US Advisory Commission on Public Diplomacy: Can Public Diplomacy Survive the Internet? – Bots,
Echo chambers & Disinformaton
Forward: public diplomacy in a post-truth society
Modern electronic communicaton is transforming the spread and impact of ideas in
unpredictable ways. Individuals and organizatons can now share informaton widely and
instantly at no cost, bypassing conventonal media and its traditonal role in curatng news,
focusing civic agendas, and moderatng debate.
While technological advances have enabled broader partcipaton in public discussion, they
have also fractured it into silos where dubious assertons and accusatons can reverberate
unchallenged. Fake news is disseminated for proft or politcal advantage. Extremists have
new forums in which to spread hatred and lies with impunity. And foreign actors infuence
domestc policy undetected. Compounding the problems, individuals have litle ability to
discern the identty of interlocutors or basis for distnguishing fact from fcton.
The speed and scale of today’s “weaponizaton of informatonn is unprecedented. Propelled
by novelty, falsehood ofen travels faster than truth, leaving context and provenance behind.
The traditonal answer to the spread of bad informaton has been to inject good informaton
into the mix, on the assumpton that the truth would rise to the top. But in a world of trolls
and bots, where simple facts are instantly countered by automated agents, this strategy may
not be adequate. It is unclear how efectvely democratc societes can contnue to deliberate
and functon, and how hostle foreign actors can be identfed and neutralized.
Executive summary
Digital’s Dark Side focuses on the emergence of social bots, artfcial intelligence, and
computatonal propaganda. Essays in this secton aim to raise awareness regarding how
technology is transforming the nature of digital communicaton, ofer ideas for competng in
this space, and raise a number of important policy and research questons needing
immediate atenton. The Disinformaton secton confronts 羆xford English Dictonary’s 2016
word of the year – “post-truthn – with a series of compelling essays from practtoners, a
social scientst, and philosopher on the essental roles that truth and facts play in a
democratc society. Here, theory, research, and practce neatly align, suggestng it is both
crucial and efectve to double-down on fact-checking and evidence-based news and
informaton programming in order to combat disinformaton campaigns from our
adversaries. The Narrative secton concludes the report by focusing on how technology and
facts are ultmately part of, and dependent on, strategic narratves. Beter understanding
how these narratves form, and what predicts their likely success, is necessary to think
through precisely how PD can, indeed, survive the Internet. Below are some key takeaways
from the report.
IN DEFENSE OF TRUTH
people tend to trust what others say, an efect called the truth bias.
At the same tme people are also contnuously evaluatng the validity of their understanding
of the world. This process is called “epistemic vigilance,n a contnuous process checking that
the informaton that a person believes they know about the world is accurate.
COMPUTATIONAL PROPAGANDA
Computatonal propaganda refers to the coordinated use of social media platforms,
autonomous agents and big data directed towards the manipulaton of public opinion.
Social media bots (or “web robotsn) are the primary tools used in the disseminaton of
computatonal propaganda. In their most basic form, bots provide basic answers to simple
, questons, publish content on a schedule or disseminate stories in response to triggers
(e.g.breaking news). Bots can have a disproportonate impact because it is easy to create a
lot of them and they can post a high-volume content at a high frequency.
Politcal bots aim to automate politcal engagement in an atempt to manipulate public
opinions. They allow for massive amplifcaton of politcal views and can empower a small
group of people to set conversaton agenda’s online.
Advances in artfcial intelligence (AI) – an evolving constellaton of technologies enabling
computers to simulate cognitve processes – will soon enable highly persuasive machine-
generated communicatons. Imagine an automated system that uses the mass of online data
to infer your personality, political preferences, religious afliaton, demographic data and
interests. It knows which news websites and social media platforms you frequent and it
controls multple user accounts on those platforms. The system dynamically creates content
specifcally designed to plug into your partcular psychological frame and achieve a partcular
outcome.
Digital tools have tremendous advantages over humans. 羆nce an organizaton creates and
confgures a sophistcated AI bot, the marginal cost of running it on thousands or millions of
user accounts is relatvely low. They can operate 24/7/3 5 and respond to events almost
immediately.
