Indonesia Country Report

In partnership with UNESCO
Content Moderation and

Local Stakeholders in
Indonesia
June 2022
ARTICLE 19
T: +44 20 7324 2500
F: +44 20 7490 0566
E: info@article19.org
W: www.article19.org
Tw: @article19org
Fb: facebook.com/article19org
© ARTICLE 19, 2022

This publication was produced with the financial support of the European Union and UNESCO.
The designations employed and the presentation of material throughout this publication do not
imply the expression of any opinion whatsoever on the part of UNESCO or the European Union
concerning the legal status of any country, territory, city or area or its authorities, or concerning
the delimitation of its frontiers or boundaries.
The authors are responsible for the choice and the presentation of the facts contained in this
publication and for the opinions expressed therein, which are not necessarily those of UNESCO or
the European Union and do not commit the Organizations.
This work is provided under the Creative Commons Attribution-Non-Commercial-ShareAlike 3.0

licence. You are free to copy, distribute and display this work and to make derivative works,
provided you:
1) give credit to ARTICLE 19;
2) do not use this work for commercial purposes;
3) distribute any works derived from this publication under a licence identical to this one.
To access the full legal text of this licence, please visit:

https://creativecommons.org/licenses/by-sa/3.0/legalcode
2
Acknowledgements
Many thanks to Sherly Haristya, who conducted this research and authored this report in
close consultation with ARTICLE 19.
A Data Protection Officer certified by the Association of Data Privacy Professionals

Indonesia, Sherly earned her PhD in Communication Studies from the Wee Kim Wee
School of Communication and Information, Nanyang Technological University, Singapore.
Her research has been bridging the global and national (Indonesia) debates on Internet
governance, content moderation and personal data protection issues. During her PhD, she
also had the opportunity to serve on the national multi-stakeholder organising committee
of the UN Internet Governance Forum in Bali, Indonesia, in 2013 and on the Secretariat of
Internet Assigned Numbers Authority Stewardship Transition Coordination Group.
Many thanks to all the persons who have shared their expertise and views and contributed
to the research process.
Thanks also to our partners at UNESCO who have provided input and feedback at various
stages of the development of this report.
3
Contents
Executive summary 5
Introduction 7
About the project 7
Methodology and structure of the report 9
Indonesia at a glance 11
The state of content moderation in Indonesia 15

Social media landscape in Indonesia 15
Overview: Impact of content moderation on peace and stability 16
Social media platforms operate globally 18
Problematic content in Indonesia 22
Content moderation dynamics in Indonesia 29
Interim conclusion 43
Analysis of stakeholders 46
Civil society organisations 46
Media industries and journalists 52
Academia and think tanks 52
Content creators and users 54
Social media platforms 54
Public authorities 55
Conclusion 56
Recommendations 57
Sustainability of a civil society coalition 60
Creating a coalition 63
Annex A: Risk analysis 66

Annex B: Potential members of the coalition 68
Annex C: Interview sheet 69
Bibliography 70
Endnotes 75
4
Executive summary
Executive summary
• The population of Indonesia is characterised by social, cultural, and economic diversity.
Against a background of deep historical roots of segregation, there are efforts to
spread false information, hate speech, and violent extremist materials on social media
that have been damaging to individuals and civil society.
• The problem is particularly acute in relation to ‘grey-area’ speech, which is speech that
falls under categories whose prohibition is not mandated by, or not compatible with,
international standards on freedom of expression and/or community standards, but
whose amplification can result in real-world violence. In this regard, content moderation
conducted by social media platforms can play a critical role in preventing the spread of
such problematic content from turning into real-world dangers.
• While there are efforts made by platforms to conduct content moderation in line with
the local context in Indonesia, the research found a disconnect between the global
community rules of platforms and their enforcement at the local level, especially in
tackling the ‘grey-area’ speech.
• Effective content moderation in a big and diverse country such as Indonesia requires a
transparent and sustainable dialogue between platforms and local civil society groups.
The leading civil society groups in Indonesia, particularly those who are the trusted
partners of platforms, have a special line of communication to address urgent
problematic content and explain the local context to platforms. However, the
partnerships should be improved, and local partners should be empowered to create a
more inclusive, credible, and meaningful dialogue between Indonesian civil society
groups, individual users, and platforms.
• The report tested the idea that a local Coalition on Freedom of Expression and Content
Moderation could play a role to fill in the gap in current content moderation practices.
Most interviewees responded positively to this suggestion and the report concludes
with recommendations on how to facilitate the creation of such a civil society Coalition
on Freedom of Expression and Content Moderation in Indonesia. By reinforcing the
5
Executive summary
capacity of local actors, the coalition would act as a bridge to develop and nurture
relationships and effective dialogue with social media companies on content
moderation issues. The coalition would contribute to the development of content
moderation practices that uphold international standards on freedom of expression
while duly taking into consideration the local context.
6
Introduction
Introduction
This publication has been produced as part of the United Nations Educational, Scientific
and Cultural Organization's (UNESCO) project Social Media 4 Peace funded by the
European Union (EU).
About the project
This report is part of the Social Media 4 Peace project that UNESCO is implementing in
Bosnia and Herzegovina (BiH), Kenya, and Indonesia, with support of the EU. The overall
objective of the project is to strengthen the resilience of civil society to potentially harmful
content spread online, in particular hate speech and disinformation, while protecting
freedom of expression and contributing to the promotion of peace through digital
technologies, notably social media. ARTICLE 19’s contribution to the project focuses on
concerns raised by the current practices of content moderation on dominant social media
platforms in the three target countries.
ARTICLE 19 considers that social media companies are, in principle, free to restrict
content on the basis of freedom of contract, but that they should nonetheless respect
human rights, including the rights to freedom of expression, privacy, and due process.
While social media platforms have provided opportunities for expression, a number of
serious concerns have come to light. The application of community standards has led to
the silencing of minority voices. The efforts of tech companies to deal with problematic
content are far from being evenly distributed: for instance, it has been shown that ‘87% of
Facebook’s spending on misinformation goes to English-language content, despite the
fact that only 9% of its users are English speaking.’ It has also been revealed that most
resources and means in terms of content moderation are being allocated to a limited
number of countries. Generally speaking, the transparency and dispute resolutions over
content removals have so far been inadequate to enable sufficient scrutiny of social
media platforms’ actions and provide meaningful redress for their users. Finally, it is
doubtful that a small number of dominant platforms should be allowed to hold so much
power over what people are allowed to see without more direct public accountability.
7
Introduction
This report specifically looks at the situation of local actors who, while they are impacted
by the circulation of harmful content on social media or the moderation thereof, often find
themselves unable to take effective action to improve their situation in that respect. They
may feel frustrated by the inconsistencies of platforms’ application of their own content
rules; they may feel that global companies ignore their requests or misunderstand the
current circumstances of the country or region. Some may lack understanding of content
rules or of content moderation, but that is not the case of all local stakeholders.
The research then seeks to test, through the views of local stakeholders, the assumption
that a local Coalition on Freedom of Expression and Content Moderation could play a role
to fill the gap between the realities of local actors and companies that operate on a global
scale. The idea for such a coalition is based on ARTICLE 19’s work on the development of
Social Media Councils, a multi-stakeholder mechanism for the oversight of content
moderation on social media platforms. ARTICLE 19 suggested that Social Media Councils
should be created at a national level (unless there was a risk that it would be easily
captured by the government or other powerful interests) because this would ensure the
involvement of local decision-makers who are well-informed of the local context and
understand its cultural, linguistic, historical, political, and social nuances. While the
development of a self-regulatory, multi-stakeholder body such as a Social Media Council is
a long-term and complex endeavour, a local Coalition on Freedom of Expression and
Content Moderation would be a lighter approach that could be supported within a shorter
timeframe. Basing its work on international standards on freedom of expression and other
fundamental rights, such a coalition could provide valuable input to inform content
moderation practices, notably through its knowledge and understanding of the local
languages and circumstances. As a critical mass of local stakeholders, it could engage
into a sustainable dialogue with social media platforms and contribute to addressing
flaws in content moderation and improving the protection of fundamental rights online.
The coalition could provide training and support on freedom of expression and content
moderation to local civil society actors that are impacted by content moderation. Finally, it
could possibly pave the way to the creation of a Social Media Council in the country at a
later stage. Through this research, at the initial stage of the Social Media 4 Peace project,
8
Introduction
the idea of a local Coalition on Freedom of Expression and Content Moderation was
submitted to local stakeholders, whose views have enabled the formulation of
recommendations on how to approach the facilitation of a pilot coalition in the specific
context of Indonesia. In order to guarantee the effective ownership of the coalition by its
members, the process facilitating its creation will necessarily include a validation exercise
that ensures that potential members have the opportunity to discuss the findings of the
research.
For the purposes of this report, we rely on the following definitions:
• Content moderation includes the different sets of measures and tools that social media
platforms use to deal with illegal content and enforce their community standards
against user-generated content on their service. This generally involves flagging by
users, Trusted Flaggers or ‘filters’, removal, labelling, down-ranking or demonetisation of
content, or disabling certain features.
• Content curation is how social media platforms use automated systems to rank,
promote, or demote content in newsfeeds, usually based on their users’ profiles.
Content can also be promoted on platforms in exchange for payment. Platforms can
also curate content by using interstitials to warn users against sensitive content or
applying certain labels to highlight, for instance, whether the content comes from a
trusted source.
Methodology and structure of the report
This research is based on a combination of desk research and qualitative interviews with
26 key informants (representatives from various local stakeholders). The researcher also
had the opportunity to present preliminary findings and gather further feedback and
contributions from stakeholders during two focus group discussions organised by
Perludem and one public event organised by SAFEnet.
The desk research allowed for the identification of issues linked to the circulation of
problematic content on social media in Indonesia. The identified content moderation
issues were then discussed with multiple stakeholders during interviews. The interviews
9
Introduction
aimed to understand the experiences and challenges of Indonesian groups in dealing with
platforms on content moderation issues.1 The interviewees also conveyed their reflections
on the idea of a local coalition on content moderation and freedom of expression. The
potential structures, members, roles, and dynamics of the coalition were also discussed.
The researcher presented the findings of this research in two focus group discussions
conducted by the election watchdog Perludem (the Association for Elections and
Democracy). Perludem held a series of discussions to prepare a roadmap for secure
elections in 2024 in Indonesia. The first focus group discussion was on 29 December
2021 and involved participants from multiple stakeholders (including civil society groups
and electoral bodies; social media representatives were unable to attend the meeting).
The second focus group discussion took place on 18 February 2022, and only civil society
actors were invited. Overall, Perludem and the participants welcomed the idea of forming a
local civil society coalition on content moderation and freedom of expression.
The researcher was also invited to discuss the findings during the public launch of
SAFEnet’s 2021 Digital Rights in Indonesia Situation Report on 2 March 2022. The four
presenters from civil society groups and a representative of Meta commented on the
current state of digital rights in Indonesia and discussed the potential way forward and the
role that civil society could take in responding to the situation. Around 65 participants
attended the live webinar. The researcher explained the findings of this ongoing research
and discussed with the other presenters and meeting participants the need to form a local
civil society coalition on content moderation and freedom of expression. The panel
concluded with a shared view on the need to unite and empower Indonesian civil society
groups to ensure their participation in the development and enforcement of Internet-
related policies in Indonesia that balance freedom of expression with the safety of
individuals and the public.
At the time of finalising this report, the dialogue between the researcher, public authorities,
social media companies, and civil society groups continues, particularly on the preparation
of peaceful elections in 2024 and how a future coalition on content moderation and
freedom of expression could play a role in securing peaceful elections.
10
Introduction
The structure of the report is as follows:
The Introduction highlights the diversity and complexity of Indonesia society. It further
presents how the deep conflicts that exist within society have at times been exploited for
political and economic profit.
The first chapter, The state of content moderation in Indonesia, describes the landscape
of social media platforms and explores the dynamics and issues related to the use of
social media and the practices of content moderation in the country. One striking issue is
the proliferation of ‘grey-area’ content (content that contains what can be described as ‘the
seeds of hate’ but does not necessarily amount to explicit incitement; the key concern with
such content is that its massive amplification could lead to real-world violence). This
chapter looks at the need for social media to understand and appreciate the local context
when applying the global community standards, and concludes that while there are a
number of initiatives and discussions involving social media with civil society and state
actors, there is room for more meaningful dialogue within the development and
enforcement processes of platforms’ community standards.
The second chapter, Analysis of stakeholders, provides an analysis of the relevant

