| 1
DISENTANGLING UNTRUTHS ONLINE: CREATORS, SPREADERS AND HOW TO STOP THEM © OECD 2022
Disentangling untruths online:
Creators, spreaders and how to stop them
2 |
DISENTANGLING UNTRUTHS ONLINE: CREATORS, SPREADERS AND HOW TO STOP THEM © OECD 2022
This Toolkit note was written by Molly Lesher, Hanna Pawelec and Arpitha
Desai. It was reviewed by the Committee on Digital Economy Policy (CDEP),
and it was declassified by the CDEP on 16 March 2022. The note was prepared
for publication by the OECD Secretariat.
This Toolkit note is a contribution to the OECD Going Digital project, which aims
to provide policy makers with the tools they need to help their economies and
societies thrive in an increasingly digital and data-driven world.
For more information, visit www.oecd.org/going-digital
.
#GoingDigital
Please cite this publication as:
Lesher, M., H. Pawelec and A. Desai (2022), "Disentangling untruths online:
Creators, spreaders and how to stop them", OECD Going Digital Toolkit Notes, No.
23, OECD Publishing, Paris, https://doi.org/10.1787/84b62df1-en
.
Note to Delegations:
This document is also available on O.N.E. under the reference code:
DSTI/CDEP(2021)19/FINAL.
This document, as well as any data and map included herein, are without prejudice
to the status of or sovereignty over any territory, to the delimitation of
international frontiers and boundaries and to the name of any territory, city or area.
© OECD 2022
The use of this work, whether digital or print, is governed by the Terms and
Conditions to be found at http://www.oecd.org/termsandconditions
.
| 3
DISENTANGLING UNTRUTHS ONLINE: CREATORS, SPREADERS AND HOW TO STOP THEM © OECD 2022
Table of Contents
Disentangling untruths online: Creators, spreaders and how to stop them ........ 4
Why is access to accurate information important? .............................................................. 6
Disentangling the different types of untruths online ......................................................... 8
How are untruths spread online and what are the consequences? ............................... 10
Surveying the evidence base of untruths online ................................................................. 13
Approaches to fighting untruths online and mitigating their negative effects ......... 17
Conclusion ....................................................................................................................................... 22
Annex. A Selection of innovative approaches to fighting untruths online and
mitigating their negative effects ................................................................................................. 23
References .......................................................................................................................................... 30
Figures
Figure 1. A typology of untruths online....................................................................................... 9
4 |
DISENTANGLING UNTRUTHS ONLINE: CREATORS, SPREADERS AND HOW TO STOP THEM © OECD 2022
Disentangling untruths online:
Creators, spreaders and how to stop them
While false rumours, inaccurate reporting, and conspiracy theories have
existed for as long as there were people to create and spread them, the
Internet has reshaped and amplified the ability to produce and perpetuate
false and misleading content. Stopping the creators and spreaders of
untruths online is essential to reducing political polarisation, building public
trust in democratic institutions, improving public health, and more generally
improving the well-being of people and society. This Going Digital Toolkit
note discusses the importance of access to accurate information online and
presents a novel typology of the different types of untruths that circulate on
the Internet. It considers how untruths are spread online as well as the
consequences, and it surveys the evidence base of false and misleading
information online. It concludes by identifying approaches to fighting
untruths online and mitigating their negative effects.
| 5
DISENTANGLING UNTRUTHS ONLINE: CREATORS, SPREADERS AND HOW TO STOP THEM © OECD 2022
The global free flow of information was one of the main drivers of the early
Internet movement. The pioneers of the Internet’s architecture viewed an
open, interconnected and decentralised Internet as a vehicle to bridge
knowledge gaps worldwide and promote learning in disadvantaged
communities, thus becoming a great information “leveller”. Despite these
idealistic beginnings, however, societies across the world are now confronted
with the dystopian prospect that instead of being a facilitator of knowledge
and information, the Internet has become a key conduit for spreading untruths
with a speed and reach that is unprecedented (OECD, 2019
[1]
); (OECD, 2019
[2]
).
Untruths – particularly propaganda and disinformation – played an important
role in the lead up to the Russian Federation’s invasion of Ukraine, and they
continue to circulate strongly in both countries (Barnes, 2022
[3]
). As the conflict
unfolds, spreaders of untruths are employing co-ordinated efforts across
multiple platforms, with social media such as Telegram and TikTok, mainstream
media outlets, and propaganda-based websites all being used to sway the
emotions and beliefs of ordinary citizens in the two countries and around the
world (Frenkel, 2022
[4]
); (Scott, 2022
[5]
).
In the context of the COVID-19 pandemic, concerns about an “infodemic” – or
information overload – have emerged. Some of this information is false and
inaccurate, and it is being spread online (WHO et. al., 2020
[6]
); (OECD, 2020
[7]
).
Such information is largely about the COVID-19 virus, its origins and effects,
cures and remedies, and actions taken by the government or public health
officials to manage the pandemic. More recently, untrue information relating
to the safety of vaccines has caused vaccine hesitancy across the world,
undermining the efforts of governments towards nation-wide inoculation
(Jongh, Rofagha and Petrosova, 2021
[8]
); (OECD, 2021
[9]
).
Similarly, false information has prevented people across the globe from
accessing accurate and truthful information about elections, which has in turn
adversely affected democratic processes and institutions (Colomina, Margalef
and Youngs, 2021
[10]
). Some countries have experienced foreign influence on
elections in the form of disinformation campaigns that have increased voter
fraud and suppression, and reduced trust in the legitimacy of elections (Taylor,
2019
[11]
). In other countries, “influence firms” have been found to illegally
harvest personal data and profile users of online platforms for the purpose of
delivering targeted political content to such users (Rosenberg, Confessore and
Cadwalladr, 2018
[12]
).
Throughout history, false rumours, incorrect reporting, and conspiracy theories
have existed. The harm caused by such untruths has varied from being
innocuous to causing severe mental and physical harm (Ireton and Posetti,
2018
[13]
). Recent revelations, including document disclosures, have only
intensified the impression that while the dissemination of falsehoods is not
new, the Internet has reshaped and amplified the ability to create and
6 |
DISENTANGLING UNTRUTHS ONLINE: CREATORS, SPREADERS AND HOW TO STOP THEM © OECD 2022
perpetuate content in ways that we are only just beginning to understand (Wall
Street Journal, 2021
[14]
); (Politico, 2021
[15]
). Such inaccurate and misleading
information can intensify social polarisation, erode public trust in democratic
institutions, and harm people and society more broadly.
There is support from people across age groups and levels of education that
action is needed to tackle inaccurate information online (Mitchell and Walker,
2021
[16]
); (Henry, 2021
[17]
). Technology firms themselves have also called for
more oversight and regulation (Clegg, 2021
[18]
); (Schaake, 2021
[19]
). This Going
Digital Toolkit note discusses the importance of access to accurate information
online and disentangles the different types that circulate. It also considers how
untruths are spread online as well as the consequences, and it surveys the
evidence base of false and misleading information online. It concludes by
identifying approaches to fighting untruths online and mitigating their
negative effects.
Why is access to accurate information important?
The right to freedom of speech, thought and expression, coupled with a free
and independent press, are indispensable for the healthy functioning of
democratic societies. The concerns surrounding the quality and accuracy of
information available through the press or on online platforms presents a
challenge for the protection of fundamental human rights enjoyed by all
individuals under international, regional, and national legal frameworks,
including the International Covenant on Civil and Political Rights (ICCPR) and
the Universal Declaration of Human Rights (UDHR). These rights include not
only the right to freedom of thought, speech, and expression that is necessary
for interacting within the public sphere, but also the right to health, the right
to privacy, and the right to access and receive reliable information that allows
for public participation in democratic processes, one of the cornerstones of the
OECD’s acquis on open government
1
. However, the propagation of untruths
endangers these human rights, reduces trust in the media, and undermines
democratic norms, national security and public order.
Article 21 of the UDHR grants citizens the right to choose their leaders in free,
fair, and regular elections as well as the right to access accurate information
about parties, candidates and other factors that may influence voting. The
United Nations Human Rights Committee also imposes an obligation on
Member States to ensure that “voters should be able to form opinions
independently, free of violence or threat of violence, compulsion, inducement
or manipulative interference of any kind” (UNCHR, 1996
[20]
). However, surveys
1
The OECD Recommendation of the Council on Open Government defines open government as
“a culture of governance that promotes the principles of transparency, integrity, accountability
and stakeholder participation in support of democracy and inclusive growth” (OECD, 2017
[88]
).
| 7
DISENTANGLING UNTRUTHS ONLINE: CREATORS, SPREADERS AND HOW TO STOP THEM © OECD 2022
suggest that political untruths negatively impact a country's politics, causing
polarisation among communities, and also sow distrust in democratic
institutions such as governments, parliaments, and courts as well as distrust of
public figures, journalists and the media (CIGI-IPSOS, 2019
[21]
); (Green, 2020
[22]
).
In the context of the COVID-19 pandemic, which has significantly increased our
reliance on technology and the Internet, health-related untruths have caused
issues for public health systems worldwide (OECD, 2020
[23]
). Trust in the health
information disseminated by entities such as the media, governmental bodies,
and health professionals is essential, especially in pandemic times (Swire-
Thompson and Lazer, 2020
[24]
). However, some users of online platforms –
including elected representatives (Lerer, 2021
[25]
) – have taken to the Internet
to spread misinformation and disinformation related to the global pandemic,
thereby jeopardising our collective right to health.
Importantly, the right to freedom of thought, speech, and expression, which is
the cornerstone of free democracies and protected under Article 19(2) of the
ICCPR, is threatened by inaccurate and misleading information that interferes
with people’s ability to exercise socio-political and economic choices (OHCHR,
2001
[26]
). In the digital context, online platforms have become the arbiters of
communication, where they balance free speech and, at the same time, require
users to adhere to terms of service that can potentially limit speech (Heins,
2013
[27]
).