MOVING PAST “FOLK THEORIES”
Folk theories, or how people think a partcular process works, are driving far too many digital
strategies. 羆ne example of a folk theory is in the prevalence of echo chambers online, or the
idea that people are increasingly digitally walled of from one another, engaging only with
content that fts cognitve predispositons and preferences.
Research suggests that the more users rely on digital platforms (e.g. Twiter and Facebook)
for their news and informaton, the more exposure they have to a multtude of sources and
stories.
Despite increased exposure to a pluralistc media ecosystem, we are becoming more and
more ideological and partsan, and becoming more walled of at the interpersonal and
physical layers. For example, marriages today are twice as likely to be between two people
with similar politcal views than they were in 69 01.
Despite this growing ideological divide, people are increasingly willing to trust one another,
even complete strangers, when their goals are aligned (see the sharing economy, for
example).
RETHINKING OUR DIGITAL PLATFORMS AND METRICS
Virality – the crown jewel in the social media realm – is overemphasized ofen at the expense
of more important metrics like context and longevity. Many of the metrics used to measure
the efectveness of social media campaigns are vulnerable to manipulaton, and more
importantly, don’t measure engagement in any meaningful way.
STRATEGIC NARRATIVES
Strategic narratves—a means for politcal actors to construct a shared meaning of the past,
present and future of politcs in order to shape the behavior of other actors—provide the
ideological backdrop for how audiences assess the meaning and signifcance of current
events and breaking news. Put another way, they help people make sense of what would
otherwise be a dizzying onslaught of news they are exposed to on a daily basis.
REMARKS ON “PUBLIC DIPLOMACY IN A POST-TRUTH SOCIETY”
There has been much discussion in the media, academia, and within the U.S. government
about living in a post truth or post factual society and how to operate in it. Much was made
of 羆xford Dictonary’s decision to make post truth the Word of the Year in 2016 , an adjectve
, they defned as “relatng to or denotng circumstances in which objectve facts are less
infuental in shaping public opinion than appeals to emoton and personal belief.n
In such a world, the public policy debate is framed largely by what “feels true and what
correlates with people’s pre-existng set of beliefs and prejudices, which can ofen be
disconnected from actual facts and the specifcs of policy. It isn’t so much that facts are
dismissed entrely, but rather they are of secondary importance or simply not as compelling,
especially when they challenge what feels true at an instnctual level. In this context, all
opinions have equal weight, regardless of how extreme they may be.
While this is not a new concept it has played a role in politcs since antquity in our age, social
media has exacerbated the problem, acceleratng the speed at which false stories spread,
creatng digital wild fres of misinformaton.
Compounding the problem is the actve work of non state and state actors who aim not only
to disseminate misinformaton but, most damaging, to erode trust in traditonal sources of
informaton. These actors — whom Get Smart fans might collectvely call “KA羆Sn—do not
necessarily want people to believe they are telling the truth, but rather to think that no one
is. Their goal is to diminish public trust in government insttutons, established media outlets,
and subject mater experts, leaving citzens open to the infuence of an onslaught of
questonable informaton generated through re-enforcing social media loops.
While there is much that is accurate about this descripton, I would like to contest the view
that we are living in a “post-truthn society — if by that we mean truth and facts no longer
mater. Facts do exist. They are out there we cannot operate without them. And they
remain compelling when they are part of a larger truth-based narratve that is backed up by
supporting actons.
COMPETITION FROM PSEUDO-FACTS
What we are facing now is intense competton at all levels. Facts compete with pseudo-facts
on substance, on speed, and for audiences’ atenton. And yes, people accept stories that
“feeln true more readily than stories that challenge their beliefs. But they accept them
because they believe they are true.
Brexit is ofen cited as an example of the post-truth phenomenon. In the UK, trust in experts
and confdence in government have both increased since a similar poll in 20164, and both
people who voted to leave and to remain in the EU shared much the same view.
nearly 901 percent of Americans say it is “extremelyn or “very importantn that the media get
its facts correct. Furthermore, about 401 percent say they can remember a specifc incident
that eroded their confdence in the media, most ofen one involving inaccuracies or a
percepton of one-sidedness, making factual accuracy the most important component of
public trust in journalism.