stakeholder groups that deal with or are impacted by the content moderation practices.
Based on this analysis, the third chapter includes recommendations on how to facilitate
the formation of a civil society coalition on content moderation and freedom of expression
in Indonesia to bridge the dialogue between social media and local civil society.
Indonesia at a glance
Many of the ‘largest and the most’ designations are bynames for Indonesia. It is the
world’s largest island country located in the Southeast Asian region. Spanning across
1,905 square kilometres, the country is divided into Western, Central, and Eastern parts of
Indonesia. It is comprised of five main islands (Sumatera; Kalimantan; Java; Nusa
Tenggara and Bali; and Sulawesi, Maluku and Papua) that are further divided into 34
provinces and over 17,500 islands. With a population of over 277 million people, it is the
fourth most populous country and the third largest democracy in the world. In terms of
11
Introduction
economy, Indonesia is the largest economy in Southeast Asia and is predicted to be the
fifth largest economy in the world by 2024. While its economy has been impacted by the
pandemic, Indonesia is in the process of economic recovery to a pre-pandemic level, with
a gross domestic product growth rate expected of about 4% this year.
The population of Indonesia is characterised by social, cultural, and economic diversity.

While recognised as a country with the largest number of Muslims (231 million people in
2021), Indonesia also acknowledges five other official religions, namely Protestantism,
Catholicism, Hinduism, Buddhism, and Confucianism. There are more than 300 ethnic
groups living together in the country. Besides some of the largest ethnic groups, Javanese
(40% of the total population), Sundanese (15.5%), Malay (3.7%), Bataknese (3.6%),
Madurese (3%), and Betawi (2.9%), there are also other smaller ethnic groups, such as the
Buginese, Balinese, Acehnese, Papuan and Chinese Indonesians. The national language is
Indonesian. However, only 7% of the total population speak Indonesian as their mother
tongue: the rest use it as their second language. There are more than 300 different
languages and local dialects spoken in Indonesia.
The pandemic has widened the economic and educational disparity in the country. There
are increasing numbers of people living in poverty today. The Central Statistics Agency
(BPS) released the level of inequality in Indonesia’s population expenditure as measured
by the Gini Ratio, which was 0.384 in March 2021. The country has not been able to return
to September 2019’s Gini Ratio of 0.380. The government has been struggling to close the
disparity in education services and standards, and the gap has even widened during the
pandemic. Only those with sufficient facilities, access to electricity, and the Internet could
adapt. However, Internet connectivity is highly concentrated in the Western part of
Indonesia, particularly on the more urbanised island of Java. Meanwhile, the eastern
provinces, such as Maluku, West Sulawesi, North Maluku, East Nusa Tenggara, and Papua,
are struggling with Internet access.
Prior to its declaration as an independent country, there was already social, political, and
cultural segregation during colonial times in the Dutch East Indies (as Indonesia was
known at the time). The segregation between Muslims and Christians was reinforced by
12
Introduction
the discriminatory policies of the colonial regime against certain groups for economic and
political gains. A cooperative yet competitive relationship between the ethnic Chinese and
the Dutch colonial government was also formed during that period. The Dutch needed to
cooperate with the Chinese to keep the Dutch East Indies economy alive. However, the
Chinese were gradually considered a threat as the income of the colony was declining at
that time. This resulted in the killing of Chinese people in 1740 (known as ‘Geger Pacinan’
or ‘Tragedi Angke’).
The founding father of Indonesia, Sukarno, realised the opportunities and challenges that
cultural diversity represented for the country. Sukarno formulated the official state
philosophy, known as Pancasila (Five Principles), to serve as the foundation of the
country:
1. Belief in the one and only God

2. Just and civilised humanity
3. The unity of Indonesia
4. Democracy guided by the inner wisdom in the unanimity arising out of deliberations
amongst representatives
5. Social justice for all of the peoples of Indonesia
These principles are often generalised in terms of religious devotion, humanitarianism,

nationalism, consultative democracy, and social justice.
As Indonesia was established as a sovereign nation in 17 August 1945, Pancasila became

the foundational element of the 1945 Constitution of the Republic of Indonesia. Section
XA of the Constitution identifies ‘human rights’ as recognised and protected by the
government. The need to unify such a diverse country also led to the formulation of the
official national motto, Bhinneka Tunggal Ika (‘Unity in Diversity’), which is inscribed on the
national emblem of Indonesia, the Garuda (Eagle) Pancasila. It declares the essential unity
of its inhabitants regardless of ethnic, regional, social, or religious differences.
Following the 1965 incident,2 Suharto took the leadership of the country. There is a long
list of human rights violations in his era: during Suharto’s 32 years of military dictatorship,
13
Introduction
there was a cultural and political repression in the country. The government controlled the
press and circulation of information in the country and restricted freedom of expression.
Accordingly, people had to rely on rumours and non-official sources in order to get
information about, and understand, the current events of the day. The government also
introduced the official use of the discriminatory labels of pribumi (indigenous) and non-
pribumi, which was lifted after the post-Suharto era. The ethnic Chinese were not
recognised as ethnic groups and all Chinese-related culture was banned (the celebration
of Chinese New Year, the use of Chinese names and the Mandarin language). Fuelled by
the government and military officials, there was strong anti-Chinese sentiment across the
country, expressed in the May 1998 riots that led to the resignation of President Suharto.
The fall of Suharto in 1998 brought reforms and democratic gains to Indonesia. The
reformation in 1998 was a milestone for the acknowledgement of human rights principles
as the country adopted a comprehensive human rights law (Law No. 39/1999 on Human
Rights). Pluralism in politics and the media also took hold in the country.3 Acknowledging
the impressive democratic gains since 1998, Freedom House describes Indonesia as
partly free in their 2021 report, noting that the ‘the country continues to struggle with
challenges including systemic corruption, discrimination and violence against minority
groups, conflict in the Papua region, and the politicised use of defamation and blasphemy
laws’.
14
The state of content moderation in Indonesia

Social media landscape in Indonesia
Internet penetration is steadily increasing in Indonesia. In 2021, data from the Indonesian
Internet Service Provider Association (APJII) and HootSuite placed Indonesia’s Internet
penetration rate at 73.7% (202.6 million people). The penetration is due to the rapid growth
in the number of mobile Internet subscriptions. There were over 345.3 million
subscriptions in 2021, an increase of 1.2% since 2020. Accordingly, 96.4% of total Internet
users (195.3 million people) use mobile Internet in the country.
The first reason why Indonesians use the Internet is to allow them to use social media.
There are at least 16 social media platforms and chat apps being used in the country.
They are YouTube (93.8% of usership), Instagram (86.6%), Facebook (85.5%), Twitter
(63.6%), Line (44.3%), TikTok (38.7%), LinkedIn (39.4%), Pinterest (35.6%), WeChat (26.2%),
Snapchat (25.4%), Tumblr (18.4%), and Reddit (17.1%). Meanwhile, chat apps used in the
country are WhatsApp (87.7%), Facebook Messenger (52.4%), Telegram (28.5%), and
Skype (24.3%).
The research from APJII confirmed that there are five main social media platforms used in
Indonesia: YouTube, Facebook, Instagram, Twitter, and LinkedIn. Meanwhile, data from
HootSuite also shows an increasing number of TikTok users. With regard to chat apps,
APJII found that WhatsApp is predominantly used in the country (93.7%).
YouTube has to a large extent replaced the television industry in the country. There are
around 190 million Indonesian Internet users ranging from 15–64 years old that use
YouTube. This is illustrated by the fact that the popular YouTube channel of Indonesian
artist Deddy Corbuzier named #CLOSEDTHEDOOR4 put up billboards in main streets of
some big cities in Indonesia to promote the channel and equate it to a television show.
The billboard challenged the public with the question: ‘STILL WATCHING TV?’.
With more than 173 million users in Indonesia, Facebook is dominated by the younger
segments of the adult population. There are 33.6% of users from the millennial
15
generation (25–34 years old) and 30.2% from generation Z (18–24 years old), 14.3% are
in the age range of 35–44 years old. Instagram is more popular among young adults and
teenagers aged 18–24 years old (36.4%), 25–34 years old (31.6%), and 13–17 years old
(12.9%).
Indonesian users have been especially dependent on big social media platforms during
the pandemic. They go to YouTube and Facebook for entertainment as well as educational
content, news, and political and social content. Indonesian users access Twitter if they
want to look for information from experts in various fields, including information related to
Covid-19. Indonesian users use TikTok and Instagram mostly for entertainment, life style,
and e-commerce.
While they use social media for various purposes, several studies indicate that there is a
key similarity in the use of social media and chat apps among Indonesian users –
receiving and sharing information with friends and relatives.5
Overview: Impact of content moderation on peace and stability
This chapter discusses categories of problematic content (disinformation, hate speech,

online harassment, terrorism and radicalism, and online gender-based violence), how such
content is amplified by malicious actors, the relationships between stakeholders and
platforms, the availability of content rules in local languages, the effectiveness of
remedies provided by platforms, the problems related to polarisation/conflicts in society,
the marginalisation of certain groups, and the impact on freedom of expression and media
freedom.
There are efforts in Indonesia to exploit conflicts that exist in society for political and
economic gain. The spread of false information and hate speech on social media
platforms is particularly intensified during election periods. Political candidates hire social
media campaign strategists to manage their online election campaigns: this strategy often
involves hiring ‘buzzers’ and the use of chatbots to spread disinformation and steer public
debates.
16
The use of social media in Indonesian politics started during the 2012 Jakarta
gubernatorial election. In the final round of votes, two candidates remained for the
positions of governor and vice governor each: the incumbent Fauzi Bowo and Nahrowi
Rahmi and the challenging pair Joko Widodo (Jokowi) and Basuki Tjahaya Purnama
(commonly called Ahok). The language of the Bowo–Rahmi campaign heavily emphasised
the individual traits of Jokowi and Ahok. They called for people to choose a candidate of
the same faith and deliberately insinuated that Jokowi’s mother was Catholic. Governor
Jokowi and Vice Governor Ahok ended up winning the election.
The centrality of social media in election campaigns was even more apparent in the 2014
Presidential election and the 2017 Jakarta gubernatorial election. In 2014, the clash
between supporters of two candidates, Jokowi and Prabowo, was visible on social media
platforms. As Jokowi was elected to be the President of Indonesia, Ahok filled the position
of Jakarta Governor. In 2017, Ahok ran for re-election in the Jakarta gubernatorial election.
Following the second round of voting, there was increasing polarisation between the
supporters of the two remaining candidates, Ahok and Anis. The supporters of one
candidate labelled the supporters from the other camp as the enemy.
An outspoken figure with a double minority background (Chinese Indonesian and

