Who Do You Sue?
STATE AND PLATFORM HYBRID POWER OVER ONLINE SPEECH
DAPHNE KELLER 

Aegis Series Paper No. 1902

Introduction
The Liberty County Vindicator, a Texas newspaper, counted down to July 4, 2018, by posting
sections of the Declaration of Independence on Facebook. When it put up the Declaration’s
passage about “merciless Indian Savages,” Facebook promptly took it down. The Vindicator’s
editor—a man of great equanimity—wrote:
While unhappy with Facebook’s action, the editor reminds readers that Facebook is a
business corporation, not the government, and as such it is allowed to restrict use of
its services as long as those restrictions do not violate any laws. Plus, The Vindicator is
using Facebook for free, so the newspaper has little grounds for complaint other than the
silliness of it.
The problem The Vindicator faces is that it has become dependent, perhaps too
dependent, on Facebook to communicate with local residents and to promote the
newspaper. . . . So, the removal of this morning’s post puts the Vindicator in a quandary
about whether to continue with posting the final two parts of the Declaration. . . . Should
Facebook find anything in them offensive, The Vindicator could lose its Facebook page.
This is frustrating, but your editor is a historian, and to enjoy the study of history a person
must love irony. It is a very great irony that the words of Thomas Jefferson should now be
censored in America.1 

The story of Facebook taking down the Declaration of Independence illustrates
several things about internet platforms and users’ speech rights. The first is platforms’
unprecedented technological capacity to regulate individual expression. Facebook and other
large internet companies can monitor every word users share and instantly delete anything
they don’t like. No communications medium in human history has ever worked this way.
The second, related point is our ever-increasing dependence on private platforms. If, as the
Vindicator’s editor feared, Facebook had taken down the paper’s page entirely, readers’ access
to local news would have suffered. The Supreme Court has said that platforms like Facebook
and YouTube serve as “the modern public square,” providing many people’s “principal
sources for knowing current events” and exploring “human thought and knowledge.”2 An
increasing number of observers find this private ownership of the “public square” alarming.

National Security, Technology, and Law

A HOOVER INSTITUTION ESSAY

 2

In particular, they worry that platform bias—or commercial self-interest—is reshaping
political discourse and outcomes.
The third point, and one major focus of this essay, is that speakers in the United States
have few or no legal rights when platforms take down their posts. Some, like conservative
commentator Dennis Prager, have tried suing. Prager claimed that YouTube violated the First
Amendment when it limited users’ access to his videos.3 But claims like these—which I will
call “must-carry” claims, borrowing and broadening a term of art from communications
law—have consistently failed.4 In some cases, courts have said that upholding them would
violate the platforms’ First Amendment rights. As then Judge Brett Kavanaugh put it, the
government may not “regulate the editorial decisions of Facebook and Google”—despite
concerns that “the real threat to free speech today comes from private entities such as
Internet service providers, not from the Government.”5 This reasoning, and the primacy of
platforms’ own speech rights, may well carry the day if must-carry claims ultimately reach
the Supreme Court.
The fourth point, and the other major focus of this essay, is that while platforms appear
to exercise their own discretion when they take down legal speech like the Vindicator’s
Declaration of Independence post, their decisions are often profoundly influenced by
governments. In some cases, governments’ role is very public. For example, Facebook and
other platforms committed, in an agreement with the European Commission, to using their
private Community Guidelines to prohibit “hate speech” as defined by the Commission.
Platforms also developed automated content filters and high-speed removal processes in
order to comply with laws such as Germany’s NetzDG and the EU’s pending Terrorist
Content Regulation—despite the serious risk that these automated tools will take down the
wrong things.6 Governments’ influence on platforms can also be more subtle. As Facebook’s
global head of policy has described, companies are “eager to predict regulation . . . so they
can adjust their policies to keep up with the times and thereby avoid risk to their business.”7
Platforms’ anticipatory obedience spares governments the need to enact actual laws—and
deprives affected users of the opportunity to challenge them in court.
This essay is an effort to frame the big picture of online speech regulation at the
intersection of state and private power. It aims to bring more coherence to the often
overheated public conversation about platforms and online speech rights. In it, I draw on
my experiences as a long-term practitioner of platform law, including as a former associate
general counsel for Google. The essay has two major parts. Both examine how the legal
rights of speakers or publishers like the Vindicator are affected when platforms take down
their speech. In the first, I discuss claims against governments for their role in driving
platforms’ decisions. In the second, I discuss claims against platforms themselves.8 I will
suggest that the two issues are related in subtle but important ways. On the one hand,
governments can bypass constitutional limits by deputizing private platforms as censors.9
On the other, platforms can take on and displace traditional state functions, operating

Daphne Keller  •  Who Do You Sue? 

 3

the modern equivalent of the public square or the post office, without assuming state
responsibilities. Focusing on either issue in isolation may leave us ill-equipped to protect
online speech rights and the open internet.

Limits on State Power
Laws and other exercises of state power often prompt platforms to take down their
users’ speech, even if that speech is protected by the US Constitution or international
human-rights law. Even governments with no intention of promoting censorship can
unintentionally “launder” their power, using state authority in ways that lead platforms to
remove lawful expression. This section will discuss how that happens and what recourse
might be available to affected internet users. Realistically, those users are unlikely to
be protected if courts do not recognize the role of state action and instead hold only
platforms—with their lesser or nonexistent duties to protect speech—responsible for content
takedowns.

Laundering State Action through Private Platforms
Governments influence platforms’ content removal decisions in a number of ways. Below, I
will describe three broad mechanisms: direct regulation; indirect pressure, or “jawboning”;
and cross-border influence.

Regulation  The state’s role in platform content removal decisions is clearest in the case
of intermediary liability laws, such as the Digital Millennium Copyright Act (DMCA) in
the United States and the eCommerce Directive in Europe, which hold platforms legally
responsible for content posted by their users. These laws are intended only to enforce
existing speech laws, by having platforms take down any genuinely unlawful material they
find or are notified about. But they also incentivize platforms to take down speech that,
while controversial or offensive, does not violate the law. Erring on the side of removing
controversial speech can spare platforms legal risk and the operational expense of paying
lawyers to assess content. Unscrupulous accusers take advantage of this, targeting legal
speech with spurious allegations. This problem, and the resulting removal of lawful speech,
is well documented.10 Some platforms simply remove anything that an accuser claims is
illegal.11 Others attempt to weed out invalid claims, but nonetheless comply with far too
many.12 Human-rights literature and widely endorsed best-practices guidelines suggest that
the best corrections for over-removal come from robust platform takedown procedures—
giving accused users notice and an opportunity to defend their speech, for example, or
penalizing those who make bad-faith accusations. Laws that lack such protections and
foreseeably lead platforms to silence lawful speech may violate internet users’ rights to free
expression.13
American lawyers typically don’t think of intermediary liability rules as state speech
regulation. Cases and scholarship linking platform liability to the First Amendment have

Hoover Institution  •  Stanford University

 4

been rare in the past two decades, because the field has largely been occupied by two federal
statutes designed to protect online speech. One, Communications Decency Act Section 230
(CDA 230), immunizes platforms from most claims based on user content. It sets policy
strongly in favor of protecting speech—sometimes at the cost of tolerating harmful or illegal
material online. The other key US law, the DMCA, includes a procedurally rigorous noticeand-takedown system, intended to reduce the risk that platforms will remove legal speech.
If these statutes did not exist, the First Amendment would still limit platforms’ liability
for users’ speech. Experts disagree, though, about exactly what the limits would be.14 Preinternet Supreme Court cases about content distributors said, for example, that holding
booksellers liable for book contents they didn’t know about would violate the First
Amendment. These rulings turned not on the booksellers’ own speech rights, but on those
of authors and readers who relied on them.15 Strict liability, the Court noted, would lead
booksellers “to restrict the public’s access to forms of the printed word which the State
could not constitutionally suppress directly.”16 This “censorship affecting the whole public”
would be “hardly less virulent for being privately administered.”17 In early internet cases,
lower courts drew on this precedent to strike down or limit laws that effectively encouraged
intermediaries to remove online speakers’ lawful expression.18 Courts outside the United
States have done the same based on their own legal systems’ free-expression guarantees.19
Other cases, though, suggest that internet users have very limited rights to challenge
intermediary liability laws or the removal decisions that platforms make in response
to them. An example from the European Union illustrates the problem—as well as
its connection to platforms’ own discretion to take down content and the must-carry
arguments examined elsewhere in this essay.
In 2016, Spain’s data protection regulator fined Google €150,000 for telling an online
publisher that its web page had been removed from search results in response to a “right to
be forgotten” claim.20 The regulator reasoned that the publisher had no cognizable interest
in being informed, and in potentially objecting, when its page disappeared from search
results. This was because “search engines do not recognize a legal right of publishers to have
their contents indexed and displayed, or displayed in a particular order.”21 In other words,
because the publisher had no legal right to challenge Google’s private decisions about
removal and ranking, it also couldn’t object when the law required Google to remove search
results. Making Google the enforcer of the right to be forgotten effectively obscured the
state’s role—which included mandating removal in the first place and penalizing platforms
for noncompliance.
The Spanish right-to-be-forgotten example only makes explicit a risk lurking in any
intermediary liability system. Platforms will inevitably take down legal speech by mistake,
or because it is the cheapest, easiest, and safest choice. European regulators are not alone in
dismissing online speakers’ legal interest in preventing this from happening. A US court,

Daphne Keller  •  Who Do You Sue? 

 5

similarly, found that a plaintiff whose lawful ads were terminated from Craigslist based on
a new intermediary liability law had no standing to challenge the law’s constitutionality.22
If states bear no responsibility for tailoring laws to minimize harm to legal speech, or
if platforms’ own discretion to remove content effectively extinguishes users’ rights to
challenge government removal mandates, then internet users will be left with few legal
tools to protect their expression rights.

Pressure  Governments also use other kinds of influence, like the threat of future
regulation, to encourage platforms to take down speech. Scholars call this tactic
“jawboning,” in reference to the biblical story of Samson.23 Jawboning by politicians and
regulators, paired with “voluntary” self-regulation by platforms, has effectively transformed
the landscape for online speech over the past decade.
Jawboning in its subtler forms makes state action much harder to trace. Even when the
state’s role is conspicuous, though, it is hard to identify the proper legal or ethical limits,
because jawboning includes very ordinary government activity. The congressman who tells
a CEO that she “had better do something or we’re going to pass a law and you won’t like it”
is following a time-honored tradition. So is the CEO who then adjusts her business model.
The company staves off burdensome regulation, the congressman claims victory, and both
sides have a deal they can live with.
The problem for politicians who demand that platforms remove more content, though, is
that the resulting bargains affect users’ speech rights. The CEO may have the authority to
remove those people’s legal speech from her platform, but Congress does not. By relying on
informal or tacit agreements, legislators and other state actors can cause platforms to adopt
speech rules that would, if written into law, be struck down by courts on free-expression
grounds.
In the United States, the most heavy-handed forms of jawboning can violate the First
Amendment. Judge Richard Posner, for example, ruled that Cook County, Illinois, sheriff
Thomas Dart acted unconstitutionally in his campaign against the notorious classified
advertising site Backpage. Dart, who had already been told by one court that he lacked legal
authority to shut the site down, instead pressured credit-card-payment processors to cut off
its cash flow.24 The court called this “official coercion” and held that it was unconstitutional.
The site’s victory against the sheriff was somewhat hollow, though, since the damage—
losing payment-processing services—was already done.
American government officials are no strangers to jawboning, as Backpage.com v. Dart
illustrates. Payment processors and other online service providers cut off Wikileaks,
for example, following pressure from Obama administration officials and members
of Congress.25 The FBI has pressured intermediaries to take down videos that, in one
legal expert’s words, were “constitutionally protected but were not viewed by the FBI as

Hoover Institution  •  Stanford University

 6

consonant with the public interest.”26 Pennsylvania police even reached across national
borders, persuading a Canadian ISP to take down a website critical of local government.27
Executive-branch pressure is particularly conspicuous in the area of intellectual property,
where Annemarie Bridy has documented the federal “IP Czar’s” role in fostering “coerced
self-regulation” agreements. As Bridy notes, “The very fact that the government convenes
and participates in negotiations over such agreements calls their voluntariness into
question.”28 US government pressure in other areas, such as terrorism, has been less
conspicuous. But American lawmakers may see less need to push platforms in these
areas, because European governments have been so effective in convincing platforms to
“voluntarily” ban speech.
One European jawboning example is the Hate Speech Code of Conduct, an agreement
between the European Commission and four major platforms—YouTube, Twitter, Facebook,
and Microsoft.29 The code requires platforms to put in place “Rules or Community
Guidelines” that prohibit “the promotion of incitement to violence and hateful conduct.”
These rules theoretically track existing European law, which would require platforms to take
down hate speech upon notification even without this separate commitment. But the code
casts platforms’ removal decisions as the product not of law or state power but of private
agreements: platforms’ contracts with users under their Terms of Service, and platforms’
agreements with the Commission under the code itself. Since platforms are in theory
enforcing only their own rules when they take down hate speech, users have no clear
means to dispute legal interpretations or raise defenses based on European free-expression
guarantees. By putting interpretation of a uniform standard in platforms’ hands, the code
also homogenizes and “federalizes” European law, despite substantial variation in member
states’ actual laws and the states’ supposed authority to maintain their own rules for free
expression.30
Platforms have reached a different set of semiformal accommodations with governments
regarding terrorist content, responding to years of escalating pressure around the world.
EU governments in particular have long threatened legislation that would require
platforms to use technical filters to find terrorist content and remove it within an hour
or two.31 The European Commission followed through on these threats with both
a formal “Recommendation” and draft legislation in 2018.32 In the years preceding,
however, platforms agreed to a series of concessions, widely interpreted as bids to ward
off regulation. Most consequentially, the same four platforms that negotiated the code
with the Commission agreed to develop and widely share new content-filtering tools,
capable of detecting images and videos using a privately held database of content hashes,
or “fingerprints.” Little is publicly known about the database, but similar content-detection
tools are widely believed to be responsible for errors like YouTube’s deletion of more than
100,000 human-rights videos.33 As of January 2018, twelve platforms were using the
database to police users’ posts, and it was said to contain some 40,000 hashes.34

Daphne Keller  •  Who Do You Sue? 