Recently, some examples of untruths have also unfairly interfered with the
right to privacy and data protection of users of online platforms. Through
content distribution techniques (e.g. “micro-targeting”), as well as algorithmic
bias that delivers specific content to users based on their personal data, some
individuals and entities have leveraged technology to spread falsehoods. Well-
established systems of data collection contribute to this phenomenon and, in
some cases, intrude on people’s right to privacy and their right to form their
ideas free from manipulation. The United Nations Special Rapporteur on the
Promotion and Protection of the Right to Freedom of Opinion and Expression
highlights in its latest report on disinformation that sensitive personal data or
information such as one’s ethnicity or political affiliation could potentially be
dangerous when untruths online fuel political violence (Khan, 2021
[28]
).
Beyond fundamental rights, there are other reasons that access to accurate
information is important. Key issues in this respect include information about
climate change, including its causes and impacts, non-COVID-19 health issues
(such as the dangers of smoking tobacco), as well as conspiracy theories (e.g.
related to cults or emotional events such as the origins of the 9/11 attacks) and
hoaxes of different types.
8 |
DISENTANGLING UNTRUTHS ONLINE: CREATORS, SPREADERS AND HOW TO STOP THEM © OECD 2022
Disentangling the different types of untruths online
Given that there is no generally recognised typology of untrue content online,
this Toolkit note surveys the literature to propose a coherent set of definitions
to bring clarity to the international debate around untruths online. False,
inaccurate, and misleading information often assumes different forms based
on the context, source, intent and purpose. It is critical to distinguish between
the various types of untrue information to help policymakers design well-
targeted policies and facilitate measurement efforts to improve the evidence
base in this important area.
• Disinformation refers to verifiably false or misleading information that
is knowingly and intentionally created and shared for economic gain or
to deliberately deceive, manipulate or inflict harm on a person, social
group, organisation or country (EC, 2019
[29]
). Fake news
2
, synthetic
media, including deepfakes,
3
and hoaxes are forms of disinformation,
among others.
• Misinformation refers to false or misleading information that is shared
unknowingly and is not intended to deliberately deceive, manipulate or
inflict harm on a person, social group, organisation or country (Ireton
and Posetti, 2018
[13]
). Importantly, the spreader does not create or
fabricate the initial misinformation content.
• Contextual deception refers to the use of true but not necessarily
related information to frame an event, issue or individual (e.g. a headline
that does not match the corresponding article), or the misrepresentation
of facts to support one’s narrative (e.g. to deliberately delete
information that is essential context to understanding the original
meaning). While the facts used are true (unlike disinformation) and
unfabricated (unlike misinformation), the way in which they are used is
disingenuous and with the intent to manipulate people or cause harm.
• Propaganda
4
refers to the activity or content adopted and propagated
by governments, private firms, non-profits, and individuals to manage
collective attitudes, values, narratives, and opinions (EAVI, 2017
[30]
).
2
Fake news refers to false information that is "purposefully crafted, sensational, emotionally
charged, misleading or totally fabricated information that mimics the form of mainstream news"
(Zimdars and McLeod, 2020
[89]
). Fake news can be wholly fabricated or a mix of fact and fiction.
3
Deepfakes are synthetic media applications (e.g. videos or sound recordings) that alter a
person’s appearance or voice in an attempt to deceive viewers or listeners that what they are
seeing or hearing is real (Somers, 2020
[86]
). Like fake news, deepfakes can be a mixture of real and
unreal elements or completely fabricated.
4
Hate speech and terrorist and violent extremist content (TVEC) would also be considered
propaganda in a broader sense, but it is not considered in the context of this note which deals
exclusively with content that contains at least one untrue element or which is misleading.
| 9
DISENTANGLING UNTRUTHS ONLINE: CREATORS, SPREADERS AND HOW TO STOP THEM © OECD 2022
While propaganda can contain both true and untrue elements, it is often
used to appeal to an individual’s or social group's sentiments and
emotions rather than being informative (Neale, 1977
[31]
).
• Satire is defined as language, film or other works of art that use humour
and exaggeration to critique people or ideas, often as a form of social or
political commentary (OED, 2013
[32]
). Satire is an important form of
social and political criticism, using humour and wit to draw attention to
issues in society, and when satire is first published, the viewer often
recognises the content as satire in part because of where and how they
view it (e.g. directly from a satirical newspaper). However, as the
content is shared and re-shared, this connection is sometimes lost
intentionally (or not) by the spreader, leading new viewers to
misunderstand the original meaning (Wardle, 2019
[33]
).
These definitions support a typology of false and misleading content that can
be differentiated along two axes: 1) The intent (or not) of the information
disseminator (spreader) to cause harm and 2) the degree of fabrication (if any)
by the creator of the information content (e.g. altering photos, writing untrue
article, creating synthetic videos) (Figure 1). This distinction is important insofar
that some types of false and misleading content are not deliberately created
with a view to deceive (e.g. satire) or they are not intentionally spread with a
view to inflict harm (e.g. misinformation).
Figure 1. A typology of untruths online
Source: Authors.
From a policy perspective, it is important to differentiate the untrue content
creators from the spreaders. There may be policies better suited to addressing
10 |
DISENTANGLING UNTRUTHS ONLINE: CREATORS, SPREADERS AND HOW TO STOP THEM © OECD 2022
false content creators than spreaders, particularly those that disseminate
falsehoods unknowingly (misinformation) or as part of societal or political
commentary (satire). Before policy options can be devised, however, it is
important to consider the context in and modalities by which false information
diffuses online.
How are untruths spread online and what are the
consequences?
The issues around inaccurate and misleading information online have emerged
in tandem with the Internet’s rise as a major news source. The share of people
in the European Union (EU) who read online newspapers and news magazines
nearly doubled in 10 years, with 65% of individuals aged 16 to 75 consuming
news online in 2020 (Shearer, 2021[34]). Likewise, 86% of adult Americans
access news on a digital device and it is the preferred medium for half of
Americans (Shearer and Mitchell, 2021[35]). The Internet is also an important
source of health advice. In 2020, 49% of men and 60% of women aged 16-75
sought health information online in the EU.
Mainstream and traditional media are one source of untruths, and once such
content is disseminated, even if articles and posts are amended, the damage
has already been done in many cases (Moschella, 2022[36]). Research shows
that national figures, including politicians, are also a source of false and
inaccurate information, and when such views are reported by mainstream
news, some of this content circulates on social media without important
context (Newman et al., 2021[37]). However, untruths online are primarily
shared by individuals or organisations via online platforms such as social media
platforms, private messaging services, and search engines.
A particular characteristic of the digital age is that false information can be
more easily spread with digital technologies that were created for entirely
different purposes (i.e. to increase user engagement, monitor user interaction,
and deliver curated content without the aim or ambition of spreading
inaccurate content) (Ávila, Ortiz Freuler and Fagan, 2018[38]). The digital
technologies used to curate content are typically driven by algorithms or adopt
AI-based approaches, making it sometimes difficult to track the source of
misleading information, monitor its flow, and limit access to or block such
content (Ávila, Ortiz Freuler and Fagan, 2018[38]). It also makes transparency
about how these technologies work critical. Sophisticated disinformation
attacks use bots, trolls, and cyborgs that are specifically aimed at the rapid
dissemination of untruths (Paavola et al., 2016[39]).
Researchers from MIT Sloan showed that tweets containing false information
were 70% more likely to be retweeted than the truth, and that false and
misleading content reaches the first 1 500 people faster than true content
| 11
DISENTANGLING UNTRUTHS ONLINE: CREATORS, SPREADERS AND HOW TO STOP THEM © OECD 2022
(Brown, 2020[40]). The authors also show that using bots is not necessarily
required to spread untruths online; individuals are in fact more likely to share
such content themselves. Another study from researchers at New York
University and the Université Grenoble Alpes found that false information on
Facebook attracts six times more engagement than factual posts (Edelson et al.,
2021[41]).
The amplification of echo chambers and filter bubbles is a feature of the
proliferation of untruths in the digital age. While such phenomena exist in an
analogue world – for example, newspapers with a particular political inclination
– it is easier and faster to spread information of any kind on the Internet. User-
specific cookies log an individual’s revealed preferences, and memberships in
social networks and linkages to people or groups helps reinforce the type of
content that is seen by individuals. When users consistently interact with or
share specific content within their social networks that reinforces such beliefs,
echo chambers that confirm existing biases emerge and grow (Karsten and
West, 2016[42]). Recent document disclosures also reveal that targeting very
small groups with falsehoods, in a tactic called “narrowcasting”, has been
successful in the spread of highly-viral untrue content (Wall Street Journal,
2021[14]).
Recent modelling work of the spread of misinformation over social media
platforms suggests that filter bubbles can indeed help explain the spread of
misinformation (Acemoglu, Ozdaglar and Siderius, 2021[43]). This research
indicates that social media users are more likely to inspect articles that do not
conform to their prior beliefs, and that a user will be more hesitant to share an
article that does not conform to the views of those in his or her sharing
network. According to this model, misinformation spreads when users decide
to share an article without inspecting it, and users tend to share articles with
others with similar beliefs (e.g. filter bubbles).
Regardless of the veracity of specific content, online platforms curate the news
feed of users based on their past engagement with similar posts and
preferences to engage with particular topics, thereby creating and facilitating
communication within echo chambers and reinforcing filter bubbles. In such
situations, it is not technically feasible to monitor content in real time and hold
users liable. As a result, some online platforms have adopted transparency and
accountability measures in the form of community codes of conduct or content
guidelines that restrict or prohibit specific forms of speech such as hate speech,
obscene content, and misinformation and disinformation. Such transparency
and accountability measures are important and have been implemented at a
global level, but thought should be given as to whether national initiatives may
also add value (e.g. the DIGI code of conduct that applies to Australia and online
platforms, see Annex). Such initiatives may be particularly useful to ensure that
culture and language are appropriately taken into account.
