allowed “digital entrepreneurs” to be “free to
launch bold new ideas without fear of punishing lawsuits or regulatory shenanigans,” and has, in turn,
“boosted economic growth and dramatically broadened consumer information and communications options.”6 In short, Section 230 “allowed * * * the digital
economy as we know it[.]”7
6 Jonathan Cannon & Adam Thierer, Quite a Fall for Digital
Tech, Discourse (Dec. 7, 2022), https://tinyurl.com/yd43h3ua.
7 Elizabeth Nolan Brown, 4 Cases That Show the Scope of
Services, Speech, and Conduct Protected by Section 230, Reason
(Jan. 25, 2019), https://tinyurl.com/3v63bvb5.
15
Section 230 was able to do this because it protected internet platforms from liability for the content
created by others. This in turn allowed companies to
experiment with new technologies to improve the experiences of their users without fear. That has led to
incredible advancements in technology that benefit
the providers and users of interactive computer services, and the public as a whole. That many providers
and users of such services also often receive substantial financial rewards from such advancements benefits both them and the economy and incentivizes further innovations.
Consider the algorithms at issue here. They are
one such technology that has developed because of
Section 230’s protections. Now, rather than awaiting
active input and effort from a new user to match the
content of an interactive content provider to that user, algorithms allow providers and users of interactive computer services to predict what information
would be relevant to a particular user based on information provided by that user in the past and by
that user’s actual behavior in the past, present, and
future. See Force, 934 F.3d at 70 (“The algorithms
take the information provided by Facebook users and
‘match’ it to other users—again, materially unaltered—based on objective factors applicable to any
content, whether it concerns soccer, Picasso, or
plumbers.”). Throughout the process, the user—who
has control over whether to use any given service at
all and generally over what information she shares
with the interactive computer service—is in control of
the information she receives, even if the content is
shared with her predictively.
16
Staying within the Google umbrella, without the
protections of Section 230, it would be easy to imagine the countless ways that Google’s search engine
could face potential liability for the answers it provides to users for even the most innocuous search. By
protecting Google from liability, Section 230 allowed
Google to create a search engine—guided by a different, though similar, algorithm—that is more responsive to the needs of its user base by predicting which
of the millions of answers to a query most likely meet
the needs and interests of the particular user. Going
back to YouTube itself, one commentator has explained that, “[i]f Google loses its case, the probable
outcome will be a significant reduction of video recommendations on YouTube, making it harder for
people to find videos related to what they’re viewing
and harder for content providers to reach viewers.”8
Another quipped that “it is hard to imagine what the
entire social media ecosystem would look like if platforms could be held liable for hosted third-party content.”9 Both are right—particularly because, whatever the uncertainty of what could have been without
Section 230, what is certain is that both the providers
and the users of interactive computer services would
be more cautious in the organizational services they
provide and the third-party content they present,
“like,” repost, recommend, or even reference lest they
Scott Shackford, Supreme Court To Hear 2 Cases About
Social Media Moderation and Liability for Terrorism, Reason
(Oct. 3, 2022), https://tinyurl.com/mt24k3bm.
8
9 David Post, Content Moderation, Social Media, and the
Constitution,
Reason
(Nov.
1,
2022),
https://tinyurl.com/ms2tcwru.
17
be held liable for the information content of such
third-party materials.
For that reason, it is no exaggeration to say that
the fallout from a narrow reading of Section 230
would affect the entire Internet. “[M]any common
features of websites, such as user reviews and comments, exist because of the liability protection offered
by Section 230.”10 Wikimedia—the company that
runs the online user-driven encyclopedia Wikipedia—
expressly credits Section 230 with making Wikipedia
possible.11
While it is a truism that the Congress faced with
protecting a “nascent consumer Internet” could not
have fully anticipated the future twists and turns of
the Internet and may have “struggled to envision how
this new technology would work,”12 the limited ability
to predict the future is true of every legislative effort
and irrelevant to this Court’s task. Regardless
whether the members of Congress could have predicted the precise details of the future they were
helping create, they were certainly correct about the
tremendous innovation they sought to encourage. In
the nearly 30 years since Section 230 became law, the
10 Ashley Johnson & Daniel Castro, Overview of Section 230:
What It Is, Why It Was Created, and What It Has Achieved, Information Technology & Innovation Foundation (Feb. 22, 2021),
https://tinyurl.com/2kfhs7wh.
11 Leighanna Mixter, Three principles in CDA 230 that make
Wikipedia
possible,
Diff
https://tinyurl.com/yc5d4w55.