 7

As with the hate-speech code, the terrorism hash database is nominally used to enforce
platforms’ private rules, not state-imposed law. Companies add videos or images to the
database that violate their own Community Guidelines. Each platform using the database
can then find matching material and remove it—or, in theory, not remove it—based on its
own, separate rules. Governments don’t maintain the tools or, for now, compel platforms
to use them. They do, however, pressure smaller platforms to adopt them.35 The EU’s 2018
draft EU Terrorist Content Regulation, which is expressly designed to reach smaller and
non-EU-based companies, would make the hash database or similar tools mandatory,
and require ongoing reporting and coordination with national authorities. Adding to
this snarl of state and private power, the Regulation would cement an EU-wide role for
law enforcement agencies’ Internet Referral Units (IRUs), which review and seek removal
of terrorist content—not for breaking the law, but for violating platforms’ Community
Guidelines.36
Internet users whose legal speech falls victim to one of these removal systems—like the
Syrian Archive, a human-rights organization that lost key videos intended for use in
prosecuting human-rights abusers—have no clear legal remedy in most countries. At best,
they can ask platforms to reconsider. The lack of further legal recourse is troubling given the
government’s extensive role in shaping platforms’ “private” takedown decisions.

Cross-Border Influence  States that regulate or influence platforms often also, intentionally
or not, shape speech rules that the platforms apply in other countries. Platforms have both
practical and political reasons to make speech restrictions global. The legal doctrines in
this area—including jurisdiction, choice of law, and comity—are complex. But the upshot
is simple. Platform operators are free to silence users in the United States based on foreign
governments’ laws. American law does not, and perhaps constitutionally could not, restrict
this choice. This freedom—the fact that US law does not require platforms to transmit
speech that is legal here—encourages foreign courts to order global deletions in the first
place.
States’ influence on expression and information outside their borders is nothing new.
Voice of America radio, for example, long broadcast prohibited content into Soviet bloc
countries. But the internet provides an unprecedented vehicle for extraterritorial influence
over speech.37 For many years, US-based platforms effectively exported American speech
laws and norms, whether deliberately or as an accidental consequence of being based in the
United States and accessible elsewhere. But this same global reach makes platforms equally
effective as vectors for speech restriction. Any government with sufficient leverage—which
could mean any government that controls an important enough market—can potentially
shape speech rules for platform users everywhere.38

The Mechanics of Cross-Border Speech Restraints  Large platforms like Google or Facebook
typically remove content that violates national law from nationally targeted versions of

Hoover Institution  •  Stanford University

 8

their services.39 A post that violates German laws, for example, can be taken down from just
the German version of Facebook—though it will be taken down globally if it also violates
the platform’s Community Guidelines. In some cases, like one that is pending between
Facebook and Austria before the EU’s highest court, platforms fight hard to keep national
law enforcement within national borders.40 But global compliance is easier and causes less
legal or political grief.
States can directly shape speech access worldwide by ordering platforms to comply globally
with their laws. The Canadian Supreme Court recently did so in a trade-secret case, ordering
Google to remove search results around the world.41 US courts would likely do the same for
American copyright laws.42 China, which historically has largely ignored material outside its
borders or simply blocked users from seeing it, also now asserts authority to restrict speech
in the rest of the world.43 In Europe, France’s data-protection regulator maintains that
Google must apply French right-to-be-forgotten laws everywhere. That claim, along with
Facebook’s Austrian case, is pending at the EU Court of Justice.44
States also indirectly influence platforms’ global speech rules, even when they only intend
to enforce their laws domestically. The reasons are practical. One has to do with platforms’
operational preference for a single set of rules. Teams that review massive volumes of user
content struggle with logistics and enforcement consistency in the best of circumstances.
Enforcing dozens of different rules around the world would, as Facebook’s Monika Bickert
has pointed out, be “incalculably more difficult” than applying a single, consistent set
of Community Guidelines.45 For social networks and other communications platforms,
inconsistent rules also create bad user experiences, interfering with communication
between people in different countries. Maintaining a single set of standards—and perhaps
expanding them to accommodate national legal pressure as needed—is much easier.
Another reason that state influence spreads is technical. Building and maintaining
differently engineered systems around the world is difficult, expensive, and inefficient.
When one country requires a technical change, like the content filters the EU seeks to
mandate, platform engineers may prefer to simply make the same change everywhere.
Political and legal calculations can shape that decision too. Once a powerful enforcement
tool like a content filter exists, demands for new uses in other countries are inevitable.
Companies that don’t want to take on expensive and politically bruising fights may prefer
to do what Mark Zuckerberg did in his House and Senate testimony: commit to using the
technology voluntarily.46

The Law of Cross-Border Speech Restraints  Neither the US First Amendment nor laws of
jurisdiction, choice of law, or comity prevent platforms from enforcing foreign speech
restrictions in the United States. The reasons relate to the overall theme of this essay:
laws generally protect speakers from state censorship but allow private platforms to take
down speech as they see fit. If a country with speech-restrictive rules compels a platform

Daphne Keller  •  Who Do You Sue? 

 9

to take content down globally, the platform can comply without violating laws in speechpermissive countries. The result is a one-way flow, with speech rules from more restrictive
countries being enforced in less restrictive ones.
US law puts no meaningful restraint on platforms’ enforcement of foreign takedown
mandates, as illustrated by two of the US “must-carry” cases I will discuss later in this
essay. Plaintiffs in both Zhang v. Baidu and Sikhs for Justice v. Facebook were human-rights
advocates who said platforms had silenced them at the behest of foreign powers—China and
India, respectively.47 In both cases, the courts held the platforms had every right to do so.
The Baidu court said that forcing the defendant, China’s leading search engine, to reinstate
content would violate its First Amendment rights. In other words, US free-expression law
protecting platforms’ speech rights ensured that China could restrict platform users from
speaking and accessing information here.
Early internet cases like Yahoo’s famous dispute with France are often misunderstood as
protecting speakers’ rights in such situations, preventing platforms from taking down US
speech based on foreign laws.48 But that case, too, only addressed constraints on state power.
The issue was whether a US court would enforce France’s order for Yahoo to stop showing
Nazi memorabilia—not whether Yahoo would or could comply voluntarily. In fact, Yahoo
chose to comply globally with the French law, even while litigating the case in US courts.49
The case’s questions about what orders US courts will enforce, while interesting to lawyers,
may be beside the point for global platforms. A company that fears having its foreign assets
seized or employees arrested, or that doesn’t want to lose access to lucrative foreign markets,
may see good reasons to follow foreign court orders and to do so globally if that’s what the
foreign court requires.50
Platforms’ discretion to choose what content they remove matters to foreign courts
considering whether to issue global removal orders too. Under traditional principles of
international law, courts may avoid issuing orders that, in the Canadian Supreme Court’s
words, “require [the defendant] to violate the laws of another jurisdiction.”51 The Canadian
Court, in the case ordering Google to remove search results globally, concluded that the
order was reasonable because no other country’s law compelled Google to leave the results
up. On remand, a lower Canadian court explained that such a conflict might exist, and the
Canadian order might be modified, if a US court issued an order “requiring Google to link”
to the disputed websites. But softer conflicts based on Google’s mere discretionary right
to include the links in search results under US law—or American users’ rights to see the
web pages, or the Canadian order’s lack of enforceability in US courts—were not enough.
Because US law did not require Google to show the search results, the Canadian court had
no reason to stay its hand.52
The same analysis could easily carry the day in the pending EU cases and lead the Court of
Justice to order Facebook and Google to apply European speech laws in the United States

Hoover Institution  •  Stanford University

 10

and around the world. It would also support Turkish courts ordering global enforcement
of laws against insulting President Erdogan, Russian courts enforcing laws against “gay
propaganda,” or Saudi courts enforcing blasphemy laws. When platforms enter markets in
speech-restrictive countries and subject themselves to those countries’ laws, there is little
in US black-letter or constitutional law to prevent the subsequent export of those countries’
speech rules.

Protecting Online Speech from State Power
Users silenced by indirect or laundered state action have relatively underdeveloped legal
options. They can use the First Amendment (or its international analogs) to challenge actual
legislation, or perhaps even to challenge some forms of non-legislative jawboning. But users
can only assert rights against the state when courts recognize state action at work. If any
intervening private action extinguishes speakers’ right to object—as in the Spanish rightto-be-forgotten case—states will not be held accountable. For this reason, the first job of
free-expression advocates in this area may be simply to call attention to the role states play
in platform-content removal. They can also develop constitutional or human rights–based
arguments and seek court judgments to better protect users’ rights.
Another option is to seek legislation. Laws could require greater transparency about state
influence on platforms, for example. At a minimum, that might include disclosing what
content platforms remove at the request of police IRUs. Lawmakers could in principle also
pass statutes to limit jawboning. As Derek Bambauer points out, though, “The demands of
the modern administrative state make regulators wary of limiting informal enforcement,”
and lawmakers’ self-interest may generally lie in favor of jawboning, in any case.53
In some ways, the problems of cross-border enforcement seem more tractable than those
of purely domestic law. All but the most repressive states have something to lose from
global enforcement of other countries’ speech restrictions. Few European lawmakers, for
example, want Facebook or Google to apply Russian antigay laws or Saudi blasphemy
laws to their citizens. Black-letter law changes to prevent or limit issuance of cross-border
speech suppression orders are not hard to imagine. Jurisdiction scholars have suggested
improvements such as a formal “scope of remedies” assessment for courts considering
extraterritorial enforcement, for example.54 Courts making such assessments could examine
not only whether the defendant has the right to silence speakers in other countries but also
whether governments do.55
The problem is that such judicial restraint would work only if it were mutual, with
democratic countries agreeing not to interfere with each other’s locally legal online
information. A new multilateral agreement of this sort seems unlikely anytime soon.
Instead, we can expect national courts to look uneasily to one another’s examples. With

Daphne Keller  •  Who Do You Sue? 

 11

Canada throwing its judicial prestige behind extraterritorial enforcement, and Europe likely
to follow, the opportunity for mutually agreed restraint may soon pass.
Unilateral fixes, or national laws limiting the reach of foreign orders, are harder. As the US
must-carry cases illustrate, lawmakers may simply have no power, under the Constitution,
to prohibit platforms from enforcing any speech rules they want. Platforms’ free hand in
removing user speech makes it hard to check even domestic exercises of state power, and
nearly impossible to check international ones.

Limits on Platform Power
For individuals prevented from participating in important online forums, the visible
cause is usually the platform operator. If users get mad enough, that’s who they sue. This
section discusses what I call “must-carry” claims, which would compel platforms to keep
users’ content online in order to protect their speech rights. It will not consider other legal
theories, such as discrimination or promissory estoppel, that would not affect platforms’
overall ability to set speech rules.
Below, I will briefly review must-carry cases to date, including important new cases outside
the United States. I will then consider what outcomes might look like if must-carry claims
did succeed. In the longest subsection, I will examine relevant US law more closely. Finally,
I will discuss models of hybrid, “partial must-carry” regimes that might provide some
protection for users’ speech without stripping all control from platforms. I conclude that
these models all have very serious problems—but also that they suggest lines of inquiry that
can better advance our thinking about platforms and online speech.

The Story So Far
The intuition that major platforms must, somehow, owe it to users to host or transmit
lawful speech is widely held. Indeed, both of the major US intermediary liability statutes
presuppose the existence of such claims and immunize platforms against them.56 In the
United States, though, no claim to date has succeeded. Outside the United States, to the best
of my knowledge (and I’ve tracked the topic for years), none did before 2018, when several
important cases emerged.

In the United States, Cases against Internet Platforms Have Failed  To date, more than two
dozen US plaintiffs have sued platforms under must-carry theories.57 The cases have gained
increasing political attention and brought ever greater legal firepower in recent years. But
plaintiffs have been arguing at least since the 1990s that platforms act as public forums
and thus can be compelled to carry users’ speech.58 So far, platforms have won every case,
defeating a long list of legal claims. In addition to arguing that platforms have violated their
First Amendment rights,59 plaintiffs have unsuccessfully sued platforms for constitutional

Hoover Institution  •  Stanford University

 12

due process60 and equal protection61 violations, breach of contract,62 unfair competition,63
interference with prospective economic advantage,64 and a kitchen sink of other claims.
Platforms’ defenses have been almost as varied. Sometimes platforms win simply because
plaintiffs can’t make out the merits of their claims.65 Other times they win based on their
Terms of Service, which typically prohibit content that violates Community Guidelines, as
well as reserving rights to terminate service for any reason.66 Platforms also sometimes win
because of Section 230 of the Communications Decency Act (CDA), which immunizes them
both from claims treating them as publishers and from claims arising from their good-faith
efforts to take down “objectionable” speech.67
Finally, platforms defeat must-carry claims by asserting their own constitutional rights.68
They argue that laws requiring them to host content against their will would force them
to speak, in violation of the First Amendment. This constitutional argument is particularly
important. If the current pro-platform-regulation zeitgeist leads to other changes in the law,
it will very likely be litigated to the Supreme Court.