12 |
DISENTANGLING UNTRUTHS ONLINE: CREATORS, SPREADERS AND HOW TO STOP THEM © OECD 2022
Though intermediary liability laws allow for take down of content when such
content is reported by users or ordered by the government or courts, creators
and spreaders of disinformation are often not held liable for their actions. More
recently, several countries have either enacted (e.g., Singapore, Malaysia) or
proposed disinformation laws (e.g., the United States, EU, and Korea) that call
for imposing criminal sanctions on agents of disinformation in addition to take
down of the false content from online platforms (Yadav et al., 2021[44]).
Though the constitutional contours of the right to freedom of speech and
expression vary across jurisdictions, human rights advocates have raised
concerns regarding such laws on the basis that they unreasonably restrict free
speech and disregard the normative considerations for restricting free speech,
namely the principles of 1) legality; 2) necessity; and 3) proportionality.
However, some academics have also argued that criminalisation of untrue
content could potentially curb the spread of false information online owing to
their deterrent effect (Helm and Nasu, 2021[45]). However, it is sometimes
difficult to come to an agreed upon definition of “truth”, a question that has
perplexed philosophers since the times of Aristotle (Blackburn, 2020[46]).
Events, topics, and beliefs are subject to a range of individual and idiosyncratic
factors that can lead to different interpretations, creating challenges
particularly for technology-driven solutions that may be less nuanced for
addressing untruths online. When such judgements are enshrined in law,
freedom of expression may be negatively impacted.
AI and big data analytics can be leveraged to fight untruths online to help
identify and remove false content online. In addition to the enhanced accuracy
with which AI can detect false information or recognise disinformation tactics
deployed through bots and deepfakes (Marcellino et al., 2020[47]), AI solutions
are more cost-effective because they reduce the time and human resources
required for detecting and removing false content. However, at the same time,
the effective use of AI for countering untruths online depends on large volumes
of data as well as supervised learning without which such tools run the risk of
false positives and human biases (Woolley, 2020[48]).
There is a need to measure and evaluate the extent to which falsehoods are
circulating online, as well as assess people’s susceptibility to encountering and
engaging with false content online. This can help identify the root causes of
untruths and help people, firms and governments to develop measures to
prevent the spread of untruths and ensure the protection of fundamental
rights and access to accurate information on other important issues (e.g.
climate change).
| 13
DISENTANGLING UNTRUTHS ONLINE: CREATORS, SPREADERS AND HOW TO STOP THEM © OECD 2022
Surveying the evidence base of untruths online
While public discourse and policymakers have shown increasing interest in
fighting untruths online, measuring this phenomenon has made less progress.
This is in part because there is not a mutually agreed upon definition of the
range of phenomena used to describe untruths online (Figure 1). Moreover, for
the most part data on this topic is held privately and it is not systematically
gathered either by national statistical organisations or by other entities.
However, there is some data that can be pieced together to begin to show the
extent of untruths online.
Individuals’ perceptions of exposure to false and misleading
content online
Social media users often share false content unintentionally because they
believe it, which is why the problem of inaccurate and misleading information
is so widespread. A CIGI–IPSOS poll (CIGI-Ipsos, 2019
[49]
) that surveyed 25 000
respondents in over 25 economies found that 86% of people around the world
reported that they have been exposed to fake news, and 86% of them initially
believed the false news at least once. As disinformation usually relies on highly
emotional content that provokes shock or anger, those false social media posts
draw more attention.
According to a Eurobarometer survey, 37% of EU respondents reported that
they were exposed to fake news every day or almost every day, while four in
five respondents indicated that they were exposed at least several times a
month. Moreover, 85% perceived fake news as a problem in their country. In
11 emerging economies, research suggests that between 44% and 78% of social
media and messaging app users indicated that they see occasionally or
frequently articles or content that seems obviously false or untrue (Silver,
2019
[50]
). However, such surveys are inherently conceived with self-reporting
bias and, in the context of “fake news”, they measure more respondents
awareness and perception rather than the real extent of the misinformation
problem.
In a 2018 survey by Pew Research, more Americans reported that made-up
news is a bigger problem than climate change or racism (Pew Research Center,
2019
[51]
). Most Americans indicated that they had come across inaccurate news,
and over one-third said that their preferred news source reported “made-up
information intended to mislead the public”. At the end of April 2020, 63% of
American adults indicated that they had seen some or a lot of news about
COVID-19 that seemed entirely made-up (Pew Research Center, 2019
[51]
). A
survey conducted in May 2020 among Japanese daily Internet users suggested
that 72% of respondents reported that they saw or heard at least one piece of
false or misleading content about COVID-19, and almost 36% had shared such
14 |
DISENTANGLING UNTRUTHS ONLINE: CREATORS, SPREADERS AND HOW TO STOP THEM © OECD 2022
information with others (Japanese Ministry of Internal Affairs and
Communications, 2020
[52]
).
In New Zealand, survey results of 2 301 people between February and March
2021 show that 75% of respondents considered misinformation as an "urgent
and serious threat to New Zealand society", and they indicated that the
Internet was an important vehicle for disseminating misinformation (Talbot
and Nusiebah, 2021
[53]
). Over 80% indicated that misinformation is becoming
more prevalent, and almost 60% reported exposure to misinformation in the
past six months. The report also estimates that half of all Kiwis held at least
one belief associated with misinformation, with as many as 19% of respondents
holding three or more such beliefs.
Trends in untruths related to elections and democratic processes
and institutions
Much media attention on inaccurate information online has focused on content
related to elections and democratic processes or institutions. Researchers from
the University of Oxford have been monitoring governments and political party
actors engaging in manipulation of public opinion on social media annually
between 2015 and 2020 (Bradshaw, Howard and Bailey, 2021
[54]
). In 2020, they
found evidence of the use of social medial for political disinformation and
propaganda in 81 countries, up from 70 in 2019. While almost all countries rely
on accounts managed by humans, automated bots were also used in 57
countries. Since 2018, they also identified more than 65 “influence firms”
providing political communication services (sometimes called “computational
propaganda”) to State actors , and the activities of these firms spread from 25
countries in 2019 to 48 countries in 2020 (Bradshaw, Howard and Bailey,
2021
[54]
). Given the extent and increasing reach of such content, it is perhaps
unsurprising that in a 2018 Flash Eurobarometer Survey, 83% of the EU
respondents agreed that fake news is a problem for democracy in general (EC,
2018
[55]
).
Some co-ordinated operations to manipulate public opinion have been taken
down by Facebook and Twitter. Between 2017 and July 2021, Facebook
(Facebook, 2021
[56]
) removed 180 networks engaged in influence operations or
what the company calls “coordinated inauthentic behaviour”. Overall, more
than 60 thousand assets (e.g. Facebook accounts, pages and groups as well as
Instagram accounts) were deleted. Around half of the networks were engaged
in domestic interference, one-third in foreign interference, and the remaining
groups did both. While identifying the country origin of false and inaccurate
content can be challenging (i.e. spreaders may use a VPN to conceal their
location), between 2017 and 2020 Facebook reported that the largest number
of networks originated in the Russian Federation (27), followed by Iran (23),
Myanmar (9), the United States, (9) and Ukraine (8). Foreign influence
| 15
DISENTANGLING UNTRUTHS ONLINE: CREATORS, SPREADERS AND HOW TO STOP THEM © OECD 2022
operations targeted most often the United States (26), Ukraine (11) and the
United Kingdom (11).
Researchers followed a sample of Internet users in the United States with their
prior consent to understand their online behaviour and exposure to
disinformation (Guess, Nyhan and Reifler, 2018
[57]
). They analysed web traffic
of 2 525 Americans during the weeks preceding the 2016 United States
presidential election and estimated that 27.4% of adult Americans visited an
article from an unreliable news site in that period. However, the articles
containing inaccurate or misleading content represented only 2.6% of all
articles read on news websites focusing on national and global politics.
Trends in inaccurate content related to public health
The outbreak of the COVID-19 pandemic brought about a surge of false and
misleading information worldwide. The Center for Countering Digital Hate
(CCDH), a non-profit organisation, focuses on spreaders of disinformation in
the context of the COVID-19 pandemic. Through their analysis of over 800 000
pieces of anti-vaccine content from Facebook and Twitter between February
and March 2021, they identified 12 people responsible for 65% of all identified
anti-vaccine posts (Center for Countering Digital Hate, 2021
[58]
). Following the
publication of this report, Facebook closed some of the accounts linked to this
group, which resulted in a loss of 5.8 million followers out of their total of 14.2
million followers (O’Sullivan, 2021
[59]
).
In total, the CCDH tracked 425 anti-vaccine accounts with 59.2 million followers
across the platforms. While the report was criticised by some for
overestimating the importance of these super-spreaders by not taking into
account anti-vaccine accounts that were already removed by the company,
another study confirmed that most interactions about COVID-19 were
generated by a very small group of users. Indeed, researchers from Italy
analysed 200 million interactions on Twitter related to the pandemic and
concluded that 0.1% of users account for up to 45% of activities and 10% of
the news that is shared (Sacco et al., 2021
[60]
).
Another study on misinformation analysed tweets with hashtags related to
COVID-19 posted before March 2020 (Kouzy et al., 2020
[61]
). Out of the 673
tweets identified, about 25% contained misinformation and another 17%
contained information that could not be verified. Another study showed that
exposure to anti-vaccine misinformation decreased the share of people who
indicated that they would definitely get a COVID-19 vaccine by 6.2 percentage
points in the United Kingdom and by 6.4 percentage points in United States
compared to the group that was exposed to factual information (Loomba et al.,
2021
[62]
).
Before the COVID-19 pandemic, social media companies were often hesitant to
moderate content posted on their websites, in part because they did not want
16 |
DISENTANGLING UNTRUTHS ONLINE: CREATORS, SPREADERS AND HOW TO STOP THEM © OECD 2022
to limit free speech. But with the onset of the COVID-19 pandemic and the
ensuing disinformation that followed, in March 2020 a group of seven
platforms
5
proactively published a joint statement in which they committed to
combat fraud and misinformation about the COVID-19 virus (Statt, 2020
[63]
).