Christian), Ahok has been subject to racist comments. In a campaign speech, Ahok
criticised his political opponents for using Islam as a campaign tool. A man named Buni
Yani edited the speech and added a caption that made Ahok seem like he was insulting
the Quran. The edited video was uploaded to Yani’s social media accounts and went viral,
which led to a flare up. Massive protests from Indonesian Muslim groups took place on 4
November 2016 (Action 411) and 2 December 2016 (Action 212). These huge pressures
forced Jokowi to decide that Ahok must be charged and prosecuted. Ahok did not win the
election, and eventually he was jailed for blasphemy.
The circulation of disinformation that sought to delegitimise the election process and
results continued into the Presidential election in 2019. Building on the sentiment against
minority groups developed in the previous elections, there was a circulation of hoaxes on
social media that claimed some protestors were shot by Chinese police during the post-
17
election demonstrations in May 2019. While this message contained no explicit statement
that promoted violence, anti-Chinese sentiment triggered chaos (this is a clear example of
the risks linked to ‘grey-area’ content, as shall be further explored in this report). The
incident prompted the government to block access to social media platforms and chat
apps to curb the spread of such hateful disinformation. However, the spread of anti-
Chinese disinformation found its way to the encrypted platform Telegram.
In August 2019, the government ordered further Internet restrictions during violent
protests in Papua. Despite the decision of the Jakarta State Administrative Court in 2020
that there is no legal basis in the Electronic Information and Transactions (ITE) Law that
regulates Internet shutdowns, Internet users in Papua experienced issues in accessing the
Internet in April 2021.6
Social media platforms operate globally
This section examines the disconnect between the global community rules of platforms
and their enforcement, particularly in tackling ‘grey-area’ speech in the specific context of
Indonesia. To do so, this section focused on examining the hate speech policies of
platforms as they are particularly relevant to understanding the complexity of ‘grey-area’
content.
For several reasons, the global content rules appear to be problematic.7 A first serious
flaw in the system is that the level of commitment to provide the full details of community
standards in the Indonesian language varies from one platform to another. As the
community standards are living documents that are regularly updated, the availability of
the Indonesian translations appears not to be updated. For example, as shown in the
Figure 1 captured on 4 March 2022, the Indonesian version of Facebook’s misinformation
guidelines announces that ‘Some of the content on this page is not yet available in
Indonesian language.’ In the meantime, in its English Community Standards, Facebook
states ‘Please note that the US English version of the Community Standards reflects the
most up-to-date set of the policies and should be used as the master document.’
18
Figure 1: Screenshot of Facebook’s misinformation guidelines in Indonesia language.8
Moreover, while global content rules elaborated by social media companies typically
explain the types of content that are prohibited on their platforms along with some
specific examples to help their users understand the scope of the restrictions, they do not
seem to take into consideration ‘grey-area’ content and how, in the context of Indonesia,
such content can lead to real-world harm. Here are some examples:
• Twitter
‘We prohibit content that makes violent threats against an identifiable target. Violent
threats are declarative statements of intent to inflict injuries that would result in serious
and lasting bodily harm, where an individual could die or be significantly injured, e.g., ‘I
will kill you.’’9
• TikTok
‘Do not post, upload, stream, or share:
• Hateful content related to an individual or group, including:
– calling for or justifying violence against them.’10
19
• YouTube
‘Don’t post content on YouTube if the purpose of that content is to do one or more of
the following.
• Encourage violence against individuals or groups based on any of the attributes

noted above. We don’t allow threats on YouTube, and we treat implied calls for
violence as real threats. You can learn more about our policies on threats and
harassment.
• Incite hatred against individuals or groups based on any of the attributes noted
above.’11
• Instagram
‘We remove content that contains credible threats or hate speech, content that targets
private individuals to degrade or shame them, personal information meant to blackmail
or harass someone, and repeated unwanted messages.’12
• Facebook
‘We aim to prevent potential offline harm that may be related to content on Facebook.
While we understand that people commonly express disdain or disagreement by
threatening or calling for violence in non-serious ways, we remove language that incites
or facilitates serious violence.’13
In addition to difficulties related to the definition of categories of prohibited content,14 it

would be helpful to know how platforms assess ‘grey-area’ speech in relation to the
potential harms that are likely to occur in light of the sensitivity of certain issues in the
local context. This may include, for example, the motives of the poster, the frequency of
past violations made by the poster, the profile of the victim, the power relationship
between the poster and the victim, and the harm likely to result in relation to the sensitivity
of the issue from the local perspective. Furthermore, clear indications are necessary about
the appeals mechanisms, appropriate content moderation measures that companies may
take (such as the types of sanctions), and the timeframe within which companies intend to
20
manage the content in question. In that respect, for example, Facebook briefly explains its
considerations for deciding if a threat is credible and needs to be removed:
We remove content, disable accounts and work with law enforcement when we believe
there is a genuine risk of physical harm or direct threats to public safety. We also try to
consider the language and context in order to distinguish casual statements from
content that constitutes a credible threat to public or personal safety. In determining
whether a threat is credible, we may also consider additional information like a person’s
public visibility and the risks to their physical safety.15
In response to a question on whether it is possible for community guidelines to specify

how to handle ‘grey-area’ speech, a platform representative in Indonesia underlined how it
is not always straightforward to translate the policy texts into action particularly when an
abundance of the content requires a comprehensive understanding of many contexts.
‘Grey-area’ content is, consequently, difficult to define and therefore content moderation in
this area could be complemented by a dialogue with locally-relevant multi-stakeholder
expert groups to guide the processes.
This finding highlights the gap in the current state of content moderation that is primarily
conducted by platforms. There is a need for the community standards of platforms to
provide a reliable and consistent baseline to guide content moderation decision-making
processes. However, given the prevalence and complexity of ‘grey-area’ content in
Indonesia (see below), community standards may not be able to capture all the
considerations and enable effective responses. This means that platforms should work
hand-in-hand with the relevant local stakeholder groups to enhance content moderation
mechanisms. Academic research funded by Facebook came to a similar conclusion of a
need to complement community standards with guidance from local groups in order to
fully grasp the complexity of the local context and the degree of harm experienced by the
related groups.
21
Problematic content in Indonesia
This section addresses the complexity of several categories of content that are distributed
massively in Indonesia. These include disinformation, hate speech, online harassment,
terrorism, radicalism, and online gender-based violence. While these are different
categories of problematic content, they all present the same difficulty related to what this
study describes as ‘grey-area’ content – ‘seeds of hate’ directed to certain individuals or
groups, which, although it does not amount to an explicit incitement to discrimination,
hostility, or violence, may be amplified through social media and lead to serious violence.
The case studies below showcase the gap between the global community rules of
platforms and their enforcement at the local level.
Case study 1
This example looks at the absence of shared understanding between a platform and an
election authority regarding content moderation measures.
As explained in the ‘Introduction to the context of the country,’ there are deep historical
social, political, and cultural divides in Indonesia, particularly between Muslims and
Christians. Since 2012, societal tensions, particularly through social media, have
intensified. In 2014, the clash between supporters of two Presidential candidates,
Jokowi and Prabowo, gained visibility on social media platforms. In 2017, there was
increasing polarisation between the supporters of the two final candidates, Ahok and
Anis Baswedan. Ahok was jailed for blasphemy after a manipulated video gave the
impression that he was insulting the Quran. The circulation of disinformation using anti-
Chinese and anti-communist sentiments continued into the Presidential election in
2019.
Observing such a critical situation, a representative of an election authority argued with

the local representatives of a platform that the proliferation of anti-Chinese and anti-
Communist content during the campaign period for the 2019 Presidential election was
dangerous. They argued that such content might lead to violence and therefore
22
requested the platform to moderate that type of content. However, the platform was
consistently resistant to taking any actions because the content was perceived as not
containing incitement to violence. The consistent refusal frustrated the election
authority and eventually stopped them from flagging that ‘grey-area’ content to the
platform.
Case study 2
This case study shows how a platform refused requests to take down online gender-
based violence with no explanation for the factors that determine how the platform
assess the severity of a ‘grey-area’ speech, the appropriate content moderation
measures, and the approximate timeframe to manage the content in question.
Two Muslim women reached out to Andreas Harsono, the Indonesian researcher of
Human Rights Watch, asking for help. They had been subjected to online bullying,
intimidation, and death threats on their social media accounts after one of them had
talked in a webinar, arguing that Islam allows Muslim women decide how they want to
dress, including whether they want to wear a hijab (headscarf and long-sleeve shirt) or
not. She had talked about how women should be able to control their own bodies. Her
friend and another male colleague in Cairo, Egypt, had defended the same position
during the webinar. Due to the online bullying and later the threats of being hacked and
poisoned to death, the women decided to move to Jakarta, and they reported the threats
to the National Police.
Harsono read more than 60 pages of material, checked the online material and
corresponding hyperlinks, and developed a legal analysis of the case, which he emailed
to the headquarters of the social media platform. He provided an analysis of the overall
context, such as the context of the webinar (which had been discussing a new
government regulation to allow state-school female students and teachers to choose to
wear the hijab or not). He also described the multiple posts against the three victims.
23
Harsono had just written a report for Human Rights Watch on the bullying, intimidation,
discrimination, and sometimes violence to force Muslim girls and women to wear the
hijab in Indonesia, a predominantly Muslim country: ‘I Wanted to Run Away – Abusive
Dress Codes for Women and Girls in Indonesia’ was published in March 2021. The email
concluded with a request to the platform to take down the death threats messages.
Harsono sent the email on 27 April 2021 and a follow-up on 15 June, copying several
other people at Human Rights Watch. A representative of the platform eventually replied
on 4 August. They wrote, ‘There is real nuance to our hate speech and bullying policies,
however, and a lot of what folks experience as threatening or unpleasant doesn’t meet
the (very classic human rights oriented) standard.’ The decision came as a huge
disappointment for the three victims.
Disinformation
There are concerted efforts in Indonesia to produce and spread disinformation online
through paid commenters and bots. The spread of disinformation is used to deepen the
existing social, racial, and religious divisions in the country, and such efforts are more
aggressive during election periods.
This growing problematic trend is reflected in the latest reports from the Oxford Internet
Institute released in 2020 and 2019. Back in 2017 and 2018, specific groups designed to
destabilise the country with disinformation and hate speech had been identified, such as
the Saracen and Muslim Cyber Army movements. Today, an increasing number of paid
commenters, called ‘buzzers’, as well as automated accounts, seek to manipulate the
online information landscape on behalf of political parties and private contractors. The
Oxford Internet Institute found that such teams work to support certain narratives and
attack their opposition by exploiting the prevailing divisions and delegitimising the
electoral process.
Election watchdog Perludem further emphasised that the absence of regulation on paid
political advertising and the lack of advertising transparency on social media platforms
24
have enabled the distribution of disinformation as paid political advertisement targeted at