Outside the United States, Some Must-Carry Claims Have Succeeded  Until very recently,
there were no reported rulings on must-carry claims against platforms outside the United
States. A few claims were filed and received press attention—most notably, a 2016 claim by
a French man who said Facebook had violated his rights by taking down the anatomically
detailed Gustave Courbet painting L’Origine du monde.69 But none resolved the core question
of whether a user can compel a platform to host his or her speech.
Things changed in at least two countries, Germany and Brazil, in 2018. In Germany, a
series of users sued Facebook for taking down their posts under its Community Guidelines.
The results were mixed, but at least a handful of judges had ordered Facebook to reinstate
material as of late 2018.70 The cases draw on the German Drittwirkung doctrine, which uses
fundamental rights as a strong guide in interpreting obligations between private parties. In
Brazil, at least two appellate courts have ordered YouTube to reinstate parody videos that
had been removed on copyright grounds, and to pay damages to the parody creator.71 The
courts in those cases relied in part on Brazil’s intermediary liability legislation, which is
expressly framed as a free-expression protection for internet users and in most cases requires
platforms to take down user content only if a court determines that it is illegal.72 Although
the legislation does not prohibit platforms from removing content prior to a court decision,
these cases held that YouTube violated speakers’ rights by doing so.
These rulings are, so far, unusual. But the fact that they were issued at all reflects important
divergences in legal culture and understanding of free-expression rights. One difference has
to do with the “horizontal” application of free expression and other human-rights laws—
meaning that plaintiffs can assert rights in some cases against private defendants, as well
as against the government.73 This approach may have particular traction in Latin America.

Daphne Keller  •  Who Do You Sue? 

 13

Sophisticated practitioners in countries including Brazil and Argentina have long told
me that they expect plaintiffs to prevail in at least some must-carry cases there. Another
difference may stem from broader conceptions of the state’s positive obligation to support
free expression. Both scholars and human-rights officials in Europe have argued that states
may violate citizens’ rights by encouraging, or perhaps even by tolerating, platform-contentremoval policies that harm free expression.74 Finally, courts outside the United States are
generally less receptive to the idea that platforms’ own free-speech rights should be a barrier
to claims from their customers.

What Happens If Must-Carry Claims Succeed?
It is not always clear exactly what must-carry claimants are asking for, or what the internet
would look like if they got it. One question is exactly what speech platforms would have to
carry. Another is which specific platform functions would change.

What Speech Would Be Affected?  Some must-carry proponents seemingly aim to hold
platforms to the same rules as the government, or want to convert them to common
carriers—bound to deliver any message at all, or at least any that isn’t illegal.75 Such a
standard might leave platforms free to apply content-neutral “time, place, and manner”
restrictions, as the government may in places like public parks and streets. But it would
also require platforms to preserve speech that many people find obnoxious, immoral, or
dangerous. Real-world examples of legal online speech that have attracted widespread
outrage include “history of why jews ruin the world” and “how to burn jews.”76 Examples
of speech protected under recent First Amendment case law include signs held by picketers
near a soldier’s funeral saying “Thank God for IEDs” and “You’re Going to Hell.”77 There is a
reason public-interest groups and internet users typically urge platforms to take down more
legal-but-offensive speech—not less.
Requiring platforms to carry speech that most users don’t want to see would also have
serious economic consequences. Among other things, platforms would lose revenue from
advertisers who do not want their brands associated with hateful or offensive content.78
Converting platforms from their current, curated state to free-for-alls for any speech not
banned under law would be seen by some as tantamount to nationalization. Platforms
would almost certainly challenge it as an unconstitutional taking of property.
Even for committed free-expression advocates, it is not clear that requiring platforms to
preserve all legal speech is in the public interest. There are speech rights on all sides of the
issue. For one thing, platforms have their own First Amendment rights to include or exclude
content. For another, platforms sometimes silence one aggressive user—or many—in order
to help another user speak. Without the platform’s thumb on the scales, some speakers, like
female journalists barraged with not-quite-illegal threats of rape and violence, might be
driven offline entirely.

Hoover Institution  •  Stanford University

 14

If making platforms carry all legal expression seems too extreme, must-carry proponents
might argue that platforms should be held to some other standard. For example, they might
be permitted to exclude only highly offensive speech, or could be required to apply rules
fairly. I will explore some hypothetical regimes of this sort in the final section of this essay
and argue that they all have real problems. Perhaps most troublingly, legally prescribed
“decency” or “fairness” standards would take platforms out of the job of deciding what
currently legal speech users can share, only to have the government do it instead. The
degree of regulatory intervention required to do so at internet scale would dwarf anything
from the heyday of US broadcast regulation. And, unlike rules for mass media, these rules
would govern speech by ordinary individuals. Letting platforms prohibit some legal speech
would also fail to address the core concern that dissident or unpopular voices are being shut
out of the most important forums for discussion today.

What Platform Operations Would Be Affected?  Today’s major internet platforms typically
offer not one but many products and features. Facebook’s homepage, for example, has
included manually curated news headlines and an algorithmically sorted feed of friends’
posts, as well as advertisements, event invitations, and notices about private messages. To
get a handle on must-carry claims, we need to know which of these things claimants want
to change.
Even within specific, high-profile product features—like Google’s web search results or
Twitter’s news feed—platforms use a variety of mechanisms in response to disfavored
content. To understand must-carry claims, we need to know which of these would be
affected. Broadly, these mechanisms fall into two categories. First, platforms can exclude
content, applying their content removal policies. Second, they can increase or decrease its
visibility through their content ranking systems.79
Content removal, in which platforms erase or block access to material that violates their
Community Guidelines, is the kind of curation most familiar in must-carry discussions.
Platforms’ content-removal policies often begin as ad hoc responses to particular cases.
But they can become culturally defining over the years, in part by driving recruitment or
departure of key employees or board members.
Content ranking is the curation that platforms do on an ongoing basis in order to provide
their basic user experience. Ranking algorithms, maintained and updated by teams of
engineers, determine what material is considered most relevant or highest priority. For
search engines, this ranking is the entire value proposition. But many social networks and
other hosts also work hard to determine content order or layout, and strongly defend their
right to do so. Ranking decisions may in practice be nearly as consequential as removal
decisions, since few users will ever find information that is buried at the bottom of a
news feed or search results. Public discussion and advocacy around ranking have become
common in recent years.

Daphne Keller  •  Who Do You Sue? 

 15

Advocates for must-carry regimes often focus solely on removal rather than ranking. For
the sake of simplicity, I follow suit and use the term “removal” in much of this essay. But
the distinction matters for some moral or legal questions. For example, some argue that
platforms bear more responsibility for amplifying content than they do for merely hosting
it—and as a result, that they should be more willing to down-rank content than to remove
it completely.80 The distinction could also matter for legal disputes about platforms’ First
Amendment rights (as discussed in the next section) as well as for the complexity and
feasibility of potential regulations (in the final section).

The Law
The core must-carry contention is that private platforms have a legal duty to carry users’
speech. There is a reason why courts keep rejecting this argument. Considerable Supreme
Court precedent stands against it. Only a few cases in history have allowed plaintiffs to
assert First Amendment or other speech rights against private entities. As discussed below,
the doctrines created in those cases don’t work well for plaintiffs suing internet platforms.
If they could find a claim that worked, or if Congress created one, plaintiffs would still
encounter another formidable barrier in platforms’ own First Amendment rights.

Arguments That Platforms Are Acting Like States  Must-carry proponents sometimes argue
that platforms are effectively standing in the shoes of the state, and thus must assume the
state’s duties toward citizens. This theory, which builds on a line of Supreme Court cases
starting with the 1946 Marsh v. Alabama, is unlikely to succeed.81 In Marsh, the Court
upheld a speaker’s right, under the First Amendment, to distribute religious literature within
the defendant’s company-owned town.82 Over succeeding decades, the Court narrowed
this “company town” doctrine considerably. The rule now, as characterized by a lower
court in one must-carry case, only supports claims against “a private entity that owns all
the property and controls all the municipal functions of an entire town.”83 Because this
does not describe platforms like YouTube or Facebook, must-carry plaintiffs are unlikely to
succeed using “company town” arguments.
A more relevant offshoot of Marsh concerns shopping malls—private spaces that do not
serve a full range of municipal functions but that may nonetheless be forums for speech.
Plaintiffs’ best support comes from a federal Supreme Court case, PruneYard v. Robins,
that upheld the California Supreme Court’s application of the state’s constitution.84 The
Court affirmed plaintiffs’ rights under the California constitution to enter a Silicon Valley
shopping mall to distribute leaflets. Plaintiffs suing today’s platforms argue that they fulfill
the public-forum function at least as much as shopping malls ever did, and in consequence
must tolerate unwanted speech.
A claim that succeeded based on PruneYard would be a remarkable development for a
number of reasons. One is that it would create state-to-state variance in internet speech

Hoover Institution  •  Stanford University

 16

rules. But in part for this very reason, a California-only must-carry rule would probably be
preempted by federal legislation. CDA 230, which was enacted in part to solve the problem
of varying state laws and includes immunities from must-carry claims.85
A third line of cases, most famously including one about President Trump’s Twitter
account, has established limited public-forum status for government-held accounts on
private platforms.86 These cases are ongoing as of this essay’s publication. They sound
similar to the must-carry claims against private platforms discussed here and, as a result,
will probably encourage more of them. Legally, though, the Trump case is very different,
because plaintiffs are only asserting First Amendment rights against the usual defendant:
the government. The district court ruling in the Trump case turned on the conclusion that
the president and an aide functioned as state actors in operating the @RealDonaldTrump
Twitter account.87 Nothing in its reasoning indicates that Twitter itself is a state actor or is
subject to the First Amendment.

Arguments That Platforms Are Essential Channels for Communication  Another major mustcarry argument is that platforms with high market share should have special obligations
as de facto gatekeepers of public discourse. A number of legal doctrines—some focused
on speech and ideas, others on economic competition—speak to this issue. Many are old,
and in my opinion none create cognizable legal claims today. But these laws are more
malleable than the constitutional law in the company town and shopping mall cases. The
current push for platform regulation could drive legislative change, or spur courts to revive
doctrines to support plaintiffs’ claims.
One widely discussed idea holds that major platforms are like utilities: essential,
unavoidable, and monopolistic services to which customers should be guaranteed access.
This argument, while not actionable by plaintiffs in private litigation, could in principle
be embraced by Congress in new legislation. It is bolstered by the idea that platforms,
particularly social networks, may be natural monopolies, unavoidably tending toward
concentration and best managed as a single operation. The “platform as utility” model has
gained traction across the political spectrum, including from sources as far to the political
right as Steve Bannon.88
A related approach comes from the essential-facilities doctrine in antitrust, which in the
past led to things like compulsory access to the only railroad line going into a town.89
The doctrine’s vitality today is debated, as is its wisdom. Like other historical approaches
to competition law, though, it could see revival as part of the current wave of platform
regulation proposals.
Similar claims could also be raised based on traditional common-carriage law, which
required entities such as railroads or shipping companies to accept all lawful packages and
not modify their contents. At least one of the recent must-carry cases unsuccessfully raised

Daphne Keller  •  Who Do You Sue? 

 17

this argument, based on California’s codification of common law.90 Like the California
constitutional claim discussed above, though, state common-carriage arguments are likely
to be preempted by federal legislation.
Traditional sources of law about control over channels for speech—as opposed to
channels for trade and competition—are less common. Laws expressly designed to limit
communications channel owners’ power over the content they carry emerged in response to
twentieth-century technologies. They were eventually codified in the dense body of federal
law administered by the Federal Communications Commission (FCC).91 Over time, these
included rules like the fairness doctrine for broadcasters and net-neutrality obligations for
internet access providers.
This body of law does not currently support must-carry claims against user-facing platforms
like Facebook or YouTube, because Congress emphatically declined to extend it to them
in the 1996 Telecommunications Act. The CDA, which passed as part of that act, declared
Congress’s intention to “preserve the vibrant and competitive free market” for internet
services “unfettered by Federal or State regulation,” and immunized platforms from most
obligations to either carry or remove content. But Congress could, as Republican members
have recently threatened, try to change the law to create carriage obligations. If they did,
the massive body of past and current federal communications law would be highly relevant.
For one thing, these laws provide the dominant and familiar model for US regulation of
speech and communication intermediaries. Any serious proposal to legislate must-carry
obligations would draw on this history. For another, and importantly for plaintiffs in today’s
cases, these laws have been heavily litigated and are still being litigated today. They provide
important precedent for weighing the speech rights of individual users against those of
platforms.

Platforms’ First Amendment Defenses  If must-carry proponents ever convinced courts
to recognize their claims or got Congress to create new statutory rights, their work would
not be done. The next and very major hurdle would be to overcome platforms’ objections
based on their own First Amendment rights. This set of arguments, made most prominently
by Eugene Volokh and Donald Falk in work commissioned by Google, says that platforms
express editorial judgment through their ranking and removal choices.92 Laws overriding
these choices and forcing platforms to feature different content would compel them to say
something else instead, in violation of the First Amendment.
This is a strong argument. As discussed above, it has carried the day for platforms
in a number of recent cases. The Supreme Court has recognized analogous rights for
other entities resisting must-carry obligations, ranging from cable companies to parade
organizers.93 In Miami Herald v. Tornillo, for example, it overturned a law requiring
newspapers to print responses to articles.94 Similarly, in Hurley v. Irish-American Gay, Lesbian
and Bisexual Group of Boston, it upheld a parade organizer’s right to exclude a gay-rights

Hoover Institution  •  Stanford University

 18

organization, calling it a protected editorial choice.95 But these rights do not provide an
absolute trump card. In other cases, the Court has said they these rights must yield in order
to protect the speech and information rights of the larger public. The Court required cable
companies to carry local broadcast stations, for example, in the 1990s Turner rulings. A
number of broad propositions can be distilled from these cases, but many questions remain.

Legislation  One takeaway from the Supreme Court cases is, unsurprisingly, that it helps to
have Congress on your side. Federal statutes and FCC-administered regulations supported
the must-carry mandates for cable companies in Turner, and similar laws supported rulings
in the broadcast space.96 PruneYard, which recognized activists’ right to speak in a shopping
mall based on California’s state constitution, is a rare exception. No such statutes support
plaintiffs in today’s must-carry claims against platforms, at least so far. That makes it much
harder for them to overcome platforms’ First Amendment arguments.