These platforms are also signatories of the EU Code of Practice on
Disinformation and the Australian Code of Practice on Disinformation and
Misinformation through which they commit to share progress reports.
Since the beginning of the pandemic and until the end of June 2021, 20 million
pieces of content as well as 3 000 accounts, pages, and groups were removed
from Facebook and Instagram for violating COVID-19 policies (Rosen, 2021
[64]
).
In June 2021, deleted posts from the EU represented 11% of all deleted
content, which is slightly more than in the previous months. Between March
and December 2020, 110 000 pieces of content were removed in Australia
which represents around 0.8% of all content taken down during that period
(Facebook, 2021
[65]
). Since the start of the COVID-19 pandemic, Twitter
reported suspending around 2 000 accounts, and it removed 50 000 pieces of
content between January 2020 and August 2021 (Twitter, 2021
[66]
). To put the
latter number in context, there are currently close to 10 000 tweets shared
every second or around 25 billion a month, according to an estimate from
Internet Live Stats (Internet Live Stats, 2021
[67]
) and in line with the figure
shared by Twitter in 2013 (Twitter, 2013
[68]
).
Language plays an important role in content moderation
Fact-checking organisations have proliferated around the world and they
emerged as one of the key actors during the COVID-19 pandemic. Social media
platforms and researchers collaborate with fact-checking organisations to
identify untruths online. Facebook marks the content that they find false or
inaccurate with special warnings and downgrade it in the recommendation
algorithms. So far, 190 million posts on COVID were labelled with a warning.
For context, while Facebook does not share the global number of new posts,
the platform had over 1.9 billion daily active users in June 2021.
In April 2020, an analysis of over 100 pieces of misinformation content that
were flagged by fact-checkers to assess the effectiveness of Facebook’s
methods (AVAAZ, 2020
[69]
). They found that 68-70% of content in Italian and
Spanish were not labelled with a warning, while the problem concerned only
29% of English-speaking content. Likewise, in a June 2021 report Twitter
indicated that their machine-learning model that will be used to identify
content violating COVID-19 policies will be trained on English-language content
first (EC, 2021
[70]
). Other languages will only follow later.
5
Facebook, Google, LinkedIn, Microsoft, Reddit, Twitter, and YouTube.
| 17
DISENTANGLING UNTRUTHS ONLINE: CREATORS, SPREADERS AND HOW TO STOP THEM © OECD 2022
A moderation of two viral videos shows how non-English language content is
treated differently than other languages. A 26-minute conspiracy video
entitled “Plandemic the Hidden Agenda behind Covid-19” was published on
various platforms on 4 May 2020. According to Digital Trends, it reached 1.8
million views just on Facebook before being quickly deleted across platforms
(Gebel, 2021
[71]
). In November 2020, another conspiracy video, “Hold up”, was
released in French. It reached three million views in five days (Kayali, 2021
[72]
).
As reported by Politico Europe and the EU DisinfoLab, six months later the
video was still available on Facebook and YouTube. While it was initially
removed, the video was later republished. Users who would look for
“Plandemic documentary” were led to a COVID-19 information centre, while
users who would look for “Hold up documentary” would find new uploads of
the video, including one that reached over one million views on YouTube.
Going forward, the development of tools in multiple languages will be an
important part of the fight against untruths online. For example, the Spanish-
language Chequeabot tool (see Annex), which was developed by professional
fact-checkers, has been helping to fight untruths in Argentina. Other
approaches include co-operative arrangements between fact-checking entities
and media partners (e.g. FactCheck Initiative Japan, see Annex) to support the
identification of untruths online. Another innovative approach – crowdsourced
fact-checking (e.g. the Birdwatch programme, see Annex) – could also support
the moderation of non-English content in the future.
Approaches to fighting untruths online and mitigating
their negative effects
Tackling untrue content online requires a multistakeholder approach where
people, firms, and governments all play an active role in identifying and
removing inaccurate content on the Internet, and all actors exercise judgment
before sharing information online. It is likewise important to promote
transparency, and to create an enabling environment for an independent,
diverse, local and public service media to thrive, and to empower public
communicators to help in the fight against untruths online (OECD, 2021
[73]
).
Given the global reach of online platforms, a global approach is needed at least
among “like minded” countries, although this will not be easy due to differences
in culture, history and legal frameworks.
A better understanding of a range of complex and intertwined issues about
untruths online is urgent to develop “best practice” policies to address this
important problem, but in the meantime, concrete steps can be taken to begin
the fight. This note argues that five broad steps can help fight untruths online:
18 |
DISENTANGLING UNTRUTHS ONLINE: CREATORS, SPREADERS AND HOW TO STOP THEM © OECD 2022
1. Promote digital literacy initiatives,
2. Develop content moderation policies in a multistakeholder process and
with independent oversight,
3. Integrate humans and technology in the fight against untruths online,
4. Increase transparency in spending on political advertisements online,
and
5. Design a measurement agenda to improve the evidence base and inform
more targeted policies to stop the creators and spreaders of untruths.
While none of these steps by themselves can effectively stop the spread of
inaccurate and misleading content online, in combination they can go a long
way toward protecting fundamental and other rights and to mitigating the
negative effects from such content.
Create awareness about untruths online by promoting digital
media literacy
People are generally not very good at identifying false and inaccurate
information, with research showing that on average people are able to identify
47% of lies as deceptive and 61% of truths as non-deceptive (Bond and
DePaulo, 2006
[74]
). As a result, an important way to stop untruths online is to
promote digital media literacy among adults and children, including in schools
(Khan, 2021
[28]
). It is also a very practical approach. Instead of trying to chase
each piece of untrue content, which is impossible, digital literacy initiatives
protect people by giving them the tools to distinguish false and misleading
information and to disregard or ignore it.
Initiatives that have been implemented in countries worldwide by
governments, schools, universities, online platforms, and non-profits help
individuals to better assess and verify the accuracy of information online.
Digital media literacy initiatives tend to focus on developing cognitive, critical,
and technical skills that help discern fact from fiction, and enable meaningful
participation in public interactions, discussions and debates (e.g. the Bad News
and Go Viral! games, see Annex). Online platforms including Facebook, Google,
and Twitter have launched extensive digital media literacy initiatives in
partnership with international organisations, governments, and fact-checking
organisations to raise awareness and educate people about how to spot
potential false and misleading content.
While these initiatives provide resources to those accessing the Internet, scaling
such initiatives to ensure widespread training across diverse demographics has
been challenging. Such initiatives typically engage with only a tiny fraction of
the population, such as politicians, journalists and school teachers that are
concentrated in large cities (ERGA, 2020
[75]
). For example, research conducted
| 19
DISENTANGLING UNTRUTHS ONLINE: CREATORS, SPREADERS AND HOW TO STOP THEM © OECD 2022
in the United Kingdom indicates that digital literacy inequalities correspond
with other key elements of economic, social, and cultural inequality (Helsper,
2016
[76]
), with people of different ages possessing varied levels of digital media
literacy. Efforts towards imparting digital media literacy would benefit from
adopting a broader and more inclusive approach.
Develop and implement online platform content moderation
policies in a multistakeholder process and with independent
oversight
In light of the COVID-19 pandemic as well as concerns around election integrity
in some countries, several online platforms have revised and expanded existing
content moderation policies to include false and misleading content. While such
activities move in the right direction, some online platforms implement content
moderation policies that have been developed without public input and
enforced with limited clarity (Kaye, 2018
[77]
). By not engaging in a
multistakeholder process, these actions may raise the risk that such policies are
not compliant with existing laws, including on free speech, notably if policies
do not require independent oversight or transparency in the decisions leading
up to the take down of problematic content.
Towards this end, the EU’s forthcoming Digital Services Act calls on online
platforms to be more transparent and accountable in their content moderation
decisions through periodic reporting obligations. In its recent Joint
Communication on Disinformation, the European Commission announced that
online platforms that are required to comply with the Code of Practice on
Misinformation would need to provide monthly reports on how they are
dealing with misinformation.
While online platforms make information relating to takedown decisions public
to enhance transparency, untruths may still go unchecked with problematic
content remaining on platforms despite being declared untrue. Creating
content moderation practices involving local stakeholders, including fact-
checking organisations and researchers, and setting up independent audits of
content moderation decisions could help make take-down decisions more
consistent and further improve online content moderation at large.
For example, Facebook’s Oversight Board, a governing body comprised of
members from a variety of cultural and professional backgrounds, reviews
content moderation decisions taken by Facebook. The Oversight Board’s aim is
to improve fairness and transparency around content and provide oversight
and accountability (Oversight Board, 2021
[78]
). The Oversight Board recently
rebuked Facebook for not being more forthcoming about how it exempts high-
profile users from its rules (the “cross check” program), and said that it is
drafting recommendations for how to overhaul the system (Schechner,
2021
[79]
). Oversight boards could potentially serve as a model for online
20 |
DISENTANGLING UNTRUTHS ONLINE: CREATORS, SPREADERS AND HOW TO STOP THEM © OECD 2022
platforms, if the Board is truly independent, it has thorough access to
information, and independent audits with well-trained auditors are carried out.
Integrate humans and technology in the fight against untruths
online
Existing approaches to reducing untruths online are often dependent on
manual fact-checking, content moderation and takedown, and quick responses
to attacks that involve human intervention and allow for a finer-grained
assessment of the degree of the accuracy of content. For example, PolitiFact’s
“Truth-o-Meter” includes six ratings
6
to assess the degree of veracity.
Collaborations between independent, domestic fact-checking entities and
platforms can further help identify untruths (e.g. DIGI in Australia and
FactCheck Initiative Japan, see Annex) and can also be useful to ensure that
cultural and linguistical considerations are taken into account.
While human understanding is essential to interpreting specific content in the
context of cultural sensitivities and belief or value systems, monitoring online
content in real time is a mammoth task that may not be feasible entirely
without technological assistance. Automation of certain content moderation
functions and developing technologies that embed such functions “by design”
could considerably enhance the efficacy of techniques used to prevent the
spread of untruths online, although such approaches often provide less nuance
on the degree of accuracy of content (i.e. content is usually identified as either
“true” or “false”).