specific users. Scholarly research has showed that this practice is dangerous for
democracy. Another study conducted by Perludem in collaboration with Facebook more
specifically reflects that the spread of disinformation, which aims to blur the public
information on technical election procedures and to delegitimise the election processes,
have the potential to eliminate a person’s right to vote.
Hate speech
Not all hate speech that circulates in Indonesia can be clearly identified.16 The less severe
forms of hate speech (which this study defines as the ‘grey-area’) are difficult to identify.
However, past incidents in the country show that such ‘grey-area’ content contributed to
deepening the polarisation in society and in some cases even resulted in riots.
The National Hate Speech Dashboard by think tank Centre for Strategic and International
Studies (CSIS Indonesia), an initiative to monitor and visualise hate speech online until the
2024’s Presidential Election and beyond, is monitoring the general trends of hate speech
on Twitter in Indonesia. Currently, the initiative is focused on examining tweets that target
vulnerable communities in Indonesia, namely Ahmadiyyah, Shi’a, and Chinese Indonesians.
The researchers noted that there is also significant hate speech in Indonesia’s online
sphere directed towards other religious and racial minority groups, as well as LGBTIQ+
communities (lesbian, gay, bisexual, transgender, intersex, and questioning).
The ‘grey-area’ type of hate speech does not necessarily include explicit incitement to
hatred or discrimination, but its amplification promotes division in society. In the long run,
it could turn into problematic speech and could potentially lead to violence.17 This is an
area in which ‘buzzers’ have become expert: they carefully craft content that claims to
promote democracy, freedom of expression, and the Unity in Diversity motto of Indonesia,
but in effect belittles opposition groups and nurtures polarisation in society.
Mixed with disinformation, polarisation is then leveraged and used to trigger animosity. An
exemplary case is the circulation of hoaxes on social media that claimed some protestors
were shot by Chinese police during the post-Presidential election demonstrations in 2019.
25
While this message contains no explicit statement that promoted violent actions, a
sensitive anti-Chinese issue was able to trigger racial hatred and resulted in riots.
The issue of hate speech in the country is also characterised by the overly broad definition
stipulated in the 2008 ITE Law. An overbroad definition of hate speech is often misused to
restrict free speech and criminalise those who are critical of power-holders. Indeed, civil
society organisations have been criticising the law.
Online harassment
There are increasing efforts in Indonesia to threaten journalists and activists through
online harassment. The Alliance of Independent Journalist (AJI) noted that worrying
doxing incidents directed against journalists started to happen back in 2018. There were
at least three online persecutions at that time. In 2020, Southeast Asia Freedom of
Expression Network (SAFEnet) reported that there were 13 cases of doxing towards
journalists, human rights activists, and citizens. This is double the number of cases
occurring in 2019.
These attacks are typically conducted after the involved actors have voiced critical
comments towards authorities through their social media accounts or media outlets. The
attacks have also been directed against activists and citizens who have joined anti-
government demonstrations. Death threats and harassment were directed at them and
their family members. Their social media accounts and chat apps were also hacked.
While death threats and harassment might not always result in real-world violence, the act
of doxing that publicly reveals someone’s personal data on social media and chat apps
(such as a home address, family photos, telephone number, and even more concerningly
their location) may lead to escalation.
For example, Freedom House highlighted that student and activist protestors who
attended a demonstration against the Omnibus Law in Yogyakarta in October 2020
received online threats. Student organisers and participants of online discussions with
topics that were critical of the authorities were also subjected to online harassment. Their
personal data, including their location, was doxed. While such cases may not always
26
cause violence, they sometimes do. For instance, Victor Mambor (a well-known West
Papua journalist, the founder of independent online media Jujur Bicara Papua
(Jubi)\Honestly Speaking about Papua), faced a series of online threats, digital attacks,
and doxing, and subsequently his car was vandalised by unknown people. The
consequences of online threats against less well-known journalists and activists in the
depths of Indonesia could be more severe.
Terrorism and radicalism
Social media and private chat groups are intensely used in Indonesia to disseminate
radical ideology, to recruit members and terrorist fighters, and to promote violent
extremism.
Research from the Institute for Policy Analysis and Conflict (IPAC) showed that one of the
most serious challenges to combat violent extremism content online is the absence of
shared understanding on how to differentiate between intolerant but legitimate political
opinion and ethno-religious hate speech that contains potentially terrorist content and
violent extremism.18 There is no shared understanding among civil society groups, the
government, and platforms. According to the study, while there could be agreement
around what constitutes violent extremist content, it remains much more challenging to
draw a clear line between ethno-religious hate speech and violent extremism.
For example, research from Bhinneka Kultura Nusantara, a research group focused on the
diversity of society in Indonesia, identified 37 social media accounts that are mostly
affiliated with religious groups that produce and promote narratives of exclusionary and
discriminatory parenting. The two main narratives promoted by those religious groups are
‘masuk surga sekeluarga’ (reach the heaven with all family members) and ‘bangun
peradaban Islam’ (build Moslem civilisation). While the first narrative insists on traditional
roles within the family, the second leaves no room for tolerance towards other diverse
religious groups. While focusing on the family unit, these two narratives contradict the
principles and values of democracy. They might barricade rooms for dialogue between
diverse groups and shut down the voices of minorities. While such narratives do not
explicitly encourage violent extremism, the researchers argued that such content nurtures
27
violence at the family level in Indonesia. In addition, family and religion-based violence are
potentially interconnected. This can be seen in the cases of family suicide bombings that
have happened several times in Indonesia. The latest was completed by a couple of
spouses at Katedral Makassar Church, South Sulawesi, in March 2021.
Online gender-based violence
The pandemic has forced millions of people to go online. However, a spike in online
violence towards individuals based on their gender identity or sexual orientation
corresponds to the growing number of Internet users in Indonesia. The National
Commission on Violence Against Women (Komnas Perempuan) recorded 940 reported
online gender-based violence cases in 2020, an increase of 241 cases from 2019. The
Legal Aid Foundation of the Indonesian Women’s Association for Justice (LBH APIK) dealt
with 307 cases in 2020, while before the pandemic, it handled 17 cases in 2019 alone.
Moreover, during 2019, the Digital At-Risks (DARK) Subdivision of SAFEnet assisted 45
victims of online gender-based violence, whereas it received 169 filed cases from March–
June 2020. Meanwhile, many other online gender-based violence victims are reluctant to
report their cases – some might not know the procedures, while others feel uncomfortable
or traumatised, and can even distrust law enforcement agencies if they need to report their
cases.
A report from SAFEnet specifies the various categories and activities that can be included
as online gender-based violence, namely privacy infringement, surveillance and stalking,
reputation damage, online and offline harassment, online and offline threats, and cyber
attacks. These activities are typically directed at someone who is involved in an intimate
relationship, public profiles (such as activists, journalists, researchers, artists), and also
survivors of physical attacks.19
While online gender-based violence content may not always cause casualties, the victims
may experience immaterial losses, such as the loss of privacy or self-confidence. The
victims also have to bear the stigma that can cause them to lose their jobs, relationships,
and future lives. ARTICLE 19 has put forward recommendations on how social media
28
platforms have a role to play in addressing gender-based harassment and violence against
women on their platforms.
Content moderation dynamics in Indonesia
Content moderation decisions in Indonesia
The previous section showed that there are serious content moderation concerns in
Indonesia. Building upon this finding, this section highlights that the government and civil
society groups have been battling to push platforms to take local context into
consideration in their content moderation decisions. The government acted through
dialogue with social media companies, also sometimes blocking platforms, and generally
through content-related regulations.20 Leading civil society organisations have facilitated a
dialogue between platforms and the wider society.
To explain the above point, it is helpful to explore the dynamics of content moderation
practices in Indonesia.
After the 2017 Jakarta gubernatorial election and in preparation for the 2019 Presidential
election, the circulation of hoaxes, hate speech, and extremist content had become a
matter of serious concern. Previously, the government had relied on public authorities and
the public to report problematic content, while also blocking websites that were deemed to
be harmful to the public. As the role of social media became more central in Indonesia, the
government started to reach out to platforms to ask for their help in tackling problematic
content.
It was difficult for the Ministry of Communication and Information Technology (MCIT) of
the Republic of Indonesia to reach out to US-based social media offices and to establish a
dialogue and a speedy review process. The Communication Minister Rudiantara publicly
complained that Facebook was slow in responding, as it needed to discuss requests
coming from the government with its legal team at headquarters. Additionally, he doubted
that the US-based policies of the platform could understand the national context. In an
interview with the media, Rudiantara stated, ‘I respect their policies, but this is happening
in Indonesia, so you should follow the rules here.’
29
However, in its requests to platforms, the government made no difference between

legitimate (although offensive) political opinion and criminal incitement and violent
extremism in the context of Indonesia – and platforms were only willing to comply with
requests that were aligned with their global community standards.
MCIT then leveraged its efforts to tackle the proliferation of problematic content through
blocking. Rudiantara said, ‘Actually, Indonesia doesn’t intend to block, but if social media
are outrageous, then we can close.’ This step was finally taken by the government in July
2017. It blocked the encrypted chat app Telegram after the owners did not respond to
MCIT’s request for removal. Many private extremist groups make use of Telegram,
especially during the 8–9 May 2018 prison riots during which terrorist suspects detained
at the Brimob headquarters went on a rampage, killing at least five police investigators
and one prisoner. The blocking acted as a wake-up call for other tech giants to respond to
the government’s content moderation requests.
In order to facilitate coordination between the government and the social media
headquarters, MCIT also pushed big platforms to establish local representation in the
country. Additionally, platforms and MCIT launched the national trust-related programmes
that involved civil society groups: YouTube’s Trusted Flagger, the trusted partners initiative
of Facebook, and Tik Tok’s Child Safety Partner (see further explanation on the involved
civil society groups in the Analysis of stakeholders). Twitter’s Global Trust and Safety
Council invited two Indonesian civil society organisations (ICT Watch and the Wahid
Institute) to join the Council in 2016 (SAFEnet and ECPAT21 Indonesia joined the Council
later). This initiative ensured a special line of communication for civil society groups to
flag any problematic content to the social media representatives in the country.
Following the 2018 Cambridge Analytica scandal that reportedly impacted Indonesia, there
has been increasing pressure from the authorities to hold Facebook accountable. In 2018,
the Parliament and the government called Facebook executives to testify in a public
hearing. Since then, the government has become more aggressive in regulating social
media content. While the spirit is to protect society from the proliferation of problematic
content, the content-related regulatory efforts in Indonesia have not been able to strike the
30
balance between the protection of freedom of expression and the safety of individuals and
the public. As a result, online content and Internet-related regulations in Indonesia have
failed to uphold international standards on freedom of expression.
Generally, international standards stipulate that any restriction must be:
• Prescribed by law: any restriction must be formulated with sufficient precision.