Scarcity and Economic Considerations  Another important theme of the communications
cases is the “scarcity rationale,” the principle that regulation, and the resulting burden on
owners’ First Amendment rights, may be justified to ensure fair access to physically finite,
exhaustible communications channels. The Court developed this doctrine in upholding
fairness requirements for broadcast spectrum and later applied similar reasoning to cable.
Some degree of government interference with the companies’ rights was justified, the Court
said, to keep them from restricting, “through physical control of a critical pathway of
communication, the free flow of information and ideas.”97
Platforms like Twitter or YouTube do not exercise the kind of physical bottleneck control
over communications that supported the Court’s rulings against cable companies. In this
sense, user-facing platforms are also different from present-day internet access providers like
ISPs, whose First Amendment challenge to net-neutrality rules was rejected by the US Court
of Appeals, DC Circuit, in 2016.98 Speakers banned from a site like YouTube can still speak
on Reddit or Twitter; users kicked off of those platforms may find a home on Gab or 4chan.
They can also put content online using their own servers.
Many critics argue, though, that the platform ecosystem has created new forms of scarcity.
Even if users can still speak on less-popular platforms, they argue, those may be inadequate
because not enough other people are there to listen or respond.99 Following this argument,
the economic success of a handful of platforms, and the resulting concentration of users—
perhaps aided by network effects or natural monopoly dynamics—effectively creates new
bottlenecks and scarcity of communications channels.
Is economic dominance—or dominance in the “attention marketplace”—enough to justify
must-carry obligations and override platforms’ own speech rights? Must-carry claimants
would say yes, arguing that major platforms, like the cable companies in Turner, control
“critical pathway[s] of communication.”100 Platforms would say that they are instead

Daphne Keller  •  Who Do You Sue? 

 19

like the parade in Hurley: “enviable vehicle[s] for the dissemination of [one’s] views” but
possessed of no “abiding monopoly of access to spectators.”101 This dispute, with its family
resemblance to competition or antitrust law, appears in some First Amendment cases as
a question about whether speakers have adequate alternate channels to speak and access
information. In other cases, the Court expressly considers economic competition and
market share.
In 1990s cases about must-carry rules for cable, the Supreme Court split over Congress’s
authority to override carriers’ First Amendment interests on grounds other than
competition. Justice Kennedy’s four-justice controlling opinion in Turner II recognized
the government’s interest in “preserving a multiplicity of broadcast outlets” and said
Congress could act “regardless of whether the conduct that threatens it is motivated by
anti-competitive animus or rises to the level of an antitrust violation.”102 But Kennedy’s
replacement on the Court, Brett Kavanaugh, sees economic competition as the core
justification for must-carry obligations. In a dissent written as a DC Circuit judge, he
said that “absent some market dysfunction,” imposing net-neutrality rules on ISPs would
violate their First Amendment rights. “If market power need not be shown,” he wrote, “the
Government could regulate the editorial decisions of Facebook and Google” or “impose
forced-carriage or equal-access obligations on YouTube and Twitter.”103
As will be discussed below, a case now pending before the Court, Manhattan Community
Access Corporation v. Halleck,104 will give Kavanaugh and the other current justices a chance
to weigh in again on the tension between cable operators’ and cable programmers’ First
Amendment rights—and, by implication, on the viability of must-carry claims for internet
platforms.

Forums’ Own Speech  Supreme Court must-carry cases also look closely at the means by
which operators of communications channels “speak,” and the ways that laws may impair
their speech. Key questions derived from the case law include these:
What platform operations count as “speech”?
Platforms like Google and Facebook claim to speak through both their ranking decisions
(how they organize user content) and their removal decisions (what content they exclude).
Ranking decisions could strike courts as more “speech-like,” since they arise from companywritten software and communicate judgments about pages’ or posts’ relative value to other
users.105 Following this reasoning, a “must-not-remove” law for today’s platforms might clear
First Amendment hurdles more readily than a “must-not-downrank” law.
On the other hand, the Court has recognized strong First Amendment interests in pure
removal decisions. In Hurley, it said that even if the parade organizer defendant had made
few considered judgments in planning the parade, it “clearly decided to exclude a message

Hoover Institution  •  Stanford University

 20

it did not like from the communication it chose to make, and that is enough” to merit First
Amendment protection.106 In another case, United States v. Playboy, the Court said that a law
requiring a platform to reorganize or limit its distribution of certain content (by scrambling
pornographic channels or showing them only at night) should receive the same degree of
scrutiny as a law that prohibited carrying the content entirely.107 These cases suggest that
the Court might view a law governing ranking and a law governing removal with equal
skepticism.
Would a must-carry obligation prevent a platform from crafting and conveying its own message?
The Court has in some cases focused on whether “the complaining speaker’s own message
was affected by the speech it was forced to accommodate.”108 In PruneYard, for example,
the Court concluded that a shopping mall owner’s own “autonomy” and communication
power were not undermined by leafleters’ presence on its premises.109 By contrast, the
Supreme Court is hostile to laws that would require platforms to carry specific messages,110
or that, like the newspaper right-of-reply law in Tornillo, discourage them from speaking on
topics that might trigger must-carry duties.111 These cases, and the strict scrutiny applied
to content-based speech laws generally, suggest that any law requiring platforms to carry
specific voices or perspectives would face serious First Amendment obstacles.
Would a must-carry obligation cause people to misattribute users’ speech to the platform?
The Court has also considered whether observers might mistakenly believe that a platform
endorsed messages it was compelled to carry. In Hurley, for example, it reasoned that viewers
would likely assume parade organizers supported a pro-gay-rights message if plaintiffs
were allowed to march in the parade. In Turner, by contrast, it said that viewers would
probably not assume that cable companies endorsed the messages of broadcast television
shows.112 The Court has also asked whether platforms can avoid this risk by disavowing any
connection with the speakers.113
These considerations arguably weaken platforms’ First Amendment arguments against mustcarry obligations. Most internet users presumably know that platforms do not endorse third
party speech, and platforms could always add still more disclaimers to their user interfaces
to make that clear. At the same time, public uproar over platforms’ failure to remove
Holocaust denial and other legal-but-offensive speech suggests that many people consider
them responsible for any speech they choose not to remove. This kind of attribution of user
statements to platforms—and resulting pressure to voluntarily remove legal speech—may
simply be a given for any platform that isn’t subject to must-carry rules.

Daphne Keller  •  Who Do You Sue? 

 21

Without access to the platform, can speakers communicate by other means?
Must-carry claims are weaker when speakers have access to adequate alternate channels
of communication. In Turner, for example, the Court saw no alternative for broadcasters if
cable companies cut them off. “A cable operator, unlike speakers in other media,” the Court
noted, can “silence the voice of competing speakers with a mere flick of the switch.”114
(Two years later, two of the same justices seemed to view cable operators’ power differently.
If pornography was not available on cable, they said, viewers could find it via alternate
channels such as theaters or videocassettes.115)
Turner also involved a related financial question: whether broadcasters could survive
without cable ad revenues. Given cable operators’ incentives to discriminate against
broadcast content, and in favor of their own affiliated content, Congress had found
that broadcasters faced particular economic risk. Protecting broadcasters was necessary,
it concluded, to ensure that households without cable could still receive television
programming.116 The Court agreed that this gave the government an interest in keeping
broadcasters in business. This aspect of Turner is so specific to the history of broadcast and
cable that it is unlikely to help plaintiffs in must-carry cases.
Reconciling These Considerations
Overall, these cases suggest that platforms’ First Amendment rights pose a major obstacle
to must-carry claims. But no case really reconciles all the considerations or tells us how
to apply them today. A 1996 case, Denver Area Educational Telecommunications Consortium
v. FCC, comes closest. In it, the justices expressly addressed conflicts between the speech
rights of cable owners and the speech rights of people who watched or shared content
on cable channels.117 Their ruling was highly fractured—and fractious—with six separate
opinions and little majority-supported analysis. Some questions left open in Denver Area
may, however, be answered in a case now pending before the Supreme Court, Manhattan
Community Access Corporation v. Halleck. The case has been billed as a potential gamechanger for internet must-carry claims.
In Denver Area, creators of cable programming challenged three legislative provisions that
permitted or required cable companies to limit pornographic content on channels they
were otherwise required to carry. Importantly for today’s must-carry cases, the Court held
that a provision permitting cable companies to exclude such content from public-access
channels violated the content creators’ First Amendment rights. In other words, content
removal decisions made by a private company under its own discretion—roughly analogous
to most platform takedowns we see today—were constitutionally impermissible. The
plurality opinion noted the “risk that certain categories of programming (say, borderline
offensive programs)” would be removed by the cable companies.118 The Court did not rule

Hoover Institution  •  Stanford University

 22

on the cable companies’ own First Amendment rights or decide whether Congress could
constitutionally compel them to carry the channels in the first place.
In their various opinions, the Denver Area Justices weighed in on nearly every issue raised
in today’s must-carry claims: the relative importance of users’ and channel owners’ speech
rights;119 the analogy to shopping malls;120 government public-forum cases;121 common-lawcommon-carriage law and its relationship to the First Amendment;122 and even the metaquestion of whether courts should reason by analogy to earlier technologies.123 The net
result is great fodder for speculation about the likely outcome if the Court ever accepts an
internet platform must-carry case. But it provides little real guidance.
The now-pending case, Halleck, picks up one of the questions left unresolved in Denver Area:
whether cable public-access channels are public forums for First Amendment purposes. The
plaintiffs are content creators whose material was banned from a Manhattan public-access
channel by the defendant, a nonprofit organization operating the channel under contract
with the city. Since the case was not brought against the cable company, it does not directly
concern the power of communications channels’ owners. In that sense, it is not entirely
analogous to Turner, Denver Area, or must-carry claims against internet platforms. It is more
similar to the case about President Trump’s Twitter account: it concerns only one “channel”
on a larger platform and claims that the channel’s operators are state actors who must
respect First Amendment rights.
There are other reasons why the Court’s eventual ruling in Halleck seems unlikely to
affect internet platforms directly. For one thing, the statutory regimes governing cable
and internet platforms are very different. For another, the argument that cable channels
are public forums mostly rests on cable operators’ specific history of agreements with local
governments, which included commitments to maintain public-access channels.
Still, the case calls for the Court to rule on broadly relevant doctrines, and it provides ample
opportunity for dicta about internet platforms specifically. Given Kavanaugh’s discussion
of YouTube and Twitter in the DC Circuit net-neutrality case, he may be particularly
likely to weigh in. It is also an opportunity for the Court to speak, at least in dicta, to
the government’s own responsibilities. If New York City owes speakers First Amendment
protection and delegates its authority to a vendor, can that vendor really take down any
speech it wants? Or should speakers have a remedy from someone—the government or
its deputy—in the chain of delegation?124 In Halleck, we again see connections between
must-carry claims against private actors and the potential impunity of state actors in cases
involving platforms.

Daphne Keller  •  Who Do You Sue? 

 23

Protecting Online Speech from Private Power
As discussed above, it is far from clear that courts or Congress could, as a constitutional
matter, compel platforms to carry particular speech or messages against their will. There
are also serious policy and philosophical questions about whether we should want them
to. Platforms’ rules against legal-but-offensive speech prevent real-world harms and curtail
behavior that many people find morally abhorrent. They also protect platforms’ economic
value as desirable destinations for users and advertisers.
On the other hand, leaving internet speech policy in the hands of private companies has
ramifications going beyond major platforms to the entire internet speech ecosystem. If any
private internet intermediary can exclude legal but unpopular speech—and has economic
incentives to do so—where is that speech supposed to go? Has the public interest been
served if it is driven offline entirely? As Justice Kennedy noted decades ago, “Minds are
not changed in streets and parks as they once were. To an increasing degree, the more
significant interchanges of ideas and shaping of public consciousness occur in mass and
electronic media.”125 In function, internet platforms have largely replaced public spaces like
parks and streets, while other private intermediaries have displaced public exchange systems
like the postal service or cash currency. If private actors throughout the internet’s technical
stack can exclude legal speech, and are pressured to do so, the online marketplace of ideas
will look very different from the one imagined in First Amendment jurisprudence.
In response to comparable questions, the Supreme Court has occasionally approved mustcarry obligations that “left room for editorial discretion and simply required [owners of
communications channels] to grant others access to the microphone.”126 This implies
that a regime mandating a limited degree of access to important speech channels could,
hypothetically at least, pass constitutional muster. It is hard to imagine what such a “partial
must-carry” regime would look like in the real world, but I will sketch out some approaches
in this section.

Regulating Bigness  Lawmakers outside the United States have experimented somewhat
with setting different rules for hosting platforms depending on their size. Germany’s
NetzDG, for example, holds social networks with more than two million German users to
stringent content-removal time lines, as well as higher standards of public transparency.127
Drafts of the EU’s pending Copyright Directive would also include special obligations for
larger hosts.128 Setting different restrictions depending on size would create problematic
incentives for growing start-ups and is generally not a common approach in American law.
It is also hard to identify a workable definition of “bigness” that would not inadvertently
sweep in complex entities like the thinly staffed, user-managed Wikipedia. Special rules for
mega-platforms would conceptually align with some thinking about competition and the
First Amendment, though, by imposing obligations only on those who control access to
“scarce” communications channels or audiences.

Hoover Institution  •  Stanford University

 24

Empowering Users  For many proponents of online civil liberties, the go-to solution for
problems of platform content moderation is to give users themselves more control over
what they see. Settings on YouTube or Twitter, for example, could include dials and knobs to
signal our individual tolerance for violence, nudity, or hateful speech. This isn’t a cure-all,
but it’s still a great idea. It’s been around at least since the 1990s, when technologies like the
Platform for Internet Content Selection (PICS) were supposed to allow users to choose what
web content appeared in their browsers. Both the Supreme Court in the seminal Reno v.
ACLU internet First Amendment case and Congress in passing CDA 230 relied in part on the
expectation that such technologies would empower consumers.
Today, there remains much to be done to give users more control over their information
diet. There is perhaps a chicken-and-egg question about the paucity of end-user content
controls today and the rise of major, centralized platforms. Did internet users stop
demanding these tools once they found search engines and curated hosting platforms to
protect them from the worst of the web? Or did they flock to centralized platforms because
good tools for true end-user control did not exist? It may be that such tools have only
limited promise as a technical matter, because they depend on accurate content labeling.
A user who wanted to block most racial epithets but retain access to rap lyrics, historical
documents, and news reporting, for example, could do so only if people or algorithms first
correctly identified content in these categories. That’s more work than humans could do at
internet scale, and algorithmic filters have so far proven highly unreliable at tasks of this
sort. Giving everyone his or her own content filter, too, might solve some problems while
exacerbating others—particularly those involving “filter bubbles,” echo chambers, and
attention scarcity.