Such approaches would also benefit from partnerships between local fact-
checking entities and online platforms to ensure cultural and linguistical biases
are addressed
7
. Advanced technologies such as automated fact checking
(Dulhanty et al., 2019
[80]
) or natural language processing and data mining
(Rahman, Chia and Gonzalez, 2021
[81]
) (Wang et al., 2018
[82]
) can be leveraged
to detect producers of inaccurate information and prevent sophisticated
disinformation attacks, although the spreaders of untruths have found ways to
circumvent such approaches (e.g. through the use of images rather than words).
In this regard, transparent use of digital technologies by online platforms to
identify and remove untrue content can improve the dissemination of accurate
information.
At the same time, the technical limitations of AI and other technologies (e.g.
potential bias) point towards the need to adopt a hybrid approach that
integrates both human intervention and technological tools in fighting untruths
online. In such an approach, digital tools can help to monitor and detect
6
1) True, 2) mostly true, 3) half true, 4) mostly false, 5) false and 6) pants-on-fire.
7
Algorithms trained mainly on American English have been shown to underperform on content
using British English (Waterson and Milmo, 2021
[87]
).
| 21
DISENTANGLING UNTRUTHS ONLINE: CREATORS, SPREADERS AND HOW TO STOP THEM © OECD 2022
inaccurate information online, and human expertise and value judgment can be
used to determine the extent to which untruths are likely to harm the public.
Approaches that marry technology-oriented solutions and human judgement
may be best suited to ensure both efficient identification of problematic
content and potential takedown after careful human deliberation, taking into
account all relevant principles, rights and laws such as on free speech.
Increase transparency in spending on political advertisements
online
With political parties and candidates spending large amounts of money on paid
political advertising and content through different channels, online platforms
have become a vehicle for disseminating untrue and misleading content. Such
untruths often mislead voters and compromise electoral outcomes,
undermining democratic elections and civic processes. In an effort to address
this issue, online platforms now sometimes publish periodic transparency
reports that disclose the identities of political advertisers as well as the
amounts spent on such advertisements or content. However, such laws do not
require disclosures by advertising agencies, consultancies or political
organisations that spend money towards political advertisements and content
on behalf of political parties and candidates.
To ensure enhanced transparency in online political advertising, including
campaign spending, mandating political parties to disclose monies spent
towards paid digital advertisements and content on a regular basis could
potentially mitigate the harms caused by political disinformation and
incentivise political parties and candidates to publish accurate and truthful
information (Dunčikaitė, Žemgulytė and Valladares, 2021
[83]
). Another step in
this direction that has been adopted by the EU is to restrict political
advertisements and content to “issue-based advertising” that focuses on clear
distinguishability from editorial content (EC, 2019
[29]
). Such issue-based
advertising allows voters to make more informed decisions based on their own
judgment and reasoning.
Design a measurement agenda to improve the evidence base
Without a solid evidence base, it is difficult to develop well-targeted policies
for fighting untruths online. Ideally, indicators along the following dimensions
would help shed important light on the scale, content and reach of untruths
online:
• Who (age, gender, language, education and income levels) is spreading
false information online?
• What types of false information online (e.g. health, elections, conspiracy
theories) are most prevalent?
22 |
DISENTANGLING UNTRUTHS ONLINE: CREATORS, SPREADERS AND HOW TO STOP THEM © OECD 2022
• Where does false information originate from and where is it
disseminated?
• Why do people spread falsehoods online (i.e. what are their aims and
ambitions, or is it accidental)?
• How and through which vehicles are untruths spread in the digital age?
While data on the dimensions noted above will be challenging to come by, not
least because much of the data needed is proprietary and dispersed among
private firms, it is nonetheless important that a co-ordinated measurement
agenda be developed and implemented in a cross-country comparable manner
and in partnership with the private sector.
Conclusion
Overall, while digital technologies are in and of themselves neutral, they can
nonetheless be used intentionally and unintentionally to spread inaccurate and
misleading information, thereby impacting individuals, social groups, and
society in a variety of ways. While responses to untruths by individuals,
governments, and firms vary across countries (see Annex), a multistakeholder
approach is needed to reduce the spread of untruths online. Information
producers, users of online platforms, and online platforms themselves all have
an important role to play in stopping the creators and spreaders of untruths
online and ensuring transparency and accountability.
| 23
DISENTANGLING UNTRUTHS ONLINE: CREATORS, SPREADERS AND HOW TO STOP THEM © OECD 2022
Annex. A Selection of innovative approaches to
fighting untruths online and mitigating their
negative effects
Create awareness about untruths online by improving
digital media literacy
Bad News and Go Viral! games
Responsible entity: University of Cambridge’s Social Decision-Making Lab in
collaboration with the UK Cabinet Office
Description: The Bad News and Go Viral! Games were developed in response
to research from the University of Cambridge that found that educating people
on the techniques used to spread false and misleading content on social media
increases their ability to identify and disregard similar content in the future.
Bad News was launched in 2018 and it has been played over one million times;
Go Viral! is a shorter game that was launched in 2020. It focusses on falsehoods
related to COVID-19 and it is easier to adapt for different languages and
cultures; there are currently French and German versions.
Read more: https://www.goviralgame.com; https://www.getbadnews.com
;
https://www.cam.ac.uk/stories/goviral.
Be Internet Awesome initiative
Responsible entity: Google
Description: Google's media literacy initiative, Be Internet Awesome (BIT),
aims to teach children how to spot untruths online and be safe, confident
explorers of the online world. BIT is a multifaceted programme that includes
an interactive, web-based game ‘Interland’, and an educational curriculum to
teach children about digital safety.
Read more: https://beinternetawesome.withgoogle.com/en_us.
24 |
DISENTANGLING UNTRUTHS ONLINE: CREATORS, SPREADERS AND HOW TO STOP THEM © OECD 2022
Check the Facts campaign
Responsible entity: Australian Associated Press (AAP) with the support of
Facebook
Description: The AAP’s “Check the Facts” digital literacy campaign aims to
raise awareness about how to recognise reliable information based on
techniques used in professional fact checking organisations. It includes videos
and other resources to help spot untrue content by considering the source,
whether the source is trustworthy, and the specific claims that are being made.
Read more: https://www.aap.com.au/factcheck-resources/
.
Media and digital literacy programmes in Finnish schools
Responsible entity: National Audio-Visual Institute (NAVI) and the Finnish
Ministry of Education and Culture, in collaboration with schools media
professionals and fact checking organisations
Description: In response to fake news campaigns focussing on immigration,
EU, Finland and the North Atlantic Treaty Organisation (NATO) members
recognised the need to strengthen the population’s resilience to digital
untruths, and instituted a cross-sector approach to improve media literacy
within the country, with a focus on children. Media and digital literacy skills are
embedded across Finland’s national school curriculum and implemented by the
NAVI and the Finnish Ministry of Education and Culture in collaboration with
media education professionals as well as a non-profit fact checking
organisation, Faktabari, that provides fact checking and media literacy
materials for schools.
Read more: https://medialukutaitosuomessa.fi/mediaeducationpolicy.pdf
.
Section 51206.4 of the California Education Code
Responsible entity: State Government of California, United States
Description: In 2018, the state of California passed a bill requiring the State’s
Department of Education website to list resources and instructional materials on
media literacy, including professional development programmes for teachers.
Specifically, the bill aims to empower students to distinguish advertisements from
news stories and make informed decisions online. The legislative intent behind this
law was based on a 2016 Stanford University study that indicated that 80% of
middle school students did not recognise an advertisement that was masquerading
as a news story despite being labelled as "sponsored content" (Breakstone et al.,
2019
[84]
).
Read more:
https://leginfo.legislature.ca.gov/faces/billTextClient.xhtml?bill_id=201720180
SB830.
| 25
DISENTANGLING UNTRUTHS ONLINE: CREATORS, SPREADERS AND HOW TO STOP THEM © OECD 2022
Develop and implement online platform content
moderation policies in a multistakeholder process and
with independent oversight
Birdwatch
Responsible entity: Twitter
Description: Launched in early 2021, Twitter’s pilot programme, Birdwatch,
aims to combat misinformation by adding fact-checking notes written by
crowd-sourced volunteers. Initially restricted to a group of enrolled
contributors, the flags on potentially misleading Tweets are now shown to
Twitter test users in the United States. Notes must be approved by
contributors who have shown to have diverging views in the past and can be
further evaluated by users. A recent study by MIT researchers showed that
ratings from a small, politically balanced group of regular people correlated
with professional fact-checkers (Allen et al., 2021
[85]
).
Read more:
https://blog.twitter.com/en_us/topics/product/2021/introducing-
birdwatch-a-community-based-approach-to-misinformation;
https://blog.twitter.com/en_us/topics/company/2022/building-a-better-
birdwatch; https://www.science.org/doi/10.1126/sciadv.abf4393.
DIGI Code of Practice
Responsible Entity: Digital Industry Group Inc. (DIGI)
Description: On 22 February 2021, DIGI launched a code of practice that
commits a diverse set of technology companies to reducing the risk of harm
from online disinformation and misinformation in Australia. The Australian
Code of Practice on Disinformation and Misinformation has been adopted by
Adobe, Apple, Facebook, Google, Microsoft, Redbubble, TikTok, and Twitter. All
signatories commit to protect Australians by providing appropriate safeguards
against harm from online disinformation and misinformation, and to adopting
a range of scalable measures that reduce its spread and visibility. Participating
companies also commit to releasing an annual transparency report about their
efforts under the code, which will help improve understanding of online
misinformation and disinformation in Australia over time. The Code was
developed in response to the Australian Government policy announced in
December 2019, where the digital industry was asked to develop a voluntary
code of practice on disinformation, drawing on learnings from a similar code in
the EU.