(Nov.
9,
2017),
12 Susan Benkelman, The Law that made the Internet what
it is today, The Washington Post (Apr. 26, 2019),
https://tinyurl.com/2rkc9b5x.
18
Internet has exploded because its protections have allowed providers and users of interactive computer
services to test new technologies—to the great benefit
of the Nation as a whole.
2. The technological innovations made possible by
Section 230 have also greatly increased the ability of
the average American to spread ideas. As the Fourth
Circuit recognized just months after Section 230 was
enacted, without Section 230 protections, the “specter
of tort liability in an area of such prolific speech
would have an obvious chilling effect.” Zeran v. America Online, Inc., 129 F.3d 327, 331 (4th Cir. 1997).
Section 230 protections, by contrast, allow interactive
computer services to provide their users, rich and
poor alike, with a reach that, historically, would not
have been available even to the most privileged classes with access to the gatekeepers of the institutional
press. Indeed, because of its ability to expand the
reach of speech, Section 230 has been described as
“the internet’s First Amendment—possibly better.”13
Further, Section 230 promotes the distribution of
all speech—even speech that may be unpopular. In a
way, the speech that Section 230 facilitates is “something like the way people actually communicate—
online and off—in a world without top-down government control of our every utterance and interaction.”14 True, free speech on the Internet can be
“messy,” “florid,” “outrageous,” “offensive,” “unpre13 Elizabeth Nolan Brown, Section 230 Is the Internet's First
Amendment. Now Both Republicans and Democrats Want To
Take
It
Away.,
Reason
(Jul.
29,
2019),
https://tinyurl.com/2x5zh9vd.
14 Ibid.
19
dictable, uncontrollable, and frustrating.”15 But at
the same time, it can also be “illuminating and informative.”16 But whatever the mix of speech on particular services and across the many interactive services collectively, that is the nature of free speech
and, at least in our constitutional system, remains an
overall public good. By protecting interactive computer services from liability for the speech of others,
Section 230 has given room for all that speech to
spread.
That is not, of course, to say that anything “in Section 230 requires that an entity maintain ‘a true diversity of political discourse.’”17 After all, if Congress
did impose such a requirement as a precondition to
Section 230 immunity, it would raise serious First
Amendment questions. Cf. Miami Herald Pub. Co. v.
Tornillo, 418 U.S. 241, 258 (1974) (holding that
newspapers cannot be forced to give access to all
viewpoints). Instead, the benefits to free speech come
from the sheer number of platforms available to the
American people that have only been able to rise to
prominence because of the protections of Section 230.
Thus, if certain platforms promote or limit speech
about a particular topic—or even a particular viewpoint that the platform owners find preferable or offensive—users are free to find another platform with
15 Ibid.
16 Ibid.
17 Elizabeth Nolan Brown, Expect More Conservative Purges
on Social Media If Republicans Target Section 230, Reason
(Nov. 28, 2018), https://tinyurl.com/bddn2mde (quoting
@HawleyMO, Twitter (Nov. 27, 2018, 11:22 AM),
https://tinyurl.com/33fa3fxm).
20
different terms of service. Put differently, if Twitter,
Instagram, or Facebook decide to promote a particular viewpoint or perspective and decide that a different viewpoint is offensive, other services are available to promote competing views and disfavor other
perspectives or content that they and their users find
objectionable. And, of course, Section 230 provides
the same protections to the smaller social media platforms that it does to the biggest platforms. They too
can—and have—engaged in censorship of particular
viewpoints.18
Furthermore, Section 230 provides its protection
not only to the “providers” of interactive computer
services, but to the “users” of such services as well.
Removing immunity from Google here would equally
remove immunity for persons hosting humble chat
rooms, interest- or politics-focused blogs, and even for
persons who “like” or repost the information content
of others on their blog, their Facebook page, or their
Twitter account. That the “algorithm” such users apply when deciding what to like, recommend, or republish to their followers may often be entirely in their
heads, as opposed to incorporated into computers,
makes no difference if—as Petitioners argue—the
point is that merely recommending or promoting the
information content of another makes you liable for
any alleged harms from such content. Narrowing Section 230’s protections thus not only would chill activi-
18 Truth Social, for example, famously deleted a user account for pretending to be a cow owned by the CEO of Trump
Media, Devin Nunes. Matt Binder, Truth Social already censoring content, bans user who made fun of Trump Media CEO,
Mashable (Feb. 22, 2022), https://tinyurl.com/yc87hunc.
21
ties by providers, large and small, it would chill the
activity of millions of users as well.