Hybrid Regimes  As discussed above, there are important questions about whether mustcarry advocates want platforms to carry all legal speech or just some of it, and about
which platform operations they think should be affected. While the more aggressive civil
libertarian position might demand that platforms carry any speech that is legal, some
claimants—and the general public—almost certainly prefer curated platforms where civil
discussion can take place. Speakers seeking an online audience in particular may not want
their platform of choice to become so unattractive that other users depart or that advertisers
cease funding platform operations.
Are there partial must-carry laws that could achieve this goal, and that would at least
partially preserve platforms’ own control over ranking and removal? Below, I list a
few possible models I have discussed over the years with experts in internet law and
technology, both informally and in print or on conference panels. They all involve very
serious trade-offs among competing values. Some of them sound simple in theory but
would be monstrously complex in practice for lawmakers, technologists, or both. Most
also have another thing in common, something that became clearer to me in the course
of writing this essay. They reraise questions that smart people thought and wrote and

Daphne Keller  •  Who Do You Sue? 

 25

legislated and litigated about not that long ago in the context of communications law.
Today’s issues are not identical, because technology and social context have changed. But
the parallels are strong. I am not the lawyer to explain that history or revive the arguments
that communications experts have made in the past. I do think, though, that lawyers and
thinkers in my field will need to understand them much better as the political conversation
about must-carry obligations and overall platform regulation continues.
One possible approach would let platforms act against highly offensive or dangerous content
but require them to tolerate more civil or broadly socially acceptable speech. That kind of
legal regime has plenty of precedent: the FCC has long enforced rules like this for TV and
radio, for example. But it would be a troubling solution for online speech for several reasons.
First, we are far from a national or even local consensus about what’s highly offensive or
dangerous. Second, rules of this sort would, like their TV and radio equivalents, require
substantial and ongoing regulatory intervention and rulemaking to determine which theory
of offensive and dangerous speech should prevail. Regulating platforms’ rapidly evolving
and technically complex ranking algorithms would be particularly challenging.129 Third,
unlike broadcast regulation, rules limiting online speech would reach deep into ordinary
people’s daily communications. And of course, fourth, laws that allowed platforms to take
down some legal posts but not others would use state power to pick winners and losers
among legal speech. That would require a massive rethinking of First Amendment law, to be
resolved by equally massive litigation.
A second variant on this idea would try to avoid state-sanctioned value judgments about
speech, and instead let platforms enforce any rules as long as they are “fair.” Fairness
mandates that simply required rules and processes to be transparent or that barred
discrimination on bases such as race or gender might be comparatively easy to define.
But that kind of fairness wouldn’t achieve what many must-carry proponents want: equal
treatment for different viewpoints. A viewpoint-neutrality rule is much harder to imagine.
Would it require equal treatment for Democrats, Republicans, Socialists, Monarchists,
and Anarchists? For people who like creamy peanut butter and people who like crunchy?
For people urging us to invest in sketchy tech start-ups and people urging us not to?
The potential rules here would also likely resemble the ones the FCC applied to older
communications channels. The equal-time doctrine, for example, required broadcasters
to give equal airtime to qualified candidates for public office. And the fairness doctrine
required “fair” coverage for issues ranging from workers’ rights to nuclear power plant
construction. Critics charged that the doctrine was unworkable and that it effectively
enabled selective enforcement by an unaccountable bureaucracy. The FCC itself eventually
decided the doctrine was unconstitutional, and President Reagan vetoed a bill that would
have brought it back.130
A third variant might say that platforms have to make room for disfavored speech, but
that they don’t have to promote it or give it any particular ranking. As Tim Lee put it,

Hoover Institution  •  Stanford University

 26

we could “think of Facebook as being two separate products: a hosting product and a
recommendation product (the Newsfeed).”131 On this model, a platform could maintain
editorial control and enforce its Community Guidelines in its curated version, which most
users would presumably prefer. But disfavored speakers would not be banished entirely and
could be found by other users who prefer an uncurated experience. Platforms could rank
legal content but not remove it.
This idea has a communications law flavor as well. Internet regulatory models have
traditionally distinguished between “network layer” intermediaries, such as ISPs, that
function as infrastructure and thus have must-carry or net-neutrality obligations; and
user-facing “application layer” services like Facebook or Google. Users of application layer
services typically want content curation and don’t want must-carry rules—or didn’t use
to.132 The increase in demands for must-carry mandates could be taken as calls to rethink
the role of major platforms and to start treating them more like essential, network-layer
internet infrastructure. A “rank but don’t remove” model would recognize this, requiring
major platforms to offer an uncurated, unranked service but preserving their discretion
over the curated version. The problems with this model are less extreme, since it at least
avoids creating new state-sponsored speech rules. But it would still require extensive and
ongoing regulation, resulting in distortion of market incentives and innovation, to decide
what count as the “network” and “application” aspects of any given platform.133 And it
presumably would not be a very satisfactory solution for most must-carry proponents, since
it would still largely deprive them of the audience they seek.
A final variant is what I think of as the “magic APIs” model.134 It is broadly analogous to
telecommunications “unbundling” requirements, which aim to insert competition into
markets subject to network effects by requiring incumbents to license hard-to-duplicate
resources to newcomers. In the platform context, this would mean that Google or Facebook
opens up access to the “uncurated” version of its service, including all legal user-generated
content, as the foundation for competing user-facing services. Competitors would then offer
users some or all of the same content, via a new user interface with their own new content
ranking and removal policies. Users might choose a G-rated version of Twitter from Disney
or an explicitly partisan version of YouTube from a political group, for example. As Mike
Masnick puts it:
Ideally, Facebook (and others) should open up so that third party tools can provide their
own experiences—and then each person could choose the service or filtering setup that
they want. People who want to suck in the firehose, including all the garbage, could do so.
Others could choose other filters or other experiences. Move the power down to the ends
of the network, which is what the internet was supposed to be good at in the first place.135

Letting users choose among competing “flavors” of today’s mega-platforms would solve
some First Amendment problems by leaving platforms’ own editorial decisions undisturbed,

Daphne Keller  •  Who Do You Sue? 

 27

while permitting competing editors to offer alternate versions and include speakers who
would otherwise be excluded. But platforms would object on innumerable grounds,
including the Constitution’s prohibition on most state takings of property. In any case,
this approach would also create a slew of new problems—beyond the ordinary downsides
of regulatory intervention and disruption of private enterprise. The technology required
to make it work would be difficult, perhaps impossible, to build well; that’s the “magic”
part. There are also serious questions about how such a system would interact with the
complex, multiplayer technical infrastructure behind online advertising. And, perhaps
most dauntingly, streamlined systems for users to effectively migrate to competing platform
versions could far too easily run afoul of privacy and data-protection laws.136
It is far from clear to me that any of these ideas could have upsides that outweigh the
downsides. But none has really been given a good tire-kicking by technical and legal
experts, either. The current conversation about must-carry is in this sense in its infancy, for
all the political light and heat it has generated.

Conclusion
Private internet companies have unprecedented power to surveil and curtail our speech.
Never before have so many of our communications shared a common infrastructure, and
hence a common point of control—and never before have so many of us convened in the
same virtual “public square” to share our creativity, our political opinions, our cat pictures,
and all of the other speech we value. We have barely begun to grapple with what this shift
means for our communications ecosystem or our constitutional rights. We should expect
evolution in the relevant law over the coming years.
For now, that evolution is being driven by widely disjointed approaches. Platforms face
constant public pressure to remove more lawful speech, punctuated by occasional demands
to remove less. Free-expression advocates who focus on what they see as abuses of private
power by platforms find themselves with few legal tools, since the Constitution protects
speech rights against state—not private—interference. Constitutional and human-rights
litigators, skilled in fighting government abuses, may see no room to object and no one
to sue when private companies silence online speakers. Governments’ ability to indirectly
restrain lawful expression is only amplified when the law provides few or no checks on
private platforms’ power over online speech.
There are concerns of constitutional dimension in the intersection of state and private
platform power. As this essay has outlined, though, legal tools to challenge either kind
of power are underdeveloped—and in some cases may simply not be possible to devise,
consistent with the Constitution and public-policy concerns. By framing the two issues
together, I have endeavored to set up lines of inquiry and identify possible building blocks

Hoover Institution  •  Stanford University

 28

for wise public policy as we grapple with state and private speech regulation in the new
public square.

Acknowledgments
I am particularly grateful to James Boyle for seeing this all coming, and to Jack Balkin
for the framing his scholarship provides. This essay began as a bulging file labeled “Who
Do We Sue, Jack?,” and the title would still be apt today. Additional thanks to Alex Abdo,
Annemarie Bridy, Josh Cohen, Eileen Donahoe, Harold Feld, Al Gidari, Mike Godwin, Eric
Goldman, Jameel Jaffer, Paddy Leersson, Blake Reid, Dan Svantesson, and Eugene Volokh
for insightful comments. Funding information from the Stanford Center for Internet and
Society is available at http://cyberlaw.stanford.edu/about-us. This work was also supported
in part by a grant from the Knight Foundation.

NOTES
1 ​Casey Stinnett, “Facebook’s Program Thinks Declaration of Independence Is Hate Speech,” Vindicator, July 2,
2018, http://www.thevindicator.com/news/article_556e1014-7e41-11e8-a85e-ab264c30e973.html.
2 ​Packingham v. North Carolina, 137 S. Ct. 1730, 1737 (2017) (holding that North Carolina could not prohibit sex
offenders from using Facebook and other platforms). Three concurring justices, perhaps mindful of the fact that
these “public squares” were also private property, criticized this as “undisciplined dicta.” Id. at 1738.
3 ​Prager University v. Google (“Prager”), 2018 WL 1471939, at *2 (N.D. Cal. March 26, 2018), https://assets​
.documentcloud.org/documents/4421826/Prager-Dismissed.pdf.
4 ​In communications law, the term has a narrower meaning, typically referring to cable companies’ obligations to
carry local broadcast stations. Other mandatory carriage requirements go by other names. Compare 47 U.S.C.A.
§§ 534, 535 (must-carry) with 47 U.S.C.A. § 315 (fairness doctrine and equal time) and 47 U.S.C.A. §§ 531, 532
(government, public access, and leased-access channels).
5 ​United States Telecom Association v. FCC (“USTA”), 855 F.3d 381, 433–34 (DC Cir. 2017) (per curiam) (denying en
banc review of a decision upholding the FCC’s net-neutrality rules) (Kavanaugh, J., dissenting) (arguing that netneutrality rules violated ISPs’ First Amendment rights).
6 ​See Daphne Keller, Internet Platforms: Observations on Speech, Danger, and Money, Hoover Institution, Aegis
Series Paper No. 1807 (June 13, 2018), 5–8, https://www.hoover.org/research/internet-platforms-observations
-speech-danger-and-money; Evan Engstrom and Nick Feamster, The Limits of Filtering: A Look at the Functionality
and Shortcomings of Content Detection Tools, Engine (March 2017), http://www.engine.is/the-limits-of-filtering;
Natasha Duarte, Emma Llansó, and Anna Loup, Mixed Messages? The Limits of Automated Social Media Content
Analysis, Center for Democracy and Technology (November 2017), https://cdt.org/files/2017/11/Mixed-Messages
-Paper.pdf.
7 ​Monika Bickert, “Defining the Boundaries of Free Speech on Social Media,” in The Free Speech Century, ed.
Geoffrey R. Stone and Lee C. Bollinger (Oxford, UK: Oxford University Press, 2018), 267; see also Ben Wagner,
“Free Expression? Dominant Information Intermediaries as Arbiters of Internet Speech,” in Digital Dominance: The
Power of Google, Amazon, Facebook, and Apple, ed. Martin Moore and Damian Tambini (Oxford, UK, and New York:
Oxford University Press, 2018), 225 (implicit bargain between platforms and regulators “ensures that government
regulators push for self-regulatory measures that would not otherwise be legal if they were enshrined in law”).

Daphne Keller  •  Who Do You Sue? 