Read more:
https://digi.org.au/wp-content/uploads/2021/10/Australian-
Code-of-Practice-on-Disinformation-and-Misinformation-FINAL-WORD-
UPDATED-OCTOBER-11-2021.pdf.
26 |
DISENTANGLING UNTRUTHS ONLINE: CREATORS, SPREADERS AND HOW TO STOP THEM © OECD 2022
FIJ Fact-Checking Guidelines
Responsible entity: FactCheck Initiative Japan (FIJ)
Description: FIJ is a Tokyo-based non-profit organisation for the promotion
of Japanese fact checking aimed at protecting society from untruths online. FIJ
supports and co-operates with media partners that publish fact-checking
articles according to FIJ Fact-Checking Guidelines that are based on the IFCN
Code of Principles. In April 2020, FIJ worked with Yahoo! Japan – one of the
largest sources of mainstream news in Japan – to create an English-version of
their website to reach more people with COVID-19 related fact-checked
information.
Read more: https://en.fij.info/about/
.
International Fact-Checking Network (IFCN)
Responsible entity: The Poytner Institute
Description: In 2015, the Poynter Institute – a non-profit organisation that
promotes freedom of expression, civil dialogue and truthful journalism –
established the IFCN to bring together the international community of fact
checkers to tackle false and misleading information. At the beginning of 2022,
the IFCN had 108 active signatories to its Code of Principles from over 50
countries. The IFCN also provides training programmes and resources (including
some in Spanish) to help develop the skills to identify untrue and misleading
content. Many online platforms are collaborating with the IFCN and fact
checkers, for example by applying fact-checked labels.
Read more: https://www.poynter.org/ifcn/
.
| 27
DISENTANGLING UNTRUTHS ONLINE: CREATORS, SPREADERS AND HOW TO STOP THEM © OECD 2022
Network Enforcement Act
Responsible entity: German government
Description: The Netzwerkdurchsetzungsgesetz, or Network Enforcement Act
(the Act), went into force in 2017 and was subsequently amended in June 2021.
Its aim is to address illegal content that meets the criteria of the Criminal Code
(e.g. hate speech). It requires online platforms to remove content found to be
offensive or “clearly illegal” within 24 hours after receiving a user complaint. If
the illegality of the content is not obvious, the online platform has seven days
to investigate and delete it. The Act requires transparency reporting, and fines
for non-compliance can be assessed up to a ceiling of 50 million euros. The
Federal Office of Justice has the power to issue fines for noncompliance and to
supervise compliance with the Act.
Read more:
https://www.bmjv.de/SharedDocs/Gesetzgebungsverfahren/Dokumente/NetzD
G_engl.pdf;jsessionid=D7F76C9A3ED468A2F3BEECBE17F330BA.2_cid324?__blob
=publicationFile&v=2.
Oversight Board
Responsible entity: Facebook
Description: Facebook’s Oversight Board is a governing body comprised of
members from a variety of cultural and professional backgrounds that reviews
content moderation decisions taken by Facebook and Instagram. The Board
aims to promote free expression by making principled, independent decisions
by issuing recommendations on the relevant content policies of the online
platforms.
Read more: https://oversightboard.com/.
Twitter’s guidance on content related to COVID-19, elections and
other civic processes
Responsible entity: Twitter
Description: In March 2020, Twitter expanded its content moderation policy
to address content that goes against guidance from global and local public
health authorities on COVID-19 protocols. This includes sharing content that
may mislead people about the nature of the COVID-19 virus’ efficacy and/or
the safety of preventative measures, treatments, or other precautions to
mitigate or treat the disease. Twitter’s content moderation policy also
addresses untrue content related to official regulations, restrictions, or
exemptions pertaining to health advisories, the prevalence of the virus, or the
risk of infection or death associated with COVID-19. Content that could
potentially endanger public health is either labelled as demonstrably false or
28 |
DISENTANGLING UNTRUTHS ONLINE: CREATORS, SPREADERS AND HOW TO STOP THEM © OECD 2022
misleading, or removed from the platform. Similarly, Twitter also prohibits
sharing content that could potentially compromise or interfere with elections
and civic processes. Such content includes false or misleading information in
relation to procedures or circumstances around participation in a civic process
or content that seeks to confuse or manipulate voters using the platform.
Read more:
https://blog.twitter.com/en_us/topics/product/2020/updating-
our-approach-to-misleading-information; https://about.twitter.com/en/our-
priorities/civic-integrity.
TikTok’s Community Guidelines
Responsible entity: TikTok/Bytedance
Description: In response to global concern about the propagation of
disinformation and misinformation, TikTok updated its policies on misleading
content to provide further clarity on what is and is not allowed on TikTok.
TikTok has also added a policy that prohibits synthetic or manipulated content
(for example, deepfakes) that misleads users by distorting the truth of events
in a way that could lead to real world harm. Specifically, TikTok’s Community
Guidelines prohibit the sharing content that could cause harm to users or the
greater public, including content that misleads people about elections or other
civic processes, content distributed by disinformation campaigns, and untrue
health information.
Read more: https://www.tiktok.com/community-guidelines?lang=en
.
Integrate humans and technology to fight untruths
online
Chequeabot
Responsible Entity: Chequeado
Description: In January 2018, the Argentine foundation Chequeado released
Chequeabot, a bot tool that incorporates natural language processing and
machine learning to identify claims made in the media and matches them with
existing fact checks. Chequeado co-ordinated the development of Chequeabot
with both the IFCN and Full Fact, underscoring the tightly linked global fact-
checking community. Chequeabot is notable in that it is in Spanish and it was
developed by professional fact-checkers. While it is used for global issues (e.g.
the conflict in Ukraine), it is nonetheless strongly focussed on issues important
in Argentina (i.e. inflation and IMF programmes are a strong focus).
Read more: https://chequeado.com/tag/chequeabot/
.
| 29
DISENTANGLING UNTRUTHS ONLINE: CREATORS, SPREADERS AND HOW TO STOP THEM © OECD 2022
Full Fact’s AI-based fact checking tools
Responsible entity: Full Fact and Google
Description: In 2020, Google provided the non-profit Full Fact with 2 million
USD and seven technical experts from the Google.org Fellowship to help Full
Fact build AI-based tools to help fact checkers verify claims made by key
politicians, then group them by topic and match them with similar claims from
across press, social networks and even radio using speech-to-text technology.
These tools helped Full Fact process 1 000 times more content, detecting and
clustering over 100 000 claims per day. Importantly, the tools gave Full Fact’s
fact checkers more time to verifying facts rather than identifying which facts
to check. Using a machine learning BERT-based model, the technology now
works in four languages (English, French, Portuguese and Spanish).
Read more:
https://blog.google/documents/37/How_Google_Fights_Disinformation.pdf
;
https://blog.google/outreach-initiatives/google-org/fullfact-and-google-fight-
misinformation/.
Increase transparency in spending on political
advertisements online
Action Plan against Disinformation
Responsible entity: European Commission together with EU Member States
Description: As part of the Action Plan against Disinformation, the European
Commission has recommended to Member States to focus on promoting the
transparency of online political advertising, including campaign expenditure,
and it invited all political parties to respect transparency recommendations. On
19 October 2020, the European Commission presented its 2021 work
programme, which included as one of its priorities 'A New Push for European
Democracy'. Under this priority, the Commission announced its intention to
issue a proposal on greater transparency in paid political advertising.
Read more:
https://ec.europa.eu/info/sites/default/files/eu-communication-
disinformation-euco-05122018_en.pdf;
https://ec.europa.eu/info/strategy/priorities-2019-2024/new-push-european-
democracy_en.
30 |
DISENTANGLING UNTRUTHS ONLINE: CREATORS, SPREADERS AND HOW TO STOP THEM © OECD 2022
References
Acemoglu, D., A. Ozdaglar and J. Siderius (2021), “Misinformation: Strategic sharing,
homophily, and endogenous echo chambers”, NBER Working Paper No. 28884,
http://dx.doi.org/10.3386/w28884.
[43]
Allen, J. et al. (2021), “Scaling up fact-checking using the wisdom of crowds”, Science
Advances, https://doi.org/10.1126/sciadv.abf4393 (accessed on 22 March 2022).
[85]
AVAAZ (2020), “How Facebook can flatten the curve of the coronavirus infodemic”,
https://secure.avaaz.org/campaign/en/facebook_coronavirus_misinformation/
(accessed on 25 October 2021).
[69]
Ávila, R., J. Ortiz Freuler and C. Fagan (2018), “The invisible curation of content”, Web
Foundation,
http://webfoundation.org/docs/2018/04/WF_InvisibleCurationContent_Screen_AW.
pdf.
[38]
Barnes, J. (2022), “Russia Steps Up Propaganda War Amid Tensions With Ukraine”,
The New York Times, https://www.nytimes.com/2022/01/25/us/politics/russia-
ukraine-propaganda-disinformation.html.
[3]
Blackburn, S. (2020), “Truth”, https://www.britannica.com/topic/truth-philosophy-
and-logic.
[46]
Bond, C. and B. DePaulo (2006), “Accuracy of deception judgments”, Personality and
Social Psychology Review, Vol. Vol. 10, No. 3,
https://www.researchgate.net/publication/6927452_Accuracy_of_Deception_Judgm
ents.
[74]
Bradshaw, S., P. Howard and H. Bailey (2021), “Industrialized disinformation: 2020
global inventory of organized social media manipulation”, Oxford Institute,
https://demtech.oii.ox.ac.uk/research/posts/industrialized-disinformation/.
[54]
Breakstone, J. et al. (2019), “Students’ civic online reasoning: A national portrait”,
https://stacks.stanford.edu/file/gf151tb4868/Civic%20Online%20Reasoning%20N
ational%20Portrait.pdf.
[84]
Brown, S. (2020), “MIT Sloan Research About Social Media, Misinformation, and
Elections”, MIT, https://mitsloan.mit.edu/ideas-made-to-matter/mit-sloan-
research-about-social-media-misinformation-and-elections (accessed on
September 15 2021).
