For the Balkinization 20th Anniversary Symposium
Robert Post
This post summarizes a
talk that I gave to the Global Constitutionalism seminar at Yale in 2022. Its
topic is how the internet might endanger democracy. The ideas contained in the
post are tentative and speculative. They are chiefly intended to propose an
agenda for further study.
Legal regulation of the internet in the United States
is currently stunted by Section 230 of the Communication Decency Act, which
broadly immunizes service providers from liability. One can easily imagine, however,
a world in which the internet is controlled by the same forms of legal regulation
as apply to other mass media. Social media platforms would then subject to
actions for defamation and invasion of privacy, in just the same way as are
newspapers. The question I wish to explore in this post is not whether the
repeal of Section 230 is desirable, but rather whether the internet poses dangers
to democracy that are distinct from those threatened by past media. Insofar as
this is true, the question arises whether the internet should be subject to new
and innovative forms of regulation that have not previously been applied to
traditional mass media.
We can begin to identify the potentially unique dangers of the internet by specifying the ways in which current social media differ from traditional mass media. Three such differences come immediately to mind: zero marginal information cost, integration with life tasks, and interactivity.
Zero marginal information cost: Information
spreads on the internet in a frictionless way that is virtually cost-free. The
price of sending information to 1000 persons is no greater than sending it to 1
person. The price of sending information to someone on the other side of the
globe is no greater than sending it to someone around the corner. As a result, the
internet differs from prior mass media in three important ways:
a. Scale:
Information spreads on the internet on a scale that is orders of magnitude greater
than that attainable by any prior medium of communication. Facebook, for
example, had 2,936,000,000 users during the first quarter of 2022.
b. Virality:
Not only is the quantity of information distributed on the internet
greater than that transmitted by prior media, but the speed of that
distribution is far faster. When we speak about the virality of information on the
internet, we refer to the almost unimaginably rapid pace by which information spreads
from person to person in the virtual public sphere.
c. Cosmopolitanism:
The scale and virality of information spread on social media renders national
borders almost irrelevant. This has put immense pressure on the integrity of national
public spheres. In the past public spheres have typically been conceptualized as
tied to particular nation states. It used to make sense to speak of the German
theater, of the English press, of the French novel, or of the American cinema.
But the medium of the internet is so cosmopolitan that we can now begin to perceive
a public sphere that is truly international.
Integration with life tasks: The internet is
now typically accessed through phones, which have become all-purpose tools for
negotiating life tasks—for getting directions, ordering food, contacting friends,
getting local news, charging expenses, etc. Traditional media did not permeate
everyday life in this way. We may have watched a great deal of television or
spent many Sunday mornings reading the newspaper, but we now participate in the
virtual public sphere at a scale that dwarfs any previous engagement with the
traditional public sphere. Because the internet is seamlessly integrated into
our ordinary lives, its influence has become unimaginably huge.
Interactivity: In traditional mass media,
speakers unilaterally address large audiences. Social media, by contrast,
are built on the principle of interactivity. They are constructed to facilitate
conversations in virtual space. Although traditional mass media always competed
for the attention of readers and viewers, the internet differs from these media
in two ways:
a. Epistemological Authority. Traditional
media had professional gatekeepers who vouched for the authenticity and
epistemological value of distributed information. The editors of newspapers and
magazines staked their reputations on the quality of the product they
distributed to readers. By contrast, the internet has no such gatekeepers. Social
media instead produces conversations. Participants are less like the readers of
a newspaper than they are like persons who gather to gossip on a street corner
or around the water cooler at work. Epistemological validation is thus not the
product of authoritative speakers, but of self-reinforcing circles of gossip.
b. Polarization. Although
traditional mass media often targeted discrete groups who were potentially at
odds with each other, social media actually create such groups. In part this is
because persons have integrated social media so seamlessly into the conduct of
their everyday lives. Social media creates the new phenomenon of gossip among
those who are otherwise strangers. Because gossip circles are virtually defined
by the distinction between those who are inside and those who are outside, the
internet strongly reinforces the tribalism that has come to dominate our public
space.
Taken together, these
three major differences between the internet and prior traditional mass media
pose at least six major unique threats to democracy.
First, the internet
has contributed to a pervasive loss of epistemological authority. Insofar as epistemological
authority in social media is established within discreet gossip circles, we are
as a society losing the ability to identify figures of authority whom the general
public can trust to distinguish truth from fiction. A symptom of this
phenomenon is the growing distrust of expertise. Without the ability to
establish common facts, without the capacity to believe in shared disciplinary
methods or trust figures of broad epistemological authority, democracy is
profoundly threatened.