 29

8 ​This essay only concerns claims against major, user-facing platforms like Facebook and Google—not the
innumerable other entities that host speech online or provide infrastructure within the internet’s technical
“stack.” I include Google’s search index, despite its significant technical differences from hosts such as social
media platforms, because its public role makes it part of the “information gatekeeper” discussion.
9 ​Scholars have long anticipated the emergence of internet companies as “private surrogates” allowing
governments to bypass “pesky constitutional constraints.” James Boyle, “A Nondelegation Doctrine for the
Digital Age?,” Duke Law Journal 50 (2000): 10–11; James Boyle, “Foucault in Cyberspace: Surveillance, Sovereignty,
and Hard-Wired Censors,” University of Cincinnati Law Review 66 (1997): 201 (describing information “guarded by
digital fences which themselves are backed by a state power maintained through private systems of surveillance
and control”). This relationship between the state and private sectors has been called the “invisible handshake,”
Michael D. Birnhack and Niva Elkin-Koren, “The Invisible Handshake: The Reemergence of the State in the
Digital Environment,” Virginia Journal of Law and Technology 8 (2003): 6, and Niva Elkin-Koren and Eldar Haber,
“Governance by Proxy: Cyber Challenges to Civil Liberties,” Brooklyn Law Review 82 (2016): 107; “collateral
censorship,” Michael I. Meyerson, “Authors, Editors, and Uncommon Carriers: Identifying the ‘Speaker’ within
the New Media,” Notre Dame Law Review 71 (1995): 118; or “censorship by proxy,” Seth F. Kreimer, “Censorship
by Proxy: The First Amendment, Internet Intermediaries, and the Problem of the Weakest Link,” University of
Pennsylvania Law Review 155 (2006): 16. See also Matt Schruers, “Copyright, Information Intermediaries, and
Internet Architecture,” in The Turn to Infrastructure in Internet Governance, ed. Francesca Musiani et al. (New
York: Palgrave Macmillan, 2016): 107 (describing emergence of “indirect regulation” in the 1990s). Important
work by Yale law professor Jack Balkin describes a “triangular” relationship between speakers, platforms, and
government, in which “restraint is not at the hands of government bureaucrats, but at the hands of privately
owned companies who act to avoid threats of liability by nation-states.” Jack M. Balkin, “Free Speech Is a
Triangle,” Columbia Law Review 118 (2018): 2019; Jack M. Balkin, “Old-School/New-School Speech Regulation,”
Harvard Law Review 127 (2014): 2296.
10 ​Daphne Keller, “Empirical Evidence of ‘Over-Removal’ by Internet Companies under Intermediary Liability
Laws,” Stanford Law School, Center for Internet and Society, (October 12, 2015), http://cyberlaw.stanford.edu
/blog/2015/10/empirical-evidence-over-removal-internet-companies-under-intermediary-liability-laws.
11 ​Jennifer M. Urban, Joe Karaganis, and Brianna Schofield, Notice and Takedown in Everyday Practice,
University of California–Berkeley, Public Law Research Paper No. 2755628 (2016), 41.
12 ​Meyerson, “Authors, Editors, and Uncommon Carriers.”
13 ​“Manila Principles on Intermediary Liability,” https://www.manilaprinciples.org; David Kaye (Special
Rapporteur), Report of the Special Rapporteur on the Promotion and Protection of the Right to Freedom of
Opinion and Expression, UN Document A/HRC/32/38 (May 11, 2016), 6, https://perma.cc/4RMF-U8EJ (Manila
Principles “establish baseline protection for intermediaries in accordance with freedom of expression
standards”).
14 ​Jeff Kosseff, “First Amendment Protection for Online Platforms,” Computer Law and Security Review 34
(forthcoming 2019); Eric Goldman, “How Section 230 Complements the First Amendment” (forthcoming 2019);
“Section 230 as First Amendment Rule” (student note), Harvard Law Review 131 (2018): 2027; Robert W. Hamilton,
“Liability for Third-Party Content on the Internet,” Seton Hall Constitutional Law Journal 8 (1998): 733; Mike
Godwin, Cyber Rights: Defending Free Speech in the Digital Age (Cambridge, MA: MIT Press, 1998): 79–108.
15 ​Smith v. California, 361 U.S. 147 (1959); Keller, Internet Platforms, 16–20 (discussing bookseller precedent).
16 ​Smith at 154.
17 ​
Id. See also Denver Area Educ. Telecommunications Consortium, Inc. v. FCC, 518 US 727 (1996) (“Denver
Area”) (laws making cable companies responsible for policing programmers’ speech violated programmers’ First
Amendment rights); Kreimer, “Censorship by Proxy,” 59.

Hoover Institution  •  Stanford University

 30

18 ​Center for Democracy and Technology v. Pappert, 337 F. Supp. 2d 606, 610 (E.D. PA 2004) (striking down
statute requiring ISPs to block child pornography); Cubby v. Compuserve, 776 F. Supp. 135 (S.D.N.Y. 1991)
(requiring heightened scienter standard for platform in libel case).
19 ​Magyar Tartalomszolgáltatók Egyesülete v. Hungary, App. No. 22947/13, Eur. Ct. H.R. 135 ¶ 86 (2016) (strict
liability for defamatory comments on news site violated European Convention on Human Rights); Rodríguez,
María Belén c. Google Inc. / daños y perjuicios, CSJN, R.522. XLIX, 28/10/2014, https://perma.cc/6876-2G3P
(Argentina) (strict liability for search engine rejected as threat to free expression); Singhal v. Union of India, (2015)
12 SCC 73, ¶¶ 100, 117 (India) (to protect free expression rights, statute must be construed to require platform
content removal only based on government order); but see Delfi AS v. Estonia, App. No. 64569/09, Eur. Ct. H.R. 586,
¶ 115 (2015) (strict liability permissible for unprotected hate speech in news forum comments).
20 ​See David Erdos, “Communicating Responsibilities: The Spanish DPA Targets Google’s Notification Practices
When Delisting Personal Information,” Inforrm’s Blog, March 21, 2017, https://inforrm.wordpress.com/2017/03/21
/communicating-responsibilities-the-spanish-dpa-targets-googles-notification-practices-when-delisting
-personal-information-david-erdos.
21 ​Id., quoting Article 29 Working Party, Guidelines on the Implementation of Google Spain at 10 (2014) (emphasis
added).
22 ​Woodhull Freedom Foundation v. United States, 334 F. Supp. 3d 185 (D.D.C. 2018) (finding no standing to
challenge constitutionality of 2018 amendments to CDA 230; author is cocounsel).
23 ​Derek E. Bambauer, “Against Jawboning,” Minnesota Law Review 100 (2015): 57, https://ssrn.com/abstract
=2581705.
24 ​Backpage.com v. Dart, 807 F.3d 229 (7th Cir. 2015, Posner, J.). Backpage itself had at the time been found
not liable for sex-trafficking ads based on CDA 230, but this was not the basis for Posner’s ruling. He wrote
that Dart “is violating the First Amendment unless there is no constitutionally protected speech in the ads on
Backpage’s website.” Id. at 231; see also Blum v. Yaretsky, 457 U.S. 991, 1004 (1982) (“[A] State normally can be
held responsible for a private decision only when it has exercised coercive power or has provided such significant
encouragement, either overt or covert, that the choice must in law be deemed to be that of the State.”); Adickes
v. S. H. Kress & Co., 398 U.S. 144, 152 (1970) (overruled on other grounds by Celotex Corp. v. Catrett, 477 U.S.
317 (1986)) (plaintiff is entitled to §1983 relief if police officer and private store personnel “somehow reached
an understanding” to discriminatorily deny her the store’s services); Paige v. Coyner, 614 F.3d 273 (6th Cir. 2010)
(finding state action where private employer fired plaintiff in response to state official’s complaint about her
public-policy advocacy); Missouri Knights of the Ku Klux Klan v. Kansas City, 723 F. Supp. 1347 (W.D. Mo. 1989)
(refusing to dismiss claim of state action and First Amendment violation where city council acted jointly with
private cable franchisee to exclude racist speakers); Cal. Educ. Code §48950 et seq. (2018) (“Leonard law” applying
First Amendment to private and public schools); Tim Wu, Is the First Amendment Obsolete?, Columbia University,
Knight First Amendment Institute, Emerging Threats series (September 2017), https://knightcolumbia
.org/content/tim-wu-first-amendment-obsolete (discussing similar cases); Eugene Volokh, “Private Employees’
Speech and Political Activity: Statutory Protection against Employer Retaliation,” Texas Review of Law and
Politics 16 (2012): 295 (discussing private employer retaliation based on employees’ political activity); Meyerson,
“Authors, Editors, and Uncommon Carriers,” 112–14 (discussing jawboning examples and case law); Bambauer,
“Against Jawboning,” 107 (discussing unconstitutional conditions doctrine).
25 ​Derek E. Bambauer, “Orwell’s Armchair,” University of Chicago Law Review 79 (2012): 892; see also Yochai
Benkler, “A Free Irresponsible Press: Wikileaks and the Battle over the Soul of the Networked Fourth Estate,”
Harvard Civil Rights–Civil Liberties Law Review 46 (2011): 314 (informal state pressure on platforms shows “a direct
vulnerability in the private infrastructure system and a potential pathway of public censorship”); Annemarie
Bridy, “Internet Payment Blockades,” Florida Law Review 67 (2015): 1525.

Daphne Keller  •  Who Do You Sue? 

 31

26 ​Kreimer, “Censorship by Proxy,” 25, citing Zieper v. Metzinger, 392 F. Supp. 2d 516, 522–23, 538 (S.D.N.Y. 2005)
(finding triable issue of fact on First Amendment claim that defendants coerced intermediaries to take down
lawful video, but finding qualified immunity).
27 ​Kreimer, “Censorship by Proxy,” 22, discussing Pilchesky v. Miller, No. 3:05-CV-2074 (M.D. Pa. Dec. 21, 2005).
28 ​Bridy, “Internet Payment Blockades,” 1545 (using Julia Black’s terminology); see also Comments of the
Computer & Communications Indus. Association in Response to Voluntary Best Practices Study 1 (Aug. 21, 2013)
(questioning the “unexamined premise that the U.S. Government should be evaluating unregulated, private
sector action in the first place”), http://www.ccianet.org/wp-content/uploads/library/CCIA%20Voluntary%20
Practices.pdf.
29 ​European Commission, Code of Conduct on Countering Illegal Hate Speech Online (May 2016), http://
ec.europa.eu/newsroom/just/item-detail.cfm?item_id=54300. See also Danielle Keats Citron, “Extremist Speech,
Compelled Conformity, and Censorship Creep,” Notre Dame Law Review 93 (2018): 1035 (“These changes are less
the result of voluntary market choices than of a bowing to governmental pressure”); Evelyn Aswad, “The Role of
U.S. Technology Companies as Enforcers of Europe’s New Internet Hate Speech Ban,” Columbia Human Rights Law
Review Online (2016): 6 (arguing that the code of conduct is inconsistent with human-rights principles).
30 ​Article 19, Responding to “Hate Speech”: Comparative Overview of Six EU Countries (2018), https://www
.article19.org/wp-content/uploads/2018/03/ECA-hate-speech-compilation-report_March-2018.pdf.
31 ​European Commission, Commission Recommendation on Measures to Effectively Tackle Illegal Content Online
(March 1, 2018), https://ec.europa.eu/digital-single-market/en/news/commission-recommendation-measures
-effectively-tackle-illegal-content-online.
32 ​European Commission, Proposal for a Regulation of the European Parliament and of the Council on Preventing
the Dissemination of Terrorist Content Online (September 12, 2018), https://ec.europa.eu/commission/sites/beta
-political/files/soteu2018-preventing-terrorist-content-online-regulation-640_en.pdf.
33 ​Kate O’Flaherty, “YouTube Keeps Deleting Evidence of Syrian Chemical Weapon Attacks,” Wired, June 26, 2018,
https://www.wired.co.uk/article/chemical-weapons-in-syria-youtube-algorithm-delete-video.
34 ​Samuel Gibbs, “EU Warns Tech Firms: Remove Extremist Content Faster or Be Regulated,” Guardian, December
7, 2017, https://www.theguardian.com/technology/2017/dec/07/eu-warns-tech-firms-facebook-google-youtube
-twitter-remove-extremist-content-regulated-european-commission; “Extremist Propaganda and Social Media,”
hearing before the Committee on Commerce, Science, and Transportation, C-SPAN, January 17, 2018, https://
www.c-span.org/video/?439849-1/facebook-twitter-youtube-officials-testify-combating-extremism&start=859.
35 ​Report from source known to author.
36 ​Europol, EU Internet Referral Unit: Year One Report (July 22, 2016), https://www.europol.europa.eu/sites
/default/files/documents/eu_iru_1_year_report_highlights.pdf. The report emphasizes that a referral from
Europol “does not constitute an enforceable act. Thus, the decision and removal . . . is taken by the concerned
service provider under their own responsibility and accountability” (p. 4).
37 ​The US State Department long embraced the idea of exporting US speech, speech rules, and speechenhancing internet technologies. See Jack L. Goldsmith, The Failure of Internet Freedom, Columbia University,
Knight First Amendment Institute, Emerging Threats series (June 2018), https://knightcolumbia.org/content
/failure-internet-freedom.
38 ​For important early discussion of government leverage, see Jack L. Goldsmith, “The Internet and the Abiding
Significance of Territorial Sovereignty,” Indiana Journal of Global Legal Studies 5 (1998): 475.
39 ​See Graham Smith, “Geoblocking Tools and the Law,” 5–8, and Daphne Keller, “Glossary: Internet Content
Blocking Options and Vocabulary,” 51–55, in Law, Borders, and Speech Conference: Proceedings and Materials, ed.
Daphne Keller (Stanford, CA: Stanford Center for Internet and Society, 2017).