[40]
| 31
DISENTANGLING UNTRUTHS ONLINE: CREATORS, SPREADERS AND HOW TO STOP THEM © OECD 2022
Center for Countering Digital Hate (2021), “The disinformation dozen: Why platforms
must act on twelve leading online anti‐vaxxers”,
https://www.counterhate.com/disinformationdozen.
[58]
CIGI-IPSOS (2019), “CIGI IPSOS Global Survey”, Vol. 3,
https://www.cigionline.org/sites/default/files/documents/2019%20CIGI-
Ipsos%20Global%20Survey%20-
%20Part%203%20Social%20Media%2C%20Fake%20News%20%26%20Algorithms.
pdf (accessed on 5 August 2021).
[21]
CIGI-Ipsos (2019), Fake News: A Global Epidemic, https://www.ipsos.com/en-us/news-
polls/cigi-fake-news-global-epidemic (accessed on 15 September 2021).
[49]
Clegg, N. (2021), “Facebook’s Nick Clegg calls for bipartisan approach to break the
deadlock on internet regulation”, https://www.cnbc.com/2021/05/24/facebooks-
nick-clegg-a-bipartisan-approach-to-break-the-deadlock-on-internet-
regulation.html (accessed on 29 October 2021).
[18]
Colomina, C., H. Margalef and R. Youngs (2021), “The impact of disinformation on
democratic processes and human rights in the world”,
https://www.europarl.europa.eu/RegData/etudes/STUD/2021/653635/EXPO_STU(2
021)653635_EN.pdf.
[10]
Dulhanty, C. et al. (2019), “Taking a stance on fake news: Towards automatic
assessment via deep bidirectional transformer language models for stance
detection”, https://arxiv.org/pdf/1911.11951.pdf (accessed on 12 August 2021).
[80]
Dunčikaitė, I., D. Žemgulytė and J. Valladares (2021), “Paying for Views: Solving
transparency and accountability risks in online political advertising”,
https://images.transparencycdn.org/images/2021_Report_PayingForViews-
OnlinePoliticalAdvertising_English.pdf.
[83]
EAVI (2017), “Beyond Fake News”, https://eavi.eu/beyond-fake-news-10-types-
misleading-info/.
[30]
EC (2021), “Reports on June Actions - Fighting COVID-19 Disinformation Monitoring
Programme”, https://digital-strategy.ec.europa.eu/en/library/reports-june-actions-
fighting-covid-19-disinformation-monitoring-programme.
[70]
EC (2019), Code of Practice on Disinformation, European Commission, https://digital-
strategy.ec.europa.eu/en/policies/code-practice-disinformation.
[29]
EC (2018), “Flash Eurobarometer 464 (Fake news and disinformation online)”,
https://data.europa.eu/data/datasets/s2183_464_eng?locale=en.
[55]
32 |
DISENTANGLING UNTRUTHS ONLINE: CREATORS, SPREADERS AND HOW TO STOP THEM © OECD 2022
Edelson, L. et al. (2021), “Understanding engagement with U.S. (mis)information
news sources on Facebook”, IMC ’21: Proceedings of the 21st ACM Internet
Measurement Conference, https://dl.acm.org/doi/10.1145/3487552.3487859.
[41]
ERGA (2020), “Improving Media Literacy Campaigns on Disinformation”,
https://erga-online.eu/wp-content/uploads/2021/01/ERGA-SG2-Report-2020-
Improving-Media-Literacy-campaigns-on-disinformation.pdf (accessed on
23 August 2021).
[75]
Facebook (2021), “Coordinated inauthentic behavior”,
https://about.fb.com/news/tag/coordinated-inauthentic-behavior/ (accessed on
15 September 2021).
[56]
Facebook (2021), “Facebook response to the Australian disinformation and
misinformation industry code. May 2021”, https://australia.fb.com/wp-
content/uploads/sites/69/2021/05/Facebook-commitments-under-disinfo-and-
misinfo-code-final-1.pdf (accessed on 23 September 2021).
[65]
Frenkel, S. (2022), “TikTok Is Gripped by the Violence and Misinformation of Ukraine
War”, The New York Times,
https://www.nytimes.com/2022/03/05/technology/tiktok-ukraine-
misinformation.html.
[4]
Gebel, M. (2021), “Facebook takes down viral ‘Plandemic’ coronavirus conspiracy
video”, Digital Trends, https://www.digitaltrends.com/news/facebook-will-take-
down-viral-plandemic-coronavirus-conspiracy-video/ (accessed on
24 September 2021).
[71]
Green, T. (2020), “Americans are confident in tech companies to prevent misuse of
their platforms in the 2020 election”, https://www.pewresearch.org/fact-
tank/2020/09/09/few-americans-are-confident-in-tech-companies-to-prevent-
misuse-of-their-platforms-in-the-2020-election/ (accessed on 5 August 2021).
[22]
Guess, A., B. Nyhan and J. Reifler (2018), “Selective exposure to misinformation:
Evidence from the consumption of fake news during the 2016 US presidential
campaign”, European Research Council, Vol. 9/3, https://about.fb.com/wp-
content/uploads/2018/01/fake-news-2016.pdf.
[57]
Heins, M. (2013), “The brave new world of social media censorship”, Harvard Law
Review Forum, Vol. 127, pp. 325-330, https://harvardlawreview.org/2014/06/the-
brave-new-world-of-social-media-censorship/.
[27]
| 33
DISENTANGLING UNTRUTHS ONLINE: CREATORS, SPREADERS AND HOW TO STOP THEM © OECD 2022
Helm, R. and H. Nasu (2021), “Regulatory responses to fake news and freedom of
expression: Normative and empirical evaluation”, Human Rights Law Review,
Vol. 21/2, pp. 302-28, https://doi.org/10.1093/hrlr/ngaa060.
[45]
Helsper, E. (2016), “The social relativity of digital exclusion: applying relative
deprivation theory to digital inequalities”, Communication Theory, Vol. 27/3,
pp. 223-242, https://onlinelibrary.wiley.com/journal/14682885.
[76]
Henry, D. (2021), “Fake news: Majority of Kiwis think spreading misinformation
should be illegal”, The New Zealand Herald, https://www.nzherald.co.nz/nz/fake-
news-majority-of-kiwis-think-spreading-misinformation-should-be-
illegal/WC3Q45YGOB2IXQI6QGNFS4TEOU/ (accessed on 23 September 2021).
[17]
Internet Live Stats (2021), “Twitter usage statistics”,
https://www.internetlivestats.com/twitter-statistics/ (accessed on
26 October 2021).
[67]
Ireton, C. and J. Posetti (2018), Journalism, fake news & disinformation: Handbook
for journalism education and training, UNESCO,
https://unesdoc.unesco.org/ark:/48223/pf0000265552.
[13]
Japanese Ministry of Internal Affairs and Communications (2020), “Information
distribution survey on novel coronavirus infectious diseases”,
https://www.soumu.go.jp/main_content/000693295.pdf (accessed on
8 February 2022).
[52]
Jongh, D., B. Rofagha and L. Petrosova (2021), “Countering online vaccine
misinformation in the EU/EEA”,
https://www.ecdc.europa.eu/sites/default/files/documents/Countering-online-
vaccine-misinformation-in-the-EU-EEA.pdf.
[8]
Karsten, J. and D. West (2016), “Inside the social media echo chamber”, Brookings
Institution, https://www.brookings.edu/blog/techtank/2016/12/09/inside-the-
social-media-echo-chamber/.
[42]
Kayali, L. (2021), “A French coronavirus conspiracy video stayed on YouTube and
Facebook for months”, Politico Europe, https://www.politico.eu/article/french-
viral-covid-19-conspiracy-documentary-stayed-months-on-youtube-facebook/
(accessed on 23 September 2021).
[72]
Kaye, D. (2018), Report of the Special Rapporteur on the promotion and protection
of the right to freedom of opinion and expression, U.N. Human Rights Council,
https://digitallibrary.un.org/record/1631686/usage?ln=en.
[77]
34 |
DISENTANGLING UNTRUTHS ONLINE: CREATORS, SPREADERS AND HOW TO STOP THEM © OECD 2022
Khan, I. (2021), “Disinformation and freedom of opinon and expression”,
https://undocs.org/A/HRC/47/25 (accessed on 12 August 2021).
[28]
Kouzy, R. et al. (2020), “Coronavirus goes viral: Quantifying the COVID-19
misinformation epidemic on Twitter”, Cureus, Vol. 12/3,
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7152572/.
[61]
Lerer, L. (2021), “How Republican vaccine opposition got to this point”, New York
Times, https://www.nytimes.com/2021/07/17/us/politics/coronavirus-vaccines-
republicans.html.
[25]
Loomba, S. et al. (2021), “Measuring the impact of COVID-19 vaccine misinformation
on vaccination intent in the UK and USA”, Nature Human Behaviour, Vol. 5/3,
pp. 337–348, https://www.nature.com/articles/s41562-021-01056-1.
[62]
Marcellino, W. et al. (2020), “Human-machine detection of online-based malign
information”, RAND Corporation, https://doi.org/10.7249/RRA519-1.
[47]
Mitchell, A. and M. Walker (2021), “More Americans now say government should take
steps to restrict false information online than in 2018”, Pew Research Center,
https://www.pewresearch.org/fact-tank/2021/08/18/more-americans-now-say-
government-should-take-steps-to-restrict-false-information-online-than-in-
2018/.
[16]
Moschella, D. (2022), “It’s not just Facebook—“Old media” spreads misinformation,
too”, https://itif.org/publications/2022/01/10/its-not-just-facebook-old-media-
spreads-misinformation-too.
[36]
Neale, S. (1977), “Propaganda”, Screen, Vol. 18/3, pp. 9-40,
https://doi.org/10.1093/screen/18.3.9.
[31]
Newman, N. et al. (2021), Reuters Institute Digital News Report 2021,
https://reutersinstitute.politics.ox.ac.uk/sites/default/files/2021-
06/Digital_News_Report_2021_FINAL.pdf.