Overbroad restrictions are not allowed.
• In pursuit of a legitimate aim: restrictions shall only be permitted for (a) the respect of
the rights or reputation of others and (b) the protection of national security or of public
order, of or public health or morals.
• Necessary and proportionate: restrictions must have a direct and immediate

connection between the expression and the protected interest. Moreover,
proportionality means that the restrictions must be specific, tailored, and the least
intrusive means to achieve the same limited result.
A study by the Institute for Policy Research and Advocacy (ELSAM) noted that the
regulations to limit online content in Indonesia have not been aligned with the international
standards on limitations on freedom of expression.22 Indonesia’s regulatory efforts on the
limitations on the right to freedom of expression typically contain an overly broad
definition of negative content and the absence of detailed procedures to implement the
restrictions in Indonesia.
Articles 27–29 of the ITE Law defined in broad terms the categories content that are
prohibited online.23 The Regulation of the Government of the Republic of Indonesia
Number 71 of 2019 on Electronic Systems and Transactions (PP 71\2019) and the
Communications and Information Ministerial Regulation No. 5/2020 on Private Electronic
System Operators (MR 5\2020) are further attempts to regulate online content. However,
those regulations are characterised by the overbroad definition of prohibited online
content and the absence of any detailed procedures.
31
The overly broad definition of prohibited online content in Indonesia’s Internet-related

regulations, along with the compliance of platforms with the government’s requests for
securing their presence and expansion in the country, may further undermine the
protection of freedom of expression in the country. For example, in the first year of
TikTok’s presence in Indonesia, MCIT blocked eight domain name system (DNS) servers
of TikTok on 3 July 2018, on the basis that the service hosted pornography, immorality,
religious harassment, and other negative content. TikTok immediately met with MCIT and
Indonesian child protection officials on 4 July and pledged to collaborate with the
Indonesian Government on removing ‘negative content’. After the block was lifted on 10
July, the next day TikTok stated its commitment to provide a secure, healthy, and good
quality platform for Indonesia, a very important market for TikTok. Meanwhile, Indonesian
users and society remain uninformed about the definition and scope of ‘negative content’
that is prohibited by the government and social media.
MR 5\2020 in particular requires digital intermediaries, such as social media platforms, to

ensure that their ‘electronic systems’ do not contain prohibited electronic information or
documents or facilitate the spreading of such prohibited content (Article 9(3)). In the
meantime, Indonesian users are uninformed about how and to what extent platforms rely
on international standards on freedom of expression and on a satisfying assessment of
the local context in their content moderation decisions.24
The lack of transparency and accountability of social media platforms are apparent in the
interactions between civil society groups and platforms. During the interview process, civil
society groups who are the trusted partners of platforms, criticised the platforms’
responses to their requests. They viewed these platforms as being more responsive to
disinformation and hate speech content, but not to online gender-based violence. The first
two categories of problematic content are more pervasive, and platforms have received a
lot of pressure from governments and the world to take responsibility for tackling them.
Meanwhile, platforms have not put similar attention and effort into the emerging issue of
online gender-based violence. So, the weight of one problematic piece of content against
another is not the same in the eyes of platforms.
32
Ellen Kusuma, the Head of Gender-Based Violence Online Subdivision of SAFEnet, argued
that platforms did not have consistent mechanisms in addressing the concerns raised by
her team. While on one occasion, platforms might attend to their requests nimbly, on
another it could take more than a week for platforms to respond to their requests.
Furthermore, there are no clear rules on the types of problematic content and appropriate
content moderation measures to be taken by platforms on the issue. Meanwhile, she
emphasised that the victims of online gender-based violence need a quick and appropriate
response from platforms before the problematic content that endangers them escalates.
Furthermore, as previously presented in the case studies, some interviewees conveyed

that platforms sometimes showed some resistance to their requests, particularly on ‘grey-
area’ content. Platforms often argue that there is no evidence that the relevant
problematic content is escalating. However, in consideration of the quick amplification
from ‘grey-area’ content to real-world violence, as evidenced in case study 1 (the post-
Presidential election riots in 2019), the question arises of whether platforms are able to
listen to signals and recommendations from local actors to assess the local context.
Another case happened to the fact-checkers of the Indonesian fact-checking organisation

Mafindo (the Indonesian Anti-Slander Society) who criticised the decisions of a platform
not to close social media accounts that spread false information on a continuous basis.
The founder of Mafindo, Harry Sufehmi, conveyed that as an organisation that supports
freedom of expression and after carefully paying attention to the repeated violations made
by the same accounts, Mafindo saw the potential damage of the hoaxes produced by
those accounts towards the polarisation in society which in the end may cause casualties.
They then contacted the related platform several times, only to receive a diplomatic reply
that restated the platform’s decision. Moreover, while Indonesian fact-checkers have
urged the platform to display clarified facts side by side with hoaxes, it preferred to reduce
the visibility of those identified hoaxes. Fact-checkers wanted to empower the public to
differentiate reliable from false information, but they presumed that the platform was
reluctant because such actions might hurt the platform’s public image and reputation.
33
A recurring pattern identified in all of these above views and experiences of civil society
actors dealing with platforms is that platforms largely hold the decision-making power in
the negotiation process. The following examples show that a closed coordination and
dialogue between platforms and Indonesian stakeholder groups is needed in order to
make content moderation decisions that appreciate the local context in Indonesia.
A lack of understanding about Indonesia is apparent in the content moderation decisions

of platforms that use mass flags from individual users. Whenever there are mass requests
from users directed towards the same account, platforms tend to move quickly,
responding to and even agreeing to these requests. For example, in June 2021, civil
society groups noted that Instagram took down content from at least two accounts of
activists that advocated for corruption eradication efforts in Indonesia based on users’
reports. Those activists received a notification that their posts contained incitement to
violence and thus infringed on community guidelines. Civil society groups highlighted that
these mass user reports and hacking attempts against activists and journalists were part
of the counterattack by corruptors to prevent the strengthening of the Corruption
Eradication Commission (KPK).
On another occasion, platforms also responded to mass flagging by closing social media
accounts of paid pro-government influencers. In August 2021, the Twitter account of Ade
Armando was suspended twice without a clear explanation. While there is no evidence,
this communication lecturer is perceived by society as one of the key figures of the pro-
government influencers as he often posts content that could steer public debates.25
Interviewed by the press, Armando stated that he did not know which of his Twitter posts
were reported, but he suspected that there were cyber teams that reported his account to
Twitter.
Siti Cotijah from the Society Participation Division of the National Commission on Violence
Against Women (Komnas Perempuan) complained that the anti-violence education
material that the organisation livestreamed on YouTube was taken down by the platform.
She explained to the researcher:
34
“The live streaming was 2 hours long from 10:00 to 12:00. However, in the first hour,
YouTube accidentally cut off the live streaming and deleted the content of the first hour.
Then we continued our live streaming again by changing the title of the streaming event
using alphabets mixed with numbers (note: from violence into v10l3nc3 or k3k3r454n in
the Indonesian language). After the live event was done, we renamed the title using the
correct spelling.”
Cotijah said that they tried to contact YouTube, but even an established public authority
like themselves found the reporting and appeal mechanisms challenging and one-sided.
They did not know the reason behind the take-down decision, but they presumed it was
because their video used the word ‘violence’ in Indonesian language (‘kekerasan’).
Consequently, the automated content moderation system ‘thought’ that the video
promoted violence.
On the other hand, the Head of External Communication of Arus Pelangi, a civil society
organisation that promotes the protection of LGBTIQ+ rights, mentioned during the
interview their observations about content posted on Instagram by an organisation that
advocates against the protection of these minority groups: those posts stayed on the
platform, presumably because the uploader typed ‘violence’ as ‘v10l3nc3’ (‘k3k3r454n’ in
the Indonesian language) in the content.
More importantly, all of these examples point out that platforms should be in close and
meaningful coordination and dialogue with Indonesian stakeholder groups in order to
make content moderation decisions that take Indonesia’s language and context into
consideration in their content moderation decisions. This need is apparent and important
in the protection of credible online social movements and actors for strengthening public
participation and democracy in Indonesia.
This can be seen in the case of attacks towards online social movements in Wadas Village
in the context of a conflict between the police and residents who reject the Bener Dam
construction and mining plans in Wadas Village in the Bener District of the Purworejo
Regency, Central Java. The General Secretary of SAFEnet Anton Muhajir reported that
SAFEnet received reports from Wadas residents and youth activists who sided with the
35
residents that their Twitter accounts were suspended due to mass flags. They appealed to
Twitter for the decisions. Meanwhile, SAFEnet supported them by corresponding with
Twitter to explain the credibility of their profiles and digital activism activities. Within a few
days, Twitter reactivated their accounts and even verified the account of Wadas
Melawan\Wadas Fights Back with a blue sign.
Actors who initiate requests for content moderation
The practice of filing content moderation requests has been coming from all stakeholder
groups in Indonesia. The Indonesian Government is active in sending content moderation
requests to social media platforms. Some leading civil society organisations are trusted
partners of several social media platforms, allowing them a direct communication channel
to raise urgent issues to platforms.
MCIT, particularly through its Directorate of Informatics Application Control under the
Directorate General of Informatics Application (Aptika), is the lead ministry in dealing with
platforms with regard to content moderation. The Directorate of Informatics Application
Control works on the basis of reports on problematic content produced by its Information
Crawler Machine and team, as well as reports filed by the public and governmental
institutions.
MCIT proactively detects content violations using the Information Crawler Machine ‘Cyber
Drone 9’. It is a crawler system driven by artificial intelligence (AI) to detect content
violations. A team of around 100 people monitors the system and reviews the material it
flags for blocking. The blocking is then done by electronic service providers, including
social media platforms. Civil society groups have criticised the legitimacy and accuracy of
the Cyber Drone 9 because it may result in the blocking of legitimate content and over-
blocking.
The team also works based on reports filed by the public on a site called aduankonten.id.
The public could report alleged problematic content they encounter on sites, social media,
mobile apps, and software. Furthermore, in September 2021, MCIT released a site,
instansi.aduankonten.id, to enable governmental agencies to be more effective and
36
coordinated in reporting problematic content.26 Upon receiving requests and