3. The technological advancements and the benefits to free speech that Section 230 facilitates also
have led to enormous economic growth.
The economic benefits that stem from Section 230
manifest in various ways. First, Section 230 “makes
user-submitted reviews” of products “possible.”19
Such user reviews increase the likelihood that a user
will find a product that meets their needs and that
such product will be purchased. In one survey into
the effects of user reviews on consumer behavior:
•
67% of respondents said they check online
reviews either most of the time or every
time before buying products in person or
online;
•
72% said it is highly important for a business to have positive online reviews before
they buy;
•
85% either strongly agreed or somewhat
agreed that they would be less likely to purchase products online that did not have any
reviews; and
19 Ethan Wham, An Economic Case for Section 230, Disruptive
Competition
Project
(Sept.
6,
2019),
https://tinyurl.com/2m6kvam5.
22
•
65% responded with a 7 or above out of 10
when asked how much they trust online reviews.20
Indeed, research shows that “[b]uyers are looking
to their peers to understand which products and services will benefit them, as peers can provide unbiased, individualized information.”21 Neither consumer reviews—nor the purchases they lead to—would be
possible without Section 230 protections, particularly
if platforms were potentially responsible for any reviews they hosted or organized in a manner useful to
other shoppers.
Second, Section 230 allows small businesses to
flourish. After all, a “single small provider may use
multiple large providers to operate their own service
or forum” by, among other things, “maintain[ing] accounts and advertis[ing] across multiple social media
and other services, in addition to relying on ISPs,
domain name registrars, and hosting providers.”22 If
larger providers lacked protections for content posted
by smaller providers, it is unlikely that they would
make their services available to them. Loss of Section
230 protections could thus harm not only digital platforms that have dominant market share, but every20 Ibid. (citing Internet Association, Best Of The Internet:
Consumers Rely On User-Generated Content (2019),
https://tinyurl.com/yjtde9tc).
21 Russell Rothstein, The gig economy and customer reviews:
Trust in the digital era, The Future of Commerce and Customer
Engagement (2016), https://tinyurl.com/yc3pupju.
22 Elizabeth Banker, Understanding Section 230 & the Impact of Litigation on Small Providers 10 (2022),
https://tinyurl.com/2nn66m87.
23
one down to the atomized worker in the gig economy
merely trying to get word out about her services and
to be matched with users most likely to be interested
in such services.
In sum, Section 230’s protections—as predicted
and intended—have led to increased innovation
which has, in turn, led to increases in the reach of
speech and significant benefits for the national economy.
II. Algorithms or other actions that merely suggest information content provided by another content provider do not generally convert
that content into “information provided by”
the recommending user or provider of an interactive computer service.
Amicus agrees with Respondent that the plain text
of Section 230 protects the information organizing activities at issue in this case. Resp. Br. 21–29. Beyond
the textual arguments discussed by others, Amicus
highlights the essential distinction in Section
230(c)(1) between information content provided by
“another” and information content provided by the
challenged provider or user themselves. Insofar as
any organizational activity by the challenged provider or user could be deemed to create its own information content, that content would be far narrower
than the underlying information content provided by
another and would not itself form the basis for a credible claim of liability.
24
A. The plain text of Section 230 draws a distinction between redistribution of “information provided by another information content provider” and information content provided by a user or interactive computer service itself.
The plain text of Section 230(c)(1) bars claims that
seek to hold a provider or user of “interactive computer service[s]” liable “as the publisher or speaker”
of any information content provided by “another” information content provider. This case turns less on
whether Gonzalez’s claim against Google would treat
YouTube as a “publisher” of the videos it suggested
for users who seemed interested in such fare than it
does on whether, by highlighting such videos to particular users, YouTube somehow adopted such information content as its own or itself became the information content provider as opposed to the underlying
videos remaining the information content of “another.”
The statutory definitions of “information content
provider” and “interactive computer service” are helpful in resolving that question. An “information content provider” is “any person or entity that is responsible, in whole or in part, for the creation or development of information provided through the Internet or
any other interactive computer service.” 47 U.S.C.
§ 230(f)(3). The provider of an “interactive computer
service,” by contrast, does not itself necessarily create
or develop information, but instead “provides or enables computer access by multiple users to a computer
server.” Id. (f)(2). Included in the definition of “interactive computer services” are “access software providers.” Ibid. As relevant here, an “access software
25
provider” means a provider of software or tools that
can “pick, choose, analyze, or digest,” “transmit, receive, display, forward, cache, search, subset, organize, reorganize, or translate content.” Id. (f)(4)(B), (C).