Hoover Institution  •  Stanford University

 32

40 ​Eva Glawischnig-Piesczek v. Facebook Ireland Limited, Case C-18/18 (pending case), https://eur-lex.europa.eu
/legal-content/EN/TXT/?uri=CELEX%3A62018CN0018.
41 ​Google Inc. v. Equustek Solutions Inc., 2017 SCC 34 [2017].
42 ​Bridy, “Internet Payment Blockades,” 1557–58 (discussing US pressure on Russia through WTO).
43 ​Paul Mozur, “China Presses Its Internet Censorship Efforts across the Globe,” New York Times, March 2, 2018,
https://www.nytimes.com/2018/03/02/technology/china-technology-censorship-borders-expansion
.html?smid=tw-share.
44 ​Google Inc. v. Commission nationale de l’informatique et des libertés (CNIL), Case C-507/17 (pending case),
https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=uriserv:OJ.C_.2017.347.01.0022.02.ENG.
45 ​Bickert, “Defining the Boundaries of Free Speech on Social Media,” 260.
46 ​Sydney Li and Jamie Williams, “Despite What Zuckerberg’s Testimony May Imply, AI Cannot Save Us,”
Electronic Frontier Foundation, April 11, 2018, https://www.eff.org/deeplinks/2018/04/despite-what-zuckerbergs
-testimony-may-imply-ai-cannot-save-us.
47 ​Zhang v. Baidu.com, Inc. (“Zhang”), 10 F. Supp. 3d 433 (S.D.N.Y. 2014); Sikhs for Justice, Inc. v. Facebook, Inc.,
144 F. Supp. 3d 1088 (N.D. Cal. 2015).
48 ​A district court held that enforcing France’s removal order would violate the First Amendment. But the
appellate court, noting that the French plaintiffs were not actually asking US courts to enforce the order,
dismissed the case. Yahoo! Inc. v. La Ligue Contre Le Racisme et L’Antisemitisme, 433 F.3d 1199, 1210 (9th Cir.
2006).
49 ​Yahoo! at 1215 (“After the French court entered the orders, Yahoo! voluntarily changed its policy to comply
with them, at least to some extent”).
50 ​Dan Jerker B. Svantesson, Solving the Internet Jurisdiction Puzzle (Oxford, UK: Oxford University Press, 2017).
51 ​Google Inc. v. Equustek Solutions Inc., 2017 SCC 34 at para 46 [2017] 1 S.C.R. 824. This case is ongoing, and the
author was involved as counsel to Google at the appellate stage. In its ruling, the Supreme Court invited Google
to submit evidence of real conflict with laws outside of Canada. Google sought and received a US “declaratory
judgment that the Canadian court’s order cannot be enforced in the United States and an order enjoining that
enforcement.” Google LLC v. Equustek Solutions Inc., No. 5:17-cv-04207-EJD, 2017 WL 50000834, at *2 (ND Cal.
Nov. 2, 2017) Order Granting Plaintiff’s Motion for Preliminary Injunctive Relief at 2–3. As discussed above,
however, on remand the lower Canadian court found this insufficient reason to vary its global removal order.
52 ​Equustek Solutions Inc. v. Jack, 2018 BCSC 610 (Supreme Court of British Columbia, April 16, 2018) para 20–22.
53 ​Bambauer, “Against Jawboning,” 109.
54 ​Dan Jerker B. Svantesson, “Jurisdiction in 3D: ‘Scope of (Remedial) Jurisdiction’ as a Third Dimension of
Jurisdiction,” Journal of Private International Law 12 (2016): 60.
55 ​Svantesson, “Jurisdiction in 3D,” 217–18; Factum of Human Rights Watch et al as Interveners, Google v.
Equustek, 2017 SCC 34 [2017] 1 S.C.R. 824 (arguing that when countries interpret rights differently within the range
permissible under international human-rights law, no country should impose its version on the others), https://
cis-static.law.stanford.edu/cis/downloads/HRW%20Equustek.pdf.
56 ​47 USC 230(c) (2) (immunity from must-carry claims based on good-faith efforts to exclude objectionable
content); 17 USC 512(g) (immunity from must-carry claims for platforms that carry out counternotice process).
57 ​Professor Eric Goldman’s blog is by far the best source of information on these cases and includes links to
rulings: https://blog.ericgoldman.org. See, e.g., Green v. Am. Online, Inc., 318 F.3d 465, 472 (3d Cir. 2003); Langdon
v. Google, Inc., 474 F. Supp. 2d 622 (D. Del. 2007); Estavillo v. Sony Computer Entertainment America, No. C–09–

Daphne Keller  •  Who Do You Sue? 

 33

03007 RMW, 2009 WL 3072887 (N.D. Cal. Sept. 22, 2009); Buza v. Yahoo, Inc., No. C 11–4422 RS, 2011 WL 5041174
(N.D. Cal. Oct. 24, 2011); e360Insight, LLC v. Comcast Corp., 546 F. Supp. 2d 605 (N.D. Ill. 2008).
58 ​Cyber Promotions, Inc. v. America Online, 948 F. Supp. 436 (E.D. Pa. 1996) (rejecting the argument that
censorship of email violated the First Amendment).
59 ​Johnson v. Twitter, Inc., No. 18CECG00078 (Cal. Superior Ct. June 6, 2018); Kamango v. Facebook, No. 3:11CV-0435, 2011 WL 1899561 (N.D. NY April 19, 2011); Estavillo, No. C-09-03007 RMW; Langdon; Prager, supra note 3;
compare hiQ Labs, Inc. v. LinkedIn, 273 F. Supp. 3d 1099 (N.D. Cal. 2017) (claiming a right of access for the purpose
of consuming, rather than posting, information).
60 ​Williams v. Life’s Rad, No. C 10–0086 SBA, 2010 WL 5481762 (N.D. Cal. May 12, 2010).
61 ​Zhang, supra note 47.
62 ​Twitter v. Superior Court ex rel Taylor, A154973 (Cal. Ct. App. Aug. 17, 2018); Lewis v. YouTube, 197 Cal. Rptr.
3d 219 (Cal. Ct. App. 2015); Hall v. Earthlink Network, Inc., No. 98 Civ. 5489 (RO), 2003 WL 22990064, at *1 (S.D.N.Y.
Dec. 19, 2003); Young v. Facebook, Inc., 790 F. Supp. 2d 1110, 1118–19 (N.D. Cal. 2011) (speculating that a better
pled contractual claim could succeed).
63 ​Williams, No. C 10–0086 SBA; KinderStart.com LLC v. Google, Inc., No. C 06-2057 JF (RS), 2007 WL 831806 (N.D.
Cal. March 16, 2007); Person v. Google, No. 07-16367, 2009 WL 3059092 (9th Cir. Sept. 24, 2009).
64 ​Williams, No. C 10–0086 SBA.
65 ​Six4three v. Facebook, No. C 17-00359 WHA, 2017 WL 657004 (N.D. Cal Feb. 17, 2017); Prager, supra note 3;
KinderStart.com, No. C 06-2057 JF (RS); Quigley v. Yelp, 2017 U.S. Dist. LEXIS 103771 (N.D. Cal. July 5, 2017);
Jayne v. Google Internet Search Engine Founders, No. 07-4083, 2008 WL 324214 (3d Cir. Feb. 7, 2008); e-ventures
Worldwide v. Google, Inc., No. 2:14-cv-646-FtM-PAM-CM, 2017 WL 2210029 (M.D. Fla. Feb. 8, 2017); Bartholomew v.
YouTube, LLC, 225 Cal. Rptr. 3d 917 (Cal. Ct. App. 2017); but see Bragg v. Linden Research, 487 F Supp 2d 593 (E.D.
Pa. 2007) (declining to dismiss in case involving monetary loss).
66 ​Song Fi v. Google, Inc., No. 14-cv-05080-CW, 2018 WL 2215836 (N.D. Cal. May 15, 2018); Darnaa v. Google
LLC, 2018 WL 6131133 (9th Cir. Nov. 21, 2018); Crawford v. Consumer Depot, Inc., No. 05C-3242 (Tenn. Ct. App. Dec.
8, 2009).
67 ​Lancaster v. Alphabet Inc., No. 15-cv-05299-HSG, 2016 WL 3648608 (N.D. Cal. July 8, 2016); Riggs v. MySpace,
Inc., No. 09–56633, 2011 WL 3020543 (9th Cir. July 25, 2011); Sikhs for Justice, supra note 47; Langdon, supra note
57; Mezey v. Twitter, Inc., No. 1:18-cv-21069-KMM, 2018 WL 5306769 (S.D. Fla. July 19, 2018); Zango v. Kaspersky,
568 F.3d 1169 (9th Cir. 2009) (immunity for provider of malware blocklist); but see e-ventures, supra note 65
(declining to dismiss on 230 grounds because plaintiff adequately pled lack of good faith); Spy Phone Labs
LLC. v. Google Inc., No. 15-cv-03756-KAW, 2016 WL 6025469 (N.D. Cal. Oct. 14, 2016) (same); Annemarie Bridy,
“Remediating Social Media: A Layer-Conscious Approach,” Boston University Journal of Science and Technology
Law (2018): 208–9 (discussing “good faith” standard); Eric Goldman, “Online User Account Termination and 47
U.S.C. §230(c) (2),” UC Irvine Law Review 2 (2012): 665.
68 ​E-ventures, supra note 65 at *4 (“[t]he First Amendment protects these decisions, whether they are fair or
unfair, or motivated by profit or altruism”); La’Tiejira v. Facebook, Inc., 272 F. Supp. 3d 981, 991–922 (S.D. Tex.
2017); Zhang, supra note 47; Langdon supra note 57; Search King, Inc. v. Google Technology, Inc., No. CIV-021457-M, 2003 WL 21464568 (W.D. Okla. May 27, 2003).
69 ​Philippe Sotto, “French Court Issues Mixed Ruling in Facebook Nudity Case,” U.S. News and World Report,
March 15, 2018, https://www.usnews.com/news/business/articles/2018-03-15/french-court-issues-mixed-ruling
-in-facebook-nudity-case.
70 ​See, e.g., 26 O 5492/18, Regional Court Munich I (July 20, 2018); 24 O 187/18, Cologne Regional Court (July
27, 2018); 5 O 69/18, Siegen Regional Court (August 22, 2018); David Meyer, “Court Tells Facebook: Stop Deleting

Hoover Institution  •  Stanford University

 34

‘Offensive’ Comment,” ZDNet, April 13, 2018, https://www.zdnet.com/article/court-tells-facebook-stop-deleting
-offensive-comment.
71 ​Civil Appeal No. 0000412-86.2016.8.24.0175 (Meleiro); Civil Appeal No. 0000447-46.2016.8.24.0175 (Meleiro).
In an interesting analog involving payment processors, US copyright holders convinced Visa and Mastercard to
cut off services globally to a site that operated legally in Russia. Before going out of business, the site successfully
sued the credit card companies in Russia for violating their Terms of Service. Bridy, “Internet Payment Blockades,”
1557.
72 ​Marco Civil da Internet, Federal Law no. 12.965 (2014).
73 ​Aleksandra Kuczerawy, “The Power of Positive Thinking: Intermediary Liability and the Effective Enjoyment of
the Right to Freedom of Expression,” Journal of Intellectual Property, Information Technology and E-Commerce
Law 8 ( 2017): 226, https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3033799.
74 ​Douwe Korff, Council of Europe, The Rule of Law on the Internet and in the Wider Digital World (December
2014), 23, https://rm.coe.int/16806da51c (states should “stop relying on private companies … to impose
restrictions that are in violation of the state’s human rights obligations” and limit even “measures implemented
by private parties for business reasons”); Christina Angelopoulos et al., Institute for Information Law, Study of
Fundamental Rights Limitations for Online Enforcement through Self-Regulation (2016), 50–51, http://www.ivir.nl
/publicaties/download/1796.
75 ​See Meyerson, “Authors, Editors, and Uncommon Carriers,” 114–15 (discussing obligations of common
carriers).
76 ​Julia Angwin, Madeleine Varner, and Ariana Tobin, “Facebook Enabled Advertisers to Reach ‘Jew Haters,’”
ProPublica, September 14, 2017, http://www.propublica.org/article/facebook-enabled-advertisers-to-reach
-jew-haters; Daphne Keller, Toward a Clearer Conversation about Platform Liability, Columbia University, Knight
First Amendment Institute, Emerging Threats series, 2018, https://knightcolumbia.org/content/toward-clearer
-conversation-about-platform-liability (discussing misperception that CDA 230 causes platforms to tolerate such
speech).
77 ​Snyder v. Phelps, 562 U.S. 443, 454 (2011).
78 ​Olivia Solon, “Google’s Bad Week: YouTube Loses Millions as Advertising Row Reaches US,” March 25, 2017,
https://www.theguardian.com/technology/2017/mar/25/google-youtube-advertising-extremist-content-att
-verizon (describing advertiser pressure on YouTube to change extremist content policies).
79 ​There are also more nuanced options, some of which blur the lines between these categories. Platforms can
put a label on disfavored content (as Facebook does with discredited news stories); put it behind a warning page
(as YouTube does for some violent or sexual material); disable comments or ads (as YouTube did with Dennis
Prager’s videos); remove it in some geographic regions (as Facebook, Twitter, YouTube, and other platforms
do for legal violations); or make it available only for users who opt in to see the content (as Google does with
pornography in search results). The demotion option, which Google has used for copyright-infringing sites, is one
reason that “removal” and “ranking” decisions are not always entirely separate.
80 ​See, e.g., Renee DiResta, “Free Speech Is Not the Same as Free Reach,” Wired, August 30, 2018, https://www
.wired.com/story/free-speech-is-not-the-same-as-free-reach.
81 ​See Jonathan Peters, “The ‘Sovereigns of Cyberspace’ and State Action: The First Amendment’s Application
(or Lack Thereof) to Third-Party Platforms,” Berkeley Technology Law Journal 32 (2017): 994; Bambauer, “Orwell’s
Armchair,” 910–13.
82 ​326 U.S. 501 (1946).
83 ​Prager, supra note 3 at 11; see also Testimony of Berin Szóka, Platform Responsibility & Section 230, Filtering
Practices of Social Media Platforms: Hearing Before the House Committee on the Judiciary (April 26, 2018), 19–21

Daphne Keller  •  Who Do You Sue? 