[37]
OECD (2021), “Enhancing public trust in COVID-19 vaccination: The role of
governments”, https://read.oecd-ilibrary.org/view/?ref=1094_1094290-
a0n03doefx&title=Enhancing-public-trust-in-COVID-19-vaccination-The-role-of-
governments.
[9]
OECD (2021), OECD Report on Public Communication: The Global Context and the
Way Forward, OECD Publishing, Paris, https://doi.org/10.1787/22f8031c-en.
[73]
| 35
DISENTANGLING UNTRUTHS ONLINE: CREATORS, SPREADERS AND HOW TO STOP THEM © OECD 2022
OECD (2020), “Combatting COVID-19 disinformation on online platforms”,
https://read.oecd-ilibrary.org/view/?ref=135_135214-
mpe7q0bj4d&title=Combatting-COVID-19-disinformation-on-online-platforms.
[23]
OECD (2020), “Transparency, communication and trust: The role of public
communication in responding to the wave of disinformation about the new
Coronavirus”, https://www.oecd.org/coronavirus/policy-responses/transparency-
communication-and-trust-the-role-of-public-communication-in-responding-to-
the-wave-of-disinformation-about-the-new-coronavirus-bef7ad6e/.
[7]
OECD (2019), An Introduction to Online Platforms and Their Role in the Digital
Transformation, OECD Publishing, Paris, https://dx.doi.org/10.1787/53e5f593-en.
[2]
OECD (2019), Going Digital: Shaping Policies, Improving Lives, OECD Publishing, Paris,
https://dx.doi.org/10.1787/9789264312012-en.
[1]
OECD (2017), Recommendation of the Council on Open Government,
https://legalinstruments.oecd.org/en/instruments/OECD-LEGAL-0438.
[88]
OED (2013), Oxford English Dictionnary, Oxford,
https://www.oed.com/viewdictionaryentry/Entry/171207.
[32]
OHCHR (2001), “Monitoring human rights tn the context of elections”, U.N. Human
Rights Office of the High Commissioner,
https://www.ohchr.org/Documents/Publications/Chapter23-MHRM.pdf.
[26]
O’Sullivan, D. (2021), “White House turns up heat on Big Tech’s Covid ’disinformation
dozen’”, CNN Business, https://edition.cnn.com/2021/07/16/tech/misinformation-
covid-facebook-twitter-white-house/index.html (accessed on 29 October 2021).
[59]
Oversight Board (2021), Oversight Board, https://oversightboard.com/ (accessed on
12 August 2021).
[78]
Paavola, T. et al. (2016), “Understanding the trolling Phenomenon: The automated
detection of bots and cyborgs in the social media”, Journal of Information
Warfare, Vol. 15/4, pp. 100-111, https://www.jstor.org/stable/26487554.
[39]
Pew Research Center (2019), “Many Americans say made-up news is a critical
problem that needs to be fixed”,
https://www.pewresearch.org/journalism/2019/06/05/many-americans-say-made-
up-news-is-a-critical-problem-that-needs-to-be-fixed/ (accessed on
26 October 2021).
[51]
Politico (2021), The facebook papers, https://www.politico.com/tag/the-facebook-
papers, accessed 1 November 2021 (accessed on 1 November 2021).
[15]
36 |
DISENTANGLING UNTRUTHS ONLINE: CREATORS, SPREADERS AND HOW TO STOP THEM © OECD 2022
Rahman, M., A. Chia and W. Gonzalez (2021), “Using NLP to fight misinformation and
detect fake news”, https://omdena.com/blog/fighting-misinformation/ (accessed
on 12 August 2021).
[81]
Rosenberg, M., N. Confessore and C. Cadwalladr (2018), “How Trump consultants
exploited the Facebook data of millions”, New York Times,
https://www.nytimes.com/2018/03/17/us/politics/cambridge-analytica-trump-
campaign.html (accessed on 5 August 2021).
[12]
Rosen, G. (2021), “Community Standards Enforcement Report, Second Quarter 2021”,
Facebook, https://about.fb.com/news/2021/08/community-standards-
enforcement-report-q2-2021/ (accessed on 22 September 2021).
[64]
Sacco, L. et al. (2021), “Emergence of knowledge communities and information
centralization during the COVID-19 pandemic”, Social Science & Medicine,
Vol. 285, https://www.sciencedirect.com/science/article/pii/S0277953621005475.
[60]
Schaake, M. (2021), “Big Tech calls for ‘regulation’ but is fuzzy on the details”,
https://www.ft.com/content/a0a7f8de-f365-4e4e-a755-284df91c6e3a (accessed
on 29 October 2021).
[19]
Schechner, S. (2021), “Facebook is rebuked by Oversight Board over transparency on
treatment of prominent users”, Wall Street Journal,
https://www.wsj.com/articles/facebooks-oversight-board-says-company-wasnt-
fully-forthcoming-on-treatment-of-high-profile-users-11634817601.
[79]
Scott, M. (2022), “As war in Ukraine evolves, so do disinformation tactics”, Politico,
https://www.politico.eu/article/ukraine-russia-disinformation-propaganda/.
[5]
Shearer, E. (2021), “News use across social media platforms in 2020”, Pew Research
Center, https://www.pewresearch.org/journalism/2021/01/12/news-use-across-
social-media-platforms-in-2020/.
[34]
Shearer, E. and A. Mitchell (2021), “More than eight-in-ten Americans get news from
digital devices”, Pew Research Center, https://www.pewresearch.org/fact-
tank/2021/01/12/more-than-eight-in-ten-americans-get-news-from-digital-
devices/.
[35]
Silver, L. (2019), “Misinformation and fears about its impact are pervasive in 11
emerging economies”, Pew Research Center, https://www.pewresearch.org/fact-
tank/2019/05/13/misinformation-and-fears-about-its-impact-are-pervasive-in-
11-emerging-economies/ft_19-05-
13_misinformation_exposuretoincorrectinformationwidespread/.
[50]
| 37
DISENTANGLING UNTRUTHS ONLINE: CREATORS, SPREADERS AND HOW TO STOP THEM © OECD 2022
Somers, M. (2020), “Deepfakes, explained”, MIT Sloan School, MIT Sloan School,
https://mitsloan.mit.edu/ideas-made-to-matter/deepfakes-explained (accessed
on 29 October 2021).
[86]
Statt, N. (2020), “Major tech platforms say they’re ‘jointly combating fraud and
misinformation’ about COVID-19”, The Verge,
https://www.theverge.com/2020/3/16/21182726/coronavirus-covid-19-facebook-
google-twitter-youtube-joint-effort-misinformation-fraud.
[63]
Swire-Thompson, B. and D. Lazer (2020), “Public health and online misinformation:
Challenges and recommendations”, Annual Review of Public Health, Vol. 41/1,
pp. 433-451, https://doi.org/10.1146/annurev-publhealth-040119-094127.
[24]
Talbot, H. and A. Nusiebah (2021), “The edge of the infodemic: Challenging
misinformation in Aotearoa”,
https://www.classificationoffice.govt.nz/assets/PDFs/Classification-Office-Edge-
of-the-Infodemic-Report.pdf.
[53]
Taylor, M. (2019), “Combating disinformation and foreign interference in
democracies: Lessons from Europe”,
https://www.brookings.edu/blog/techtank/2019/07/31/combating-
disinformation-and-foreign-interference-in-democracies-lessons-from-europe/.
[11]
Twitter (2021), “COVID-19 misinformation”,
https://transparency.twitter.com/en/reports/covid19.html#item2:2020-jul-dec
(accessed on 22 September 2021).
[66]
Twitter (2013), “New tweets per second record, and how!”,
https://blog.twitter.com/engineering/en_us/a/2013/new-tweets-per-second-
record-and-how (accessed on 21 September 2021).
[68]
UNCHR (1996), “The right to participate in public affairs, voting rights and the right
to equal access to public service”, U.N. Committee on Human Rights,
https://www.equalrightstrust.org/ertdocumentbank/general%20comment%2025.p
df.
[20]
Wall Street Journal (2021), The facebook files, https://www.wsj.com/articles/the-
facebook-files-11631713039 (accessed on 1 November 2021).
[14]
Wang, L. et al. (2018), “Five shades of untruth: Finer-grained classification of fake
news”, http://dx.doi.org/10.1109/ASONAM.2018.8508256.
[82]
Wardle, C. (2019), “First Draft’s essential guide to understanding information
disorder”, http://creativecommons.org/licenses/by-nc-nd/4.0/ (accessed on
9 August 2021).
[33]
38 |
DISENTANGLING UNTRUTHS ONLINE: CREATORS, SPREADERS AND HOW TO STOP THEM © OECD 2022
Waterson, J. and D. Milmo (2021), “Facebook whistleblower Frances Haugen calls for
urgent external regulation”, The Guardian,
https://www.theguardian.com/technology/2021/oct/25/facebook-whistleblower-
frances-haugen-calls-for-urgent-external-regulation (accessed on
8 February 2022).
[87]
WHO et. al. (2020), “Managing the COVID-19 infodemic: Promoting healthy
behaviours and mitigating the harm from misinformation and disinformation”,
https://www.who.int/news/item/23-09-2020-managing-the-covid-19-infodemic-
promoting-healthy-behaviours-and-mitigating-the-harm-from-misinformation-
and-disinformation.
[6]
Woolley, S. (2020), The Reality Game: How the Next Wave of Technology Will Break
the Truth, Public Affairs, New York.
[48]
Yadav, K. et al. (2021), “Countries have more than 100 laws on the books to combat
misinformation. How well do they work?”, https://thebulletin.org/premium/2021-
05/countries-have-more-than-100-laws-on-the-books-to-combat-misinformation-
how-well-do-they-work/ (accessed on 5 August 2021).
[44]
Zimdars, M. and K. McLeod (2020), Fake news: Understanding media and
misinformation in the digital age, Cambridge, Massachusetts ; London, England :
The MIT Press.
[89]