Those distributional and organizational activities
thus stand in direct contrast to the “creation” or “development” of the underlying content.
Applying these definitions, it is—of course—true
that the providers of interactive computer services
can create their own information content and thus,
with respect to that content, act as an information
content provider. Batzel v. Smith, 333 F.3d 1018,
1031 (9th Cir. 2003) (“The reference to ‘another information content provider’ * * * distinguishes the
circumstance in which the interactive computer service itself meets the definition of ‘information content
provider’ with respect to the information in question.”); Jones v. Dirty World Ent. Recordings LLC,
755 F.3d 398, 408–409 (6th Cir. 2014) (“[A] website
may be immune from liability for some of the thirdparty content it publishes but be subject to liability
for the content that it is responsible for as a creator
or developer.”). But more often a provider of an interactive computer service organizes and displays the
content of “another” rather than its own content.
Thus, the relevant question for this case is whether
the algorithm that YouTube uses to show videos to
others is an example of Google creating its own information content or whether it is Google merely organizing and displaying the content of “another.”
26
B. Algorithms that point users of interactive
computer services to the content of others generally do not themselves create or
develop content from the recommending
person or entity.
Like every other court to decide this issue, this
Court should recognize that “[m]erely arranging and
displaying others’ content to users of [YouTube]
through * * * algorithms—even if the content is not
actively sought by those users—is not enough to hold
[YouTube] responsible as the ‘developer’ or ‘creator’ of
that content.” Force, 934 F.3d at 70 (cleaned up).
1. This conclusion flows in part from Congress’s
definition of “access software provider.” That definition allows interactive computer services to organize
the information content of “another information content provider.” What they cannot do if they want to
maintain immunity from liability for particular content is to “create” or “develop” that content. Thus,
under the plain language of the statute, unless an interactive computer service engages in the creation or
development of the underlying information content, it
cannot be treated as the information content provider.
2. Google does not create its own content when it
organizes the content of others. As Respondent explains, YouTube’s algorithms “identify the [videos]
that users might find most relevant based on user inputs and other information.” Resp. Br. 12. It then
displays those identified pieces of content to the user.
Pet. App. 38a (“Google matches what it knows about
users based on their historical actions and sends
27
third-party content to users that Google anticipates
they will prefer.”).
As the Force court rightly explained, “[m]erely arranging and displaying” the content of others does
not itself create content. 934 F.3d at 70. “[M]aking information more available is * * * an essential part of
traditional publishing” and does not “amount to ‘developing’ that information within the meaning of Section 230.” Ibid. This conclusion, moreover, holds both
for providers of interactive computer services and
their users alike. After all, Section 230(c)(1) treats
the providers of interactive computer services and
their users the same by expressly providing that neither shall be liable for the information content provided by another.
If Petitioners’ theory is correct, and Google truly
is liable for recommending the content created by its
users via an algorithm or otherwise, then every time
a user of an interactive computer service shares a
video, blog, or tweet created by another, then that
user would become a developer of the underlying content and face potential liability for such content. Indeed, under Petitioners’ approach, by sharing another user’s content, the sharing user becomes the
means by which that content reaches a broader audience. This is no different than what an algorithm
does.
And, as with algorithms, even though retweeting
or sharing the content of another may not be an endorsement of a particular message,23 both are—at the
23 “There are many reasons that someone might retweet a
statement; a retweet is not necessarily an endorsement of the
original tweet, much less an endorsement of the unexpressed be-
28
very least—recommendations that the retweeted or
shared content be viewed. Put differently, if suggesting content via an algorithm somehow falls outside of
Section 230’s gambit by magically transforming one
party’s speech into the speech of the platform, so too
does retweeting it. See Resp. Br. 33.
Petitioners’ theory is wrong and would lead to absurd results. Section 230 protects both providers and
users of interactive computer services from liability
for sharing, recommending, or displaying the speech
of another. Any attempt to split liability regimes between the “providers” and “users” of interactive computer services, or to distinguish the choices made
manually by individual users about what to recommend or highlight to others verses the automated incorporation of the same or comparable choices into an
algorithm, would be completely divorced from the
text of the statute.
3. That is not to say that the provider or user of
an interactive computer service can never take steps
that would make it an information content provider
itself and subject it to liability because of the ways
that it interacts with the content of others. There is,
after all, a difference between a provider or user suggesting the content of others to its users or followers
based on their prior history or some other predictive
judgment about likely interest and a provider or user
actively adopting such content as its own, such as by
lief system of the original tweeter[.]” Flynn v. Cable News Network, Inc., No. 1:21-CV-2587-GHW, 2022 WL 3334716, at *5
(S.D.N.Y. Aug. 12, 2022).