 35

(distinguishing Marsh from contemporary platforms), http://docs.techfreedom.org/Szoka_Testimony-Platform
_Reponsibility_&_Neutrality-4-25-18.pdf [sic].
84 ​447 U.S. 74 (1980). Compare Hudgens v. NLRB, 424 U.S. 507, 521 (1976) (federal First Amendment “has no part
to play” in a case against privately owned shopping center).
85 ​See sources cited in note 67.
86 ​Knight Institute v. Trump, 302 F. Supp. 3d 541 (S.D.N.Y. 2018); Davison v. Randall, No. 17-2002 (4th Circ. Jan. 7,
2019).
87 ​Knight Institute at 568 (noting that “further analysis may be necessary when the party exercising control over
the forum is a nongovernmental entity”).
88 ​Peter Swire, “Should the Leading Online Tech Companies Be Regulated as Public Utilities?,” Lawfare Blog,
August 2, 2017.
89 ​United States v. Terminal Railroad Association, 224 U.S. 383 (1912); Matthew Lane, “Antitrust in 60 Seconds:
What Is the ‘Essential Facilities Doctrine’ in the U.S.?,” Disruptive Competition Project, April 4, 2018, http://www
.project-disco.org/competition/040418antitrust-in-60-seconds-what-is-the-essential-facilities-doctrine-in-the
-u-s/#.Wz0Cr9hKhE4.
90 ​Taylor, supra note 62.
91 ​See Verizon v. FCC, 740 F.3d 623, 651 (DC Cir. 2014) (describing codification of common-law common-carriage
rules for transportation and communication services). See also Harold Feld, “My Insanely Long Field Guide to
Common Carriage, Public Utility, Public Forum—and Why the Differences Matter,” Wetmachine, September 5,
2017, http://www.wetmachine.com/tales-of-the-sausage-factory/my-insanely-long-field-guide-to-common
-carriage-public-utility-public-forum-and-why-the-differences-matter.
92 ​Eugene Volokh and Donald M. Falk, “Google—First Amendment Protection for Search Engine Search Results,”
Journal of Law, Economics, and Policy 8 (2012): 883.
93 ​Turner Broad. Sys. v. FCC, 512 U.S. 622, 629 (1994) (“Turner I”) (cable operators’ selection of channels is
expressive even if a cable company generally acts as “a conduit for the speech of others, transmitting it on a
continuous and unedited basis to subscribers”).
94 ​418 U.S. 241 (1974).
95 ​515 U.S. 557 (1995).
96 ​Red Lion Broadcasting Co. v. FCC, 395 U.S. 367 (1969).
97 ​Turner I, supra note 93 at 657.
98 ​USTA, supra note 5; see also John Blevins, “The New Scarcity,” Tennessee Law Review 79 (2012): 353; Bridy,
“Remediating Social Media”; Cindy Cohn, “Bad Facts Make Bad Law: How Platform Censorship Has Failed So Far
and How to Ensure That the Response to Neo-Nazis Doesn’t Make It Worse,” Georgetown Law Technology Review
2 (2018): 437–41.
99 ​See, e.g., Wu, Is the First Amendment Obsolete? (“if it was once hard to speak, it is now hard to be heard”);
Zeynep Tufekci, “It’s the (Democracy-Poisoning) Golden Age of Free Speech,” Wired, January 16, 2018, https://
www.wired.com/story/free-speech-issue-tech-turmoil-new-censorship (“[t]he most effective forms of censorship
today involve meddling with trust and attention”); Herbert Simon, “Designing Organizations for an InformationRich World,” in Computers, Communications, and the Public Interest, ed. Martin Greenberger (Baltimore, MD: Johns
Hopkins University Press, 1971), 40 (“a wealth of information creates a poverty of attention”).
100 ​Turner I, supra note 93 at 657.
101 ​Hurley v. Irish-American Gay, Lesbian and Bisexual Group of Boston, Inc., 515 U.S. 557 (1995), 577–78.

Hoover Institution  •  Stanford University

 36

102 ​Turner Broad. Sys., Inc. v. FCC, 520 U.S. 180 (1997) (“Turner II”), 194 (identifying competition as one of three
state interests along with media pluralism and preservation of local broadcast stations, and rejecting dissent
argument that ruling should turn solely on competition); 223 (rejecting contention that broadcasters should
pursue more burdensome antitrust remedies); Turner I, supra note 93 at 640 (rejecting argument that mustcarry was only a response to “market failure in a market whose commodity is speech” and noting that “the mere
assertion of dysfunction or failure in a speech market, without more, is not sufficient” to justify must-carry
obligations).
103 ​USTA, supra note 5 at 433.
104 ​No. 17-702, reviewing 882 F.3d 300 (2d Cir. 2018).
105 ​See, e.g., Denver Area, supra note 17 at 739 (cable operators’ speech interests were “relatively weak because
they act less like editors, such as newspapers or television broadcasters, than like common carriers, such as
telephone companies”).
106 ​Hurley, supra note 101 at 574.
107 ​529 U.S. 803, 812 (2000) (“It is of no moment that the statute does not impose a complete prohibition. The
distinction between laws burdening and laws banning speech is but a matter of degree”).
108 ​Rumsfeld v. Forum for Academic & Inst’l Rights, Inc., 547 U.S. 47, 63 (2006).
109 ​Hurley, supra note 101 at 579–80 (describing PruneYard holding); see also Turner I, supra note 93 at 655
(must-carry will not “force cable operators to alter their own messages”).
110 ​See Turner I, supra note 93 at 655 (contrasting carriage mandates only for “those who disagree with
appellant’s views” in Pacific Gas & Electric v. Pub. Util. Comm’n of Cal., 475 U.S. 1, 14 [1986]).
111 ​Turner I, supra note 93 at 656 (contrasting Tornillo, in which “the safe course is to avoid controversy”);
PruneYard Shopping Center v. Robins, 447 US 74 (1980), 87–88 (“no specific message is dictated by the State”).
112 ​Hurley, supra note 101 at 576–77 (distinguishing Turner’s lower misattribution risk), 580 (distinguishing
PruneYard on same basis); PruneYard, supra note 111 at 87.
113 ​PruneYard, supra note 111 at 87; Turner I, supra note 93 at 655; Hurley, supra note 101 at 580; Rumsfeld, supra
note 108 at 60 (2006).
114 ​Turner I, supra note 93 at 656; see also id. at 653–54 (distinguishing newspaper in Tornillo); Hurley, supra note
101 at 579–80 (noting that PruneYard ruled for plaintiffs without requiring them to show that they would be fully
silenced if the mall excluded them).
115 ​Denver Area, supra note 17 at 744. Justices Stevens’ and Souter’s inconsistent treatment of local broadcast
content in Turner and pornography in Denver Area lends some credence to the Turner dissent, which argued that
must-carry requirements for local broadcasting were not content-neutral. Turner II, supra note 102 at 230.
116 ​Turner I, supra note 93 at 647.
117 ​Denver Area, supra note 17 at 790 (identity of petitioners). An earlier lower court case, Midwest Video v. FCC,
lays out colorfully and at length the competing First Amendment rights of carriers and programmers, as well
as risks to programmer speech when carriers act as “a corps of involuntary government surrogates” policing
programmers’ content. 571 F.2d 1025, 1056 (8th Cir. 1978).
118 ​Breyer Plurality Op at 766; see also Kennedy Op at 782 (law “singles out one sort of speech for vulnerability
to private censorship in a context where content-based discrimination is not otherwise permitted”). Cable
companies, which at the time were still litigating their own First Amendment claims in Turner, were not parties to
Denver Area.

Daphne Keller  •  Who Do You Sue? 

 37

119 ​Kennedy Op (prioritizing programmers’ rights); Thomas Op (prioritizing cable operators’ rights); Breyer
Plurality Op (declining to prioritize).
120 ​Kennedy Op at 793; Thomas Op at 820, fn. 5.
121 ​Kennedy Op at 783 (public-access channels are public forums); Thomas Op at 826 (they aren’t); Stevens Op at
773 (declining to decide).
122 ​Kennedy Op at 783 (cable operators were common carriers for leased access channels and thus barred from
content discrimination); Thomas Op. at 824–25 (private common carriers can discriminate and are not subject to
First Amendment claims).
123 ​Breyer Op at 742 (rejecting analogies); Kennedy Op at 787 (approving). Compare Heather Whitney, Search
Engines, Social Media, and the Editorial Analogy, Columbia University, Knight First Amendment Institute,
Emerging Threats series (February 2018), https://knightcolumbia.org/content/search-engines-social-media-and
-editorial-analogy.
124 ​For a more detailed discussion, see Daphne Keller, “Online Speech Rights and the Supreme Court’s Pending
Halleck Case,” Stanford University, Center for Internet and Society, January 28, 2019, https://cyberlaw.stanford
.edu/blog/2019/01/online-speech-rights-and-supreme-courts-pending-halleck-case.
125 ​Denver Area, supra note 17 at 802–3 (Kennedy, J., concurring).
126 ​FCC v. League of Women Voters of California, 468 U.S. 364, 385 (1984).
127 ​Act to Improve Enforcement of the Law in Social Networks, Article 1, Section 1 (English translation)
(2017), https://www.bmjv.de/SharedDocs/Gesetzgebungsverfahren/Dokumente/NetzDG_engl.pdf?
__blob=publicationFile&v=2.
128 ​European Commission, Proposal for a Directive of the European Parliament and of the Council on Copyright
in the Digital Single Market (September 14, 2016), https://ec.europa.eu/digital-single-market/en/news/proposal
-directive-european-parliament-and-council-copyright-digital-single-market.
129 ​See Blevins, “The New Scarcity” (describing logistical challenges of such regulation). For a sense of the
challenge, Google reports that of more than two trillion search queries it receives each year, 15 percent are
completely new. That works out to more than 800 million unprecedented search result sets—and opportunities
for “unfair” ranking—each day. Barry Schwartz, “Google Reaffirms 15% of Searches Are New, Never Been
Searched Before,” Search Engine Land, April 25, 2017, https://searchengineland.com/google-reaffirms-15
-searches-new-never-searched-273786.
130 ​Kathleen Ann Ruane, “Fairness Doctrine: History and Constitutional Issues,” Congressional Research Service,
July 13, 2011, https://fas.org/sgp/crs/misc/R40009.pdf; see also Adam Thierer, “Why the Fairness Doctrine Is
Anything but Fair,” Heritage Foundation, October 29, 1993, https://www.heritage.org/government-regulation
/report/why-the-fairness-doctrine-anything-fair; Szóka, supra note 83 at 3 (“Opposition to reinstatement of the
Fairness Doctrine has been in every GOP platform since 2008”); Wu, Is the First Amendment Obsolete? (a fairness
doctrine for platforms would be “too hard to administer, too prone to manipulation, and too apt to flatten what
has made the Internet interesting and innovative”).
131 ​Timothy B. Lee, “Alex Jones Is a Crackpot—But Banning Him from Facebook Might Be a Bad Idea,” Ars
Technica, August 6, 2018, https://arstechnica.com/tech-policy/2018/08/op-ed-alex-jones-is-a-crackpot-but
-banning-him-from-facebook-might-be-a-bad-idea.
132 ​Blevins, “The New Scarcity”; Bridy, “Remediating Social Media.”
133 ​The line between curated and uncurated product components is difficult to draw: Would your racist uncle’s
tweets simply never appear in your feed but be visible on his profile page? Be hidden from the casual visitor to his

Hoover Institution  •  Stanford University

 38

profile page but findable through search? If Google dislikes the only web page that has the text string “fifty purple
pigeons eat potatoes,” could it rank a hundred other pages above it when users search for those exact words?
134 ​APIs, or application program interfaces, are technical tools that allow one internet service to connect with
and retrieve information from another.
135 ​Mike Masnick, “Platforms, Speech and Truth: Policy, Policing and Impossible Choices,” Techdirt, August 9,
2018, https://www.techdirt.com/articles/20180808/17090940397/platforms-speech-truth-policy-policing
-impossible-choices.shtml (emphasis omitted). Jonathan Zittrain has similarly suggested that “Facebook should
allow anyone to write an algorithm to populate someone’s feed.” Quoted in Michael J. Coren, “Facebook Needs to
Hand Over Its Algorithm If It Really Wants to End Fake News,” Quartz, December 6, 2016, https://qz.com/847640
/facebook-needs-to-hand-over-its-algorithm-if-it-really-wants-to-end-fake-news.
136 ​See generally Kevin Bankston, “How We Can ‘Free’ Our Facebook Friends,” New America, June 28, 2018,
https://www.newamerica.org/weekly/edition-211/how-we-can-free-our-facebook-friends (discussing privacy
laws’ potential “unintended consequence of locking in current platforms’ dominance by locking down their
control over your data”).

Daphne Keller  •  Who Do You Sue? 

 39

The publisher has made this work available under a Creative Commons Attribution-NoDerivs license 3.0. To view a copy
of this license, visit https://creativecommons.org/licenses/by-nd/3.0.

The author’s photo shown on the back page is available under a Creative Commons Attribution-NonCommercialShareAlike license 2.0. To view a copy of this license, visit https://creativecommons.org/licenses/by-nc-sa/2.0. Some
rights reserved by the Stanford Center for Internet and Society. Photograph by Amanda Avila.
Hoover Institution Press assumes no responsibility for the persistence or accuracy of URLs for external or third-party
internet websites referred to in this publication and does not guarantee that any content on such websites is, or will
remain, accurate or appropriate.
Copyright © 2019 by the Board of Trustees of the Leland Stanford Junior University
22 21 20 19

5 4 3 2 1

The preferred citation for this publication is Daphne Keller, Who Do You Sue? State and Platform Hybrid Power over
Online Speech, Hoover Working Group on National Security, Technology, and Law, Aegis Series Paper No. 1902 (January
29, 2019), available at https://www.lawfareblog.com/who-do-you-sue-state-and-platform-hybrid-power-over-online
-speech

Hoover Institution  •  Stanford University

 Working Group on National Security,
Technology, and Law

PHOTO BY AMANDA AVILA

About the Author

DAPHNE KELLER
Daphne Keller is the director
of intermediary liability at the
Stanford Center for Internet and
Society. Her work focuses on
the ways that laws governing
platforms’ responsibility for online
information affect the rights of
internet users. She was previously
associate general counsel to Google,
where she led the web search legal
team.

The Working Group on National Security, Technology, and
Law brings together national and international specialists
with broad interdisciplinary expertise to analyze how
technology affects national security and national security
law and how governments can use that technology to defend
themselves, consistent with constitutional values and the rule
of law.
The group focuses on a broad range of interests, from
surveillance to counterterrorism to the dramatic impact that
rapid technological change—digitalization, computerization,
miniaturization, and automaticity—are having on
national security and national security law. Topics
include cybersecurity, the rise of drones and autonomous
weapons systems, and the need for—and dangers of—state
surveillance. The group’s output will also be published on
the Lawfare blog, which covers the merits of the underlying
legal and policy debates of actions taken or contemplated
to protect the nation and the nation’s laws and legal
institutions.
Jack Goldsmith is the chair of the National Security,
Technology, and Law Working Group.
For more information about this Hoover Institution Working
Group, visit us online at hoover.org/research-teams/national​
-security-technology-law-working-group.

Hoover Institution, Stanford University
434 Galvez Mall
Stanford, CA 94305-6003
650-723-1754

Hoover Institution in Washington
The Johnson Center
1399 New York Avenue NW, Suite 500
Washington, DC 20005
202-760-3200