Communicators, experts, and ofcials may feel overwhelmed and succumb to inacton or,
worse, be seduced into adoptng “post-truth techniquesn that appeal only to emoton and
sideline facts or challenging audiences’ beliefs.
There is also the temptaton to counter the barrage of misinformaton by atemptng to rebut
every false story, but this is a losing propositon. There are too many of them, they spread
too quickly, and there are too few of us to chase them.
6) people tend to believe something when it is repeated, 2) propagandists gain the
advantage when they get to make the frst impression, and 3) subsequent rebutals may ac-
tually work to reinforce the original misinformaton, rather than dissipate it.
The way to counter pseudo-facts and misinformaton is to present a compelling narratve of
our own, one that is true, defensible, and based on the enduring values and goals that
, people share, not the least of which is strengthening our collectve security and prosperity.
To gain credibility and make our narratve relevant, we must also listen to and acknowledge
our audiences’ underlying fears, grievances, and beliefs. But it is not just a mater of telling a
good story the narratve must be ted to acton.
Finally, there is one last critcal element in this debate. In additon to ofering compelling,
truthful narratves, I believe we must also help foreign audiences targeted by concerted
disinformaton campaigns to beter understand the dangers of acceptng everything at face
value and encourage them to cultvate a “healthy skeptcism.n By this I do not mean to
promote paranoia, simply vigilance.
COMPUTATIONAL PROPAGANDA AND POLITICAL BOTS: AN OVERVIEW
Computatonal propaganda is best defned as the assemblage of social media platforms,
autonomous agents and big data directed towards the manipulaton of public opinion. Social
media bots are the primary tools used in the disseminaton of computatonal propaganda.
When bots are used to automate politcal engagement in atempts to manipulate public
opinion our team at the University of 羆xford calls them “politcaln bots. Politcal bots allow
for massive amplifcaton of politcal views, they can empower a small group of people to set
conversaton agenda’s online. They are used over social media to manufacture trends, game
hashtags, megaphone partcular content, spam oppositon and atack journalists. When in
the hands of powerful, well-resourced, political actors these automated tools can be used to
both boost and silence communicaton and organizaton among citzens in both democratc
and authoritarian regimes.
Security experts argue that more than 601 percent of content across social media websites,
and 2 per- cent of all web trafc, is generated by bots. Bots dominate many mundane tasks
on the internet, from aiding in the generaton of personal online news preferences, to ad
generaton, to promotng matches on social media. News organizatons now use bots to track
and disseminate breaking artcles. Sites like Wikipedia, which generate publically accessible
knowledge, use bots as an essental part of their labor force. The latest social bots are
automated sofware used on social media platforms to undertake tasks and mimic real users.
Twiter users’ tactcs of purchasing massive amounts of bots to signifcantly boost follower
numbers. Militaries, state-contracted frms and elected ofcials now use politcal bots to
invasively spread various forms of propaganda and food newsfeeds with politcal spam.
Politcally oriented bots are an emergent phenomena and are amongst the most important
recent innovatons in politcal strategy and communicaton technology. Bots are prevalent
and actve in social media conversatons—and their presence in these spaces contnues to
grow. The noise, spam and manipulaton inherent in many bot deployment techniques
threaten to disrupt civic conversatons and organizaton worldwide.
UNDERSTANDING COMPUTATIONAL PROPAGANDA
Many computer and social scientsts treat bot-generated trafc as a nuisance to be detected
and managed, thus extant systems work to simply identfy or block accounts that appear to
be running as automatic scripts. This approach is too simplistc and avoids focusing on the
larger, systemic problems presented by politcal bot sofware. Politcal bots suppress free
expression and civic innovaton via the demobilizaton of actvist groups and the sufocaton
of democratc free speech. Politcal bots and computatonal propaganda must, therefore, be
beter understood for the sake of free speech and digitally mediated civic engagement. The
informaton that exists on politcal bots is disjointed and ofen isolated to specifc, country or
electon-oriented, events.