recommendations from the public and state agencies, MCIT together with other
governmental bodies will verify the reports. MCIT will file content moderation requests to
social media platforms or block the sites that are verified to infringe laws and regulations
in Indonesia.
During election periods, MCIT works particularly closely with the Elections Supervisory
Agency (Bawaslu) and the General Elections Commission (KPU) to govern content
infringements related to elections. They signed the Memorandum of Action (MoA) in 2018,
2019, and 2020 to combat disinformation and hate speech. The agreements address the
responsibilities and roles of each institution and also the coordination and exchange of
data and information on online content among the three entities to tackle problematic
content related to elections. The agreements also stated the need to increase the capacity
of each institution to monitor online content and to empower the public to use the Internet
wisely. Following the signing of the MoA, each year, social media providers in the country
make joint declarations of support for the government’s efforts to eradicate hoaxes and
hate speech.
Bawaslu became involved in social media monitoring after the signing in 2018. Bawaslu
operates in collaboration with MCIT and KPU to proactively monitor online content related
to elections. Bawaslu also opened a helpline for the public to raise their concerns via email
and WhatsApp over alleged problematic online content related to elections. Additionally,
Bawaslu can directly request platforms to take down problematic content or accounts and
also ask the police to enforce any online content with election offences.
In particular, Bawaslu holds bilateral partnerships with Facebook and Google to counter
hoaxes. With Facebook, Bawaslu has developed a special line of communication to flag
infringements to the platform. Facebook and Bawaslu also conducted a series of digital
literacy training activities for election supervisory agencies at the provincial and regional
levels. Furthermore, Bawaslu and Facebook also organised a roundtable discussion with
KPU and MCIT to agree on shared perceptions in handling ‘negative content’ on social
media in the 2019’s election.
37
Google also trained Bawaslu’s officials to do reporting on problematic content on

YouTube. Bawaslu’s YouTube account is synchronised for reporting directly to Google.
Bawaslu and Google set up public service announcements to campaign against
disinformation and hate speech in the YouTube ad. In addition, Bawaslu and KPU together
with Google and some civil society organisations developed a programme called ‘Pintar
Memilih’ (clever to choose). They developed a website (pintarmemilih.id) that contains
information related to elections and delivered digital literacy campaigns to eight
campuses in Indonesia.
Non-governmental organisations and actors also play an active role as content flaggers
dealing directly with platforms. As there are trust-related programmes and partnerships
between social media platforms and Indonesian civil society groups (the safety trusted
partner initiative of Facebook for online safety issues, YouTube’s Trusted Flagger,
Twitter’s Trust and Safety Council, and Tik Tok’s Child Safety Partner), the latter have
special lines of communication to flag any problematic and urgent content to the social
media companies.
The involved civil society organisations are mostly those who work on digital rights in the
country, namely ICT Watch, SAFEnet, and Mafindo. There are also civil society
organisations who work with groups impacted by content moderation issues that were
invited to join social media trust-related initiatives, such as LBH APIK, the Wahid Institute,
Yayasan Cinta Anak Bangsa (YCAB) Foundation, and ECPAT Indonesia (see the Analysis
of stakeholders for further explanation on the social media affiliations of these trusted
partners).
Although a Meta representative claimed in the public discussion conducted by SAFEnet

that they have 12 trusted partners in Indonesia, the researcher was only able to identify
and reach out to some, since the list of trusted partners is not publicly available. This
confirms the findings of academic research with regard to the transparency and
inclusiveness issues of trust-related partners of social media.
Aside from the content flagging requests made by each stakeholder group, Indonesia also
has a multi-stakeholder WhatsApp group to manage Covid-19 related disinformation. This
38
group consists of governmental actors, representatives from social media platforms, civil
society groups, and health communities. Whenever there is any substantial disinformation
regarding Covid-19 in Indonesia, this group will discuss and propose any necessary
actions to manage that content. One of the group’s members conveyed that the group was
able to manage individual cases of Covid-19 hoaxes. This is a loose initiative and more
should be done, for example, to have clear rules and governance structures, to enable the
group to tackle hoaxes in Indonesia more massively while also ensuring compliance with
international standards on freedom of expression.
Individual users in Indonesia also actively file reports whenever they find problematic
content to platforms. However, some incidents show that the reporting mechanism can be
misused by some parties, both pro- or anti-government volunteers or paid influencers or
entities, to silence others who hold different views.
Access for individual users to an internal complaint mechanism
Most of the interviewees agreed on the need for platforms to provide simple and
interactive complaint mechanisms to protect the freedom of expression and safety of
individual users at large and, more importantly, of the vulnerable groups.
As previously explained, even an established public body such as the National

Commission on Violence Against Women (Komnas Perempuan) found that it took a lot of
effort and time to understand the process and be able to file a complaint.
Ellen Kusuma of SAFEnet mentioned in the interview that users who experienced online
gender-based violence struggled to navigate the internal complaint mechanisms of
platforms. Leading civil society organisations working on digital rights play a key role in
empowering and directing victims to report their cases to platforms.
Platforms should do more to provide accessible complaint mechanisms and they should
provide more educational materials for users to understand how to use these complaint
mechanisms.
39
Facebook’s Oversight Board
Some leading civil society actors during the interviews stated that they are aware of the
existence of Facebook’s Oversight Board. They are hopeful because the Board has the
mandate to review Facebook’s decisions. However, they asserted that they have not filed
any cases with the Board.
Meanwhile, the first transparency report from the Board shows that less than 8% of
submissions to the Oversight Board came from the Asia-Pacific region. This small number
may reflect the level of awareness of users in the region and, particularly in Indonesia, on
their rights to freedom of expression and online safety in relation to the responsibilities of
platforms to ensure those rights through appropriate content moderation mechanisms.
Furthermore, there is a widespread perception in the country that platforms have the
legitimacy, authority, and right to govern their spaces with their own rules. A study, for
example, captures how the public sector in Indonesia relies heavily on social media to
spread information to the public, and therefore they must follow the rules of social media.
Meanwhile, users, especially those with a relatively large number of followers, show an
attitude of compliance with the rules of platforms because they do not want their
accounts to be suspended. So, whenever there are users impacted by the content
moderation decisions of platforms, there are only a few that know they can appeal to
those platforms, not to mention appeal to a more high-level entity like the Facebook
Oversight Board.
Statistics and data on content moderation
There are several studies that highlight concerns and challenges with regard to content
moderation practices in Indonesia. For example, the study from the think tank Center for
Indonesian Policy Studies (CIPS) examines the impact of Indonesia’s content moderation
regulations on freedom of expression. Moreover, the study by IPAC noted the challenges
in governing and managing online terrorism and radicalism content. Civil society
organisations, such as ELSAM and SAFEnet, have issued joint statements criticising the
40
mass flaggings and content moderation decisions of platforms that stifle legitimate
content advocating for corruption eradication efforts in Indonesia.
However, there is limited to no data or research by Indonesian or international actors that

monitor online content moderation practices in Indonesia, except for the transparency
reports released by social media platforms (see the Recommendations: this report
proposes a research capacity-building programme for potential members of the coalition
to be able to monitor content on social media).
Level of information in the transparency reports
Contrary to the recommendations from the international civil society,27 transparency

reports from social media platforms have so far not been able to reassure users that
platforms include human rights and local context considerations in their content
moderation decisions and practices.
Transparency reports typically provide large amounts of information on community

standards enforcement, government requests for content restrictions, intellectual property
removal requests, and government requests for user data. However, the policy brief
produced by UNESCO to promote transparency and accountability in the digital age shows
that transparency reports produced by Internet companies contain significant gaps and
cover different issues in different ways.
This report found at least two gaps. First, transparency reports only portray the end results
of content moderation decisions but do not show how platforms conduct content
moderation processes, especially regarding the relationship between ‘grey-area’ content
and the local context. That way, users cannot be assured that platforms have carefully
undertaken an assessment of human rights implications in the local context. How do
platforms incorporate an understanding of a local context in their AI-based content
moderation systems, the intervention of human content moderators, and their decision-
making processes? With whom do platforms consult in order to understand the local
context before making content moderation decisions?
41
Second, some civil society actors interviewed for this project also urged platforms to
provide transparency on campaign advertising, in terms of disclosing information about
what kind of advertisements appear and are distributed to which users, who pays for them
(advertiser’s name, phone number, email, website, address), and also the commitment
from platforms not to display paid campaign advertisements that do not contain
disclaimer information.
In responding to the various expectations and needs of wider users that have not been
reflected in the current state of transparency reports, a representative of digital platforms
in Indonesia highlighted that transparency reports are useful for certain ends and that
other tools, particularly sustained dialogue with local trusted partners and multi-
stakeholders, can be better suited to achieve other objectives. In their view, due to their
quantitative nature, transparency reports are best to gauge the volume and characteristics
of the content removal requests and, in their current shape, not the broader contextual
challenges surrounding the policy enforcement. They considered that having a sustained
dialogue with a locally-relevant multi-stakeholder group could complement the efforts in
providing transparency to the public.
Humans or machines: Who is in charge of content moderation?
All of the civil society actors interviewed in this research stated that they have only been in
contact with representatives of social media companies in the country and or in the Asia-
Pacific region. As described in previous sections, they have been corresponding with
representatives whenever there are urgent content moderation requests that need
immediate support and responses from platforms.
The Executive Director of SAFEnet, Damar Juniarto, noted that there are three companies
outsourced to undertake the content moderation review processes globally. He conveyed
that there should be more engagement between Indonesian civil society groups and the
human content moderators of the country so that they could inform and empower the
moderators with information on the local context behind any emerging problematic
content.
42
A similar conclusion is in relation to automated content moderation. As research shows,

the opacity of an AI-based content moderation system creates Error! Hyperlink reference
not valid.. This report finds there is room for more participation of Indonesian civil society
groups in the development of automated processes, notably in relation to the assessment
of the impact of automated mechanisms of content moderation.
All of the interviewees stated that they have never been informed or involved in the
development process of an algorithmic-based content moderation system of platforms.
The impact of such systems upon Indonesia is generally unknown.
This point was apparent in an interview with the representative of a local trusted partner of
a social media platform. They stated that there was a period of time when their team was
burdened with the need to deal repeatedly with the same problematic content being
circulated through different channels. The related platform then accommodated their
complaints through developments of its automated content moderation mechanisms, but
the platform did not consult civil society and never revealed the content being moderated
by such mechanism. The Indonesian civil society groups were thus not able to assess the
impact of such mechanisms on online content. They could only observe that content from
mainstream media outlets was prioritised in the following days after the changes in the
automated system. Meanwhile, content from less credible channels appeared to be less
visible.
Interim conclusion
This research has shown that in the midst of the diverse cultural context of Indonesia,
growing misuse of social media, and the complexity of ‘grey-area’ content in the country,
there are several main flaws in the current practices of content moderation.
To address these issues, ARTICLE 19 and other civil society organisations have developed
recommendations based on international standards on human rights.28 Of particular
importance for this study is the principle of Culture Competence, set forth in the Santa
Clara Principles, which:
43
“…requires, among other things, that those making moderation and appeal decisions
understand the language, culture, and political and social context of the posts they are
moderating. Companies should ensure that their rules and policies, and their
enforcement, take into consideration the diversity of cultures and contexts in which their
platforms and services are available and used (…), and companies should ensure that
reports, notices, and appeals processes are available in the language in which the user
interacts with the service, and that users are not disadvantaged during content
moderation processes on the basis of language, country, or region.”
In addition, flaws in content moderation identified in this study should be addressed