29
endorsing the truth or correctness of a particular
message or statement.
In the former example, all a provider or user does
is acknowledge that another information content provider has created content that may be relevant to another user’s or follower’s interests and organizes that
content in a way that makes it more visible to those
users or followers. When a provider or user does
that—even through an automated algorithm—the
material that reaches a user is materially unaltered
from the material submitted by the user information
content provider in the first instance. E.g., Force, 934
F.3d at 70. As the Ninth Circuit here explained it, an
interactive computer service that actively displays or
organizes the content of an information content provider—which it is expressly allowed to do under the
statute, 47 U.S.C. § 270(f)(4)(B), (C)—does not itself
contribute to the unlawfulness of the content and
thus cannot be held liable for it. Pet. App. 32a–33a.
Put differently, such an interactive computer service
(or recommending user) has not created or developed
that content.
By contrast, in the latter example of affirmative
endorsement, a provider or user actively adopts particular information content as its own and thus may
potentially be held liable for that content. In such circumstances, rather than merely displaying a text or a
video to users based on their prior viewing history,
geography, or other data the interactive computer
service might have on them (or even based on the
provider’s own views of what might be interesting in
general), a provider or user of an interactive computer service would be endorsing and adopting the un-
30
derlying information content itself, perhaps by suggesting that such information is correct.
This Court recognizes an analogous distinction in
its First Amendment cases. This Court has held, for
example, that while the government cannot discriminate on the basis of a person’s viewpoint, it can discriminate on the basis of the broader category of content when it opens up its property for a limited purpose. Rosenberger v. Rector & Visitors of the Univ. of
Va., 515 U.S. 819, 829 (1995). The Rosenberger Court,
for example, expressly recognized that some level of
“content discrimination” may be permissible to “preserve[] the purposes of th[e] limited forum.” Id. at
830. Thus, while the government is sometimes permitted to discriminate against particular subject
matter in limited public fora, that is not necessarily
equivalent to discrimination against or in favor of
particular viewpoints, and hence not expressing the
government’s own substantive view or suppressing
the views it opposes. Ibid. Providers and users of interactive computer services, of course, are not subject
to the same limitations the First Amendment places
on government, but the broader point is that by selecting or highlighting content or even viewpoints,
the provider or user of an interactive computer service is not necessarily endorsing such information
content or adopting it as its own, it is merely organizing the information.
YouTube is not taking a stance when it, having
collected enormous amounts of data on a user’s interests, points that user to content relevant to those interests. For example, if YouTube sends a list of new
cat videos to a user that has watched cat videos in
the past, the separate information content of that or-
31
ganizational effort is no more than: “You seem to like
cats, here is more cat content.”24
But if YouTube were to ever affirmatively adopt
the view contained in one of the videos posted to it, it
would itself be creating content in a way that is categorically different from merely organizing that content on its platform—which Section 230 expressly
distinguishes from the creation or development of
content. 47 U.S.C. § 230(f)(4)(C). It is only when the
provider or user of an interactive computer service
actually creates its own information content that Section 230, properly understood, allows those services
to be held liable for such content. Because YouTube’s
algorithm merely organizes, and does not create, the
content of other users, this Court should affirm the
judgment below.
CONCLUSION
By design, Section 230(c)(1) serves a vital function
that would be lost were this Court to conclude that
the provider or user of an interactive computer service is ineligible for immunity if it uses an algorithm
or similar set of decision factors to organize and display content provided by third parties. To prevent
that result, the Court should affirm.
24 And even were a provider or user to recommend material
created by others based on their own views regarding subject
matter or viewpoint, that would not necessarily constitute an
adoption or endorsement of the underlying information content
as opposed to a more generic statement that such viewpoints are
interesting or potentially valuable or any number of other implied statements short of adoption and endorsement.
32
Respectfully submitted,
ERIK S. JAFFE
Counsel of Record
JOSHUA J. PRINCE
SCHAERR|JAFFE LLP
1717 K Street NW, Suite 900
Washington, DC 20006
(202) 787-1060
ejaffe@schaerr-jaffe.com
MANUEL S. KLAUSNER
LAW OFFICES OF
MANUEL S. KLAUSNER
5538 Red Oak Drive
Los Angeles, CA 90068
(213) 675-1776
mklausner@klausnerlaw.us
Counsel for Amicus Curiae
January 19, 2023
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.