Second, the American
demos has begun to think like a crowd instead of like a public. Democracy
presupposes a public that thinks, and thinking does not occur instantaneously. It
took time to assimilate and digest the information received from traditional
public media. The virality and interactivity of the internet, its integration
in real time into the pressing tasks of everyday life, is inconsistent with this
kind of public reflection. As a medium, the internet privileges immediate
reactions that marginalize self-conscious thoughtfulness and contemplation. The
result is that the American public is increasingly adopting a kind of mob
mentality, characterized by rapid and instantaneous responses, that is inconsistent
with democracy.
Third, the scale of
the internet produces forms of harm that may best be characterized as stochastic.
Previously we asked whether particular speech acts might cause
particular harms. The internet has rendered this kind of question almost
obsolete. Speech that is simultaneously distributed to billions of persons may
produce harm in ways that cannot meaningfully be conceptualized through the
lens of discreet causality. We will need instead to think in terms of the statistical
probability of harm. Yet at present we lack any legal framework capable of
assessing such stochastic harms in ways that will not drastically over-regulate
speech.
Fourth, most legal systems offer special
protections to speech that is about public officials or public figures, or that
is about matters of public concern. Speech that is distributed to the public at
large is often accorded such protections, as can paradigmatically be seen in
the unique privileges enjoyed by the press in almost every modern,
sophisticated legal system. The modern architecture of freedom of speech
heavily depends upon the fundamental distinction between public and private
speech. Public speech is protected insofar as modern legal systems seek to
preserve the free development of public opinion to insure public
accountability. Private speech is more heavily regulated to sustain the
community norms that define and maintain personal dignity and respect. Because it
makes it so difficult to distinguish public from private speech, the internet
threatens this fundamental legal architecture.
Even the most personal communication on the internet can achieve a wider
circulation than do the largest newspapers. On the internet the personal
becomes political; the formation of public opinion has merged into private
circles of gossip. Yet the so much in democracy depends upon clear demarcations
between public and private spheres of speech and action.
Fifth, the regulation
of speech in almost all modern sophisticated legal systems is predicated upon
the relationship between a demos and its state. All systems of freedom of
speech aspire to subordinate political authority to national public opinion.
But because the internet produces opinion that is cosmopolitan rather than
national, it threatens this entire framework of analysis. It is not clear why
the American government should hold itself accountable to Russian public
opinion. Once the internet transforms the public sphere into an international
phenomenon, therefore, we shall have to develop entirely new paradigms to
govern our protections for freedom of speech.
Sixth, and most
subtle, the scale of the internet makes its regulation incompatible with law.
This is a point that has been most elaborately conceptualized by the innovative
Israeli scholar Niva Elkin-Koren. Elkin-Koren writes that law always depends
upon the exercise of human judgment in the articulation and application of
norms. Yet the internet is far too big to be policed through the exercise of human
judgment. During the first quarter of 2022, for example, Facebook alone took
down some 151,900,000 pieces of content. These removals resulted in some 2,614,400
appeals. No court has the capacity to oversee this volume of business. No human
judgment can operate at this scale.
The real governance of
the internet therefore happens mechanically, through the algorithms of AI. These
algorithms, which continuously evolve and learn, are ontologically distinct
from the human normative judgments that are the essence of law. Whereas all
modern societies have developed methods to legitimate judicial decision-making,
no society has yet established a method of legitimizing the application of
algorithms. A potential crisis is brewing insofar as democracy depends upon
speech, and insofar as speech is controlled by politically unaccountable
algorithms.
A possible solution to this crisis might emphasize
that AI algorithms learn through iterative training, which might offer the
potential for political legitimation. Consider, for example, the Oversight
Board of Facebook, of which I am presently a trustee. The Oversight Board
decides something like 50 cases per year. The Oversight Board uses these cases
to articulate legal norms to inform Facebook’s content moderation decisions. Legal
norms have meaning only when they are applied through the medium of human
judgment. The vast majority of Facebook’s content moderation decisions,
however, are not made by human beings.
This means that the norms articulated by the Oversight Board must be translated into the AI algorithms by which Facebook polices the vast bulk of it content. Even though there is an inescapable incompatibility between law and algorithms, Facebook’s AI decision-making might nevertheless be politically legitimated were relevant stakeholders to participate in the process of its iterative training. My prediction is that in the future, society will have to develop some such mechanisms for making AI publicly accountable. Until that happens, the actual governance of the internet, which is of such great importance to the public opinion formation necessary for democracy, will remain technologically estranged both from law and from political legitimation.
Robert C. Post
is Sterling Professor of Law at Yale Law School. You can reach him by e-mail at
robert.post@yale.edu.