through the following recommendations:
• Companies should ensure that their content rules are sufficiently clear, accessible, and
in line with international standards on freedom of expression and privacy. It is of key
importance that social media companies’ content rules be made accessible and
available in local languages.
• Companies should also provide more detailed examples or case studies of the way in
which their community standards are applied in practice and conduct reviews of their
standards to ensure human rights compliance.
• Companies should be more transparent about their decision-making processes,

including the tools they use to moderate content, such as algorithms and Trusted
Flagger schemes.
• Companies should ensure that sanctions for non-compliance with their Terms of
Service are proportionate.
• Companies should put in place internal complaint mechanisms, including for the
wrongful removal of content or other restrictions on their users’ freedom of expression.
In particular, individuals should be given detailed notice of a complaint and the
opportunity to respond prior to content removal. Internal appeal mechanisms should be
clear and easy to find on company websites.
44
• Companies should publish comprehensive transparency reports, including detailed

information about content removal requests received and actioned on the basis of their
Terms of Service. Additional information should also be provided in relation to appeals
processes, including the number of appeals received and their outcome.
• Companies should collaborate with other stakeholders to develop new independent

self-regulatory mechanisms, such as a Social Media Council, modelled on effective self-
regulation archetypes in the journalism field.
45
Analysis of stakeholders
This chapter describes the relevant stakeholder groups in Indonesia that work at the
intersection of online freedom of expression and content moderation. They include civil
society groups, media industries, journalists, content creators, social media companies,
public authorities, academics, and think tanks. While it would require further research to
comprehensively map the full list of stakeholders, this section aims at providing a general
overview of the various sizes, capacities, and needs of the initial potential coalition
members, as necessary to support the report’s recommendations.
Overall, one of the key findings from this research is that all of the identified stakeholder
groups and actors would benefit from support in terms of training and information sharing
and also access to international networks. Such support would contribute to empowering
these stakeholders, notably through an update of their expertise on freedom of expression
and content moderation.
Civil society organisations
As an initial remark, it is important to keep in mind that Indonesian stakeholders have

generally been closely engaged with social media platforms. Social media platforms
funded a national digital literacy initiative called Siberkreasi, research activities, fact-
checking initiatives, and educational activities of both civil society organisations and the
government. However, Indonesian civil society actors explained in the interviews that
funding from social media platforms is not their only source of income. Most of the
interviewees also claimed that although they are funded by platforms, they are still able to
maintain a critical distance from the platforms and express their criticism. The critical
views delivered by the interviewees in this report, particularly those who are the trusted
partners of platforms, support this position.
Indonesia has a large number of civil society actors and coalitions spanning from those
who work at the national level to groups operating at the level of villages. Research from
2011 by Nugroho has shown that there are more than 250 civil society organisations and
46
groups actively using the Internet and social media. These numbers would most likely
have increased in 2022.
There are civil society organisations focused on digital and human rights issues and those
who work with groups that are impacted by content moderation issues. The first group is
skilled in digital rights and issues, and they also have a lot of experience engaging with
state actors and social media representatives (in this study, we call this group the leading
civil society organisation actors). On the other hand, the actors in the second group are
best positioned to provide expertise on the local contexts and understanding the impact of
content moderation on population (for the purposes of this study, we call these
organisations peripheral).
In the first category, there are leading civil society organisations that work on digital rights,
who are the official trusted partners of social media platforms, such as ICT Watch,
SAFEnet, and Mafindo. ICT Watch is a Trusted Flagger for YouTube29 and part of Twitter’s
Trust and Safety Council. SAFEnet is a trusted partner of Facebook and also joined the
Twitter’s Trust and Safety Council. Mafindo is a Trusted Flagger for YouTube and third-
party fact-checker for Facebook.
Based in Jakarta and Bali, ICT Watch is a civil society organisation that focuses on the
issues of digital literacy skills, online expression, and cyber governance, and works on
those issues in collaboration with other stakeholders. Since 2002, ICT Watch has been
active in various initiatives to promote these issues in the country. Together with other
national stakeholders, it initiated the Indonesia Internet Governance Forum (ID-IGF) in
2012 and co-hosted the UN Internet Governance Forum in 2013 in Bali. It has participated
in the initiation of the Digital Literacy National Movement (Siberkreasi). ICT Watch has
received various awards, for example from the UN World Summit on the Information
Society in 2016 and 2017 for its Internet Sehat programme. It also took part in the
formation of SAFEnet.
Established in June 2013, SAFEnet is a network of digital rights defenders in Southeast

Asia. It focuses on three digital rights as its working areas, namely the right to access, the
right to expression, and the right to safety. Its members are spread in 19 cities in
47
Indonesia. It monitors cases related to digital rights violations in their respective area,
conducts digital rights campaigns, and delivers digital rights capacity-building for wider
civil society actors in Indonesia. It partners with regional and international civil society
organisations and also Internet-related companies and associations.
Mafindo is an independent fact-checking organisation. It started as a grassroots

community in 2013 and since then has been growing until being formalised into a
registered civil society organisation in 2016. Besides having full-time staff, Mafindo also
has more than 95,000 members and over 1,000 individual fact-checking volunteers from
across all regions in Indonesia.
There are also leading civil society groups who focus on digital issues but are not trusted
partners of any social media platforms. They are ELSAM, Tifa Foundation, Human Rights
Watch–Indonesia, and Engage Media–Indonesia.
ELSAM is a human rights organisation based in Jakarta. It focuses on the topics of human
rights and technology, business and human rights, fundamental freedoms, transitional
justice, eco-social justice, and human rights education. It conducts research, advocacy,
and training to mainstream the protection of human rights in Indonesia.
Established in 2000, Tifa Foundation is a civil society organisation that promotes the
embodiment of open society in the areas of natural resource governance, human rights,
democracy and social movement, and transparency and accountability in digital data
ecosystem. It has been serving as a grant-making organisation and aims to be a civil
society hub in Indonesia.
Human Rights Watch is an international civil society organisation that investigates and
reports on human rights abuses in various parts of the world. Its local researcher, Andreas
Harsono, has covered Indonesia for Human Rights Watch since 2008.
Engage Media is an Asia-Pacific non-profit that promotes digital rights, open and secure
technology, and social issue documentaries. It has offices in Australia and Indonesia and
several staff in Bangkok, Kuala Lumpur, and Manila.
48
In the second category (civil society groups who work with groups that are impacted by
content moderation issues), this research identified that only some of the groups are
aware of the issues of content moderation. They may be invited to join social media trust-
related initiatives, which means they are prioritised in reporting problematic videos to the
related platforms. For example, LBH Apik is the trusted partner of Facebook. Moreover, the
Wahid Institute and YCAB Foundation are also in the Twitter’s Trust and Safety Council.
ECPAT Indonesia is the partner of several social media safety initiatives: Twitter Trust and
Safety Council, Facebook Safety Trusted Partner, TikTok Child Safety Partner, YouTube
Trusted Flagger, and Internet Watch Foundation.
Since 1995, LBH Apik has worked to promote the realisation of a gender-just legal system
and strengthen the women’s movement in empowering gender-just laws. It has been
providing free legal aid to women and children and assisting cyber-crime victims. It has 16
offices spread in various cities in Indonesia.
Founded in 2004 by the former President of Indonesia Abdurrahman Wahid, the Wahid
Institute is a Jakarta-based civil society organisation that advances the development of
progressive Muslims to promote the creation of democracy, multiculturalism, and
tolerance throughout Indonesia and the world. The organisation is currently led by his
daughter, Yenny Wahid.
The YCAB Foundation is a non-profit social foundation founded in 1999 in Jakarta. Its
mission is to improve welfare through education and inclusive financing by providing
education and economic empowerment for underprivileged youth and mothers.
ECPAT Indonesia is a national network of 22 organisations and two individual members

from 11 provinces in Indonesia. Since 2000, it has been tackling child prostitution, porn,
and trafficking for sexual purpose in Indonesia. In 2005, it collaborated with ECPAT
International, an international civil society organisation operating in 98 countries, on the
issue of child sexual exploitation. In 2012, ECPAT Indonesia became an official member of
ECPAT International.
49
There are also other civil society groups whose focus intersects with online problematic
content and who showed an interest on the topic and being involved in the coalition. This
includes general human rights organisations or organisations focused on LGBTIQ+ rights,
terrorism and radicalism, religious issues, and elections. They are Arus Pelangi; Amnesty
International–Indonesia; Warga Muda\Young Citizens; Paparisa Ambon Bergerak; the
Institute for International Peace Building\Yayasan Prasasti Perdamaian; Democracy and
Electoral Empowerment Partnership (DEEP) Indonesia; Komite Independen Sadar Pemilu
(KISP)\Independent Committee Aware of Elections; and Perludem.
Arus Pelangi is an organisation that encourages the realisation of a society that respects
the rights of LGBTIQ+ people as human rights.
Amnesty International is an international civil society movement with more than 10 million
people in various countries campaigning to end human rights abuses. It has a national
section in Indonesia with around 60 staff working on various human right issues, including
freedom of expression online.
Warga Muda is a youth collaborative network across various ethnic groups, religions,
races, classes, and professions spread across 34 provinces in Indonesia. It aims to create
a structurally and culturally friendly ecosystem for the participation and representation of
young people in the public and private sectors at both the regional and national level.
Started in 2010, Paparisa Ambon Bergerak is an initiative in the form of a basecamp for
various creative communities across racial and religious backgrounds in Ambon, Maluku,
in the eastern part of Indonesia, to gather and share ideas that promote the development
of Ambon. It has received widespread praise for its contributions to the peacebuilding
movement in Ambon through the use of digital and social media to counter false
information.
The Institute for International Peace Building is a non-governmental organisation

established in January 2008 that works to develop an integrated policy and national
strategy to reduce the level of threats from violent groups through dialogue. It focuses on
50
developing and deepening understanding of peace and conflict, political violence,

terrorism and other transnational crimes.
There are at least three election watchdogs in different parts of Indonesia who showed
interest in the development of a local content moderation coalition. Together with eight
civil society organisations they developed the Coalition for Social Media Ethics in
Indonesia, an initiative for producing ethical recommendations to guide the roles of
election candidates\parties\campaign teams, social media platforms, civil society
organisations, and media to ensure responsible election campaigns on social media.
The first election watchdog, DEEP Indonesia, is based in Depok, West Java. It works to
strengthen the principal values of democracy for the realisation of good quality elections.
It has 20 representatives across various districts in West Java. Located in Yogyakarta,
Central Java, KISP is a youth association that promotes democratic values to the public
and monitors election-related issues. It has been operating since 2018. Lastly, Perludem is
an election watchdog based in Jakarta and founded in 2005. It carries out research,
advocacy, monitoring, education, and training in the field of elections and democracy for
policymakers, organisers, participants, and voters. Perludem is currently preparing a
roadmap to secure a peaceful election in the country in 2024.
Most of the local civil society organisations who were interviewed were aware of the
impact of content moderation issues at the local level, but they were not aware of any
means to engage with platforms and seek solutions to the problematic content they are
faced with. The interviews with Moch Edward Trias of Yogyakarta-based election
watchdog KISP, and with Pierre Ajawaila of Paparisa Ambon Bergerak, provided the
opportunity to discuss the situation at the local level. Both of them shared a similar
perception that civil society actors have no bargaining power and that, as a consequence,
platforms will, for instance, take a long time to respond to their urgent requests. They also
conveyed that civil society organisations and actors that are not directly dealing with
platforms on content moderation issues may have less awareness and understanding of
the issues, as well as less resources to work on content moderation issues. Those
organisations and actors might not have enough resources in terms of staff and capacity
51
to constantly monitor online content. They therefore often seek help from the key civil
society actors active on content moderation issues whenever they encounter content that
requires moderation by platforms.
Media industries and journalists
There are at least six associations in the media industry and five journalists’
associations30 in Indonesia. The Alliance of Independent Journalists (AJI) and the
Indonesian Cyber Media Association (AMSI) are the two leading media industry and
journalists’ associations in the field of fact-checking. AMSI was initiated by 26 Indonesian
online media leaders in April 2017 with a determination to strengthen the professionalism,
trustworthiness, and independency of online media in Indonesia. AJI is a journalists’
association that promotes press freedom in Indonesia. It was founded in 1994 as a
resistance from the Indonesian press community against the arbitrariness of the regime.
Together with Mafindo and Google, AJI and AMSI supported the formation of a
collaborative fact-checking project among several online media companies, named
cekfakta. The proposed coalition could engage with AMSI and AJI in its initial formation
as they have experience dealing with false information, and later seek the interests of
other media and journalists’ associations to join the coalition.
Academia and think tanks
There are some think tanks working on content moderation, hate speech, and violent
extremism content, such as CIPS, CSIS, IPAC, and the Indonesian Institute.
CIPS is an independent, non-profit, and non-partisan think tank that promotes social and
economic reforms through research and policy engagement on the topics of food security
and agriculture, community livelihoods, and education. It has published research on
digital-related issues, such as the impact of content moderation regulation on freedom of
expression in Indonesia, the responsibility of digital platforms in Indonesia over online
content, the rights of digital consumers, and personal data protection.
Established in 1971, CSIS is an independent and non-profit organisation that focuses on

policy-oriented studies on domestic and international issues. It delivers research, dialogue,
52
and public debate in the areas of disaster management, economics, international relations,
and also political and social change. It has produced research reports and policy briefs on
digital issues, including online hate speech, digital literacy, data governance, and the
digital economy, to name just a few.
Founded in 2013, IPAC studies the dynamics and prevention of six types of conflict in
Indonesia: communal, land and resource, electoral, vigilante, extremist and insurgent, and
various forms of dispute.
The Indonesian Institute is a centre for public policy research on the topics of economic,
social, politics, and law. Since 2004, it has been operating for the realisation of public
policies that uphold human rights, law enforcement, the participation of various
stakeholders, and the application of democratic governance principles.
These think tanks could assist the proposed coalition with research and monitoring on
content moderation and hate speech trends online. They could also empower the research
capacity of the coalition members to identify any patterns of online problematic content at
the local level. The coalition could also involve Drone Emprit, a big data technology
developed by Ismail Fahmi to monitor and analyse social media traffic.
Furthermore, there are also at least three leading universities that focus on the issues of
human rights and digital technologies. They are the Center for Human Rights Law Studies
of Airlangga University, Surabaya, East Java; the Center for Cyber Law and Digital
Transformation of Padjajaran University, Bandung, West Java; and the Center for Digital
Society of Gadjah Mada University, Yogyakarta, East Java.
Formed in 2009, the Center for Human Rights Law Studies is a unit under the Faculty of
Law, Airlangga University, which conducts research and assessment activities in the fields
of business and human rights, natural resources rights, minority rights and religious
freedom, freedom of expression, academic freedom, and local government and human
rights.
53
The Center for Cyber Law and Digital Transformation is a research centre under the
Faculty of Law, Padjajaran University. It conducts research, education, advocacy, and
socialisation relating to cyber law.
The Center for Digital Society is a research centre under the Faculty of Social and Political
Sciences, Gadjah Mada University. It conducts research, publication, education, and policy
advocacy related to the issues of digital society in Indonesia.
Content creators and users
As illustrated, for instance, by a YouTube video discussing the blocking of TikTok

accounts, it is fairly evident that users generally do not know how to appeal content
moderation decisions or do not have the power to negotiate with platforms: a lot of the
comments under the video aired confusion and frustration as users felt that their TikTok
accounts were blocked without clear explanations.31
This study identified that there are at least two emerging content creators’ associations in
the country that could be a potential member of the proposed coalition. Indonesian
YouTuber Association (AYI) is an organisation for all content creators\YouTubers in
Indonesia. As of February 2022, it had 759 members. Indonesian Santri YouTuber
Association (AYSI) is a creative santri community initiated in June 2021 to develop digital
da’wah (preaching of Islam) through YouTube channels and other social media.32 Their
focus so far is on how to increase their capacity as content creators. However, this
research suggests that there can be an opportunity to involve those associations to
introduce them to content moderation issues and to empower them to join the debates.33
Social media platforms
While this study is focused on content moderation of the five largest social media
platforms in the country, the proposed coalition could also consider engaging with other
social media platforms and chat apps available in the country (see The state of content
moderation in Indonesia). The participation of these players may provide insights on the
various challenges, capacities, and resources they face in conducting content moderation
in Indonesia. An engagement with chat app companies would also shed some light on the
54
intricacies of conducting content moderation on personal chat groups. The companies

could also learn from the needs and challenges of civil society in conducting content
moderation at the local level.
Public authorities
There are several key state organisations that deal with content moderation in Indonesia,
namely MCIT, National Cyber and Crypto Agency (BSSN), cyber police together with the
ministerial and state organisations that intersect with the content being moderated (for
example, health, medicine, finance, etc).
Furthermore, there are also independent public authorities that are involved in content
moderation issues, such as the National Commission on Human Rights (Komnas HAM),
National Commission on Violence Against Women (Komnas Perempuan), and election
providers (Bawaslu and KPU).
BSSN is a government agency that handles information and cyber security issues, namely
the security of Internet protocol-based telecommunication networks and infrastructures.
Komnas HAM is an independent national human rights institution of Indonesia carrying
out studies, research, counselling, monitoring, and mediation related to human rights
issues in Indonesia. Komnas Perempuan is an independent state institution for the
enforcement of human rights of Indonesian women. Bawaslu is the election supervisory
board in Indonesia. KPU is the body that organises elections in Indonesia.
55
Conclusion
Conclusion
This research has shown that social media companies have not conducted their content
moderation practices on the basis of an informed understanding of the diversity of
Indonesian society and the rich history and context of the country. Indonesian users and
civil society groups have not been given enough information, space, or power to contribute
to the decision-making processes of content moderation. This is especially important in
relation to the ‘grey-area’ content, that is, messages that might not, per se, fit within the
categories of prohibited content under the global community standards of platforms, but
nonetheless their massive amplification could lead to real-world violence. Indeed, the
recent history of Indonesia suggests that such ‘grey-area’ content has contributed to
further polarisation and ultimately to violence in the country.
Because of the lack of a transparent and sustainable dialogue between civil society
groups and social media companies, there is no shared understanding of the
consequences that certain content can have in the context of Indonesia and of what the
appropriate content moderation measures should be.
The dynamics of content moderation practices in Indonesia show that the enforcement
processes of community standards should be informed by a solid understanding of both
international standards of freedom of expression and the local context.
Most local civil society organisations interviewed for this research see favourably the idea
of forming a civil society Coalition on Freedom of Expression and Content Moderation in
Indonesia. To guarantee the effective ownership of the coalition by its members, the
process facilitating its creation would have to start with a validation exercise that ensures
potential participants have the opportunity to discuss the findings of the research. The
coalition would act as a bridge to develop and nurture relations and dialogue with social
media companies, certain state actors, and even international organisations on content
moderation and freedom of expression issues. This coalition should consist of various
civil society groups that can provide a balanced judgement about protecting freedom of
expression and the safety of individuals and public in the specific context of Indonesia.
56
Recommendations
Recommendations
This research has shown that in the midst of the complex and diverse cultural context of
Indonesia, growing use and misuse of social media in the country, and the complexity of
‘grey-area’ problematic content in the country, there has been a lack of meaningful and
continuous dialogue between platforms and leading and peripheral civil society groups.
Civil society groups and lay users have been battling individually, instead of coordinating,
against the content moderation decisions of platforms. Most of them do not know how to
appeal against the platform’s decisions. Meanwhile, the leading civil society groups in
their capacity as the official partners of platforms have often felt powerless in the
negotiation process with platforms. Platforms usually hold the final decision-making
power, while not displaying sufficient understanding of the complexity of the local context.
Accordingly, there have been cases of over and under content moderation in the country,
that either hurt freedom of expression or the safety of individuals and public.
When we submitted the idea of a local Coalition on Freedom of Expression and Content
Moderation to the interviewees, most of them responded positively.
To be clear, there is already a number of multi-stakeholder groups and civil society

alliances working on issues of Internet governance, freedom of expression, and social
media ethics in the country, but only few have shown interest, resources, and commitment
to develop work on the issue of the contribution of local actors to content moderation on
social media.
The multi-stakeholder national committee on the ID-IGF is a yearly Internet-related

dialogue forum in Indonesia inspired by the UN IGF. ID-IGF has a multi-stakeholder
committee who is responsible for preparing the conduct of the forum every year. While the
members of ID-IGF work on various digital-related issues ranging from infrastructure, law,
and economics, to social issues related to the Internet, very few of them pay specific
attention to the issue of content moderation.
57
Recommendations
Moreover, civil society coalitions on freedom of expression in the country typically focus
on advocating for government action to protect freedom of expression. The closest one to
the topic of this research is the Coalition for Social Media Ethics in Indonesia (see the
Analysis of stakeholders, particularly civil society groups), but the interview process has
shown that not all organisations in this group are interested in the topic of content
moderation: some members of this coalition preferred to address disinformation on social
media through digital literacy and the creation of counter narratives.
This situation suggests the creation of a new coalition, but one that includes civil society
groups, academia, think tanks, and associations of media industries, journalists, and users
from the existing coalitions that show interest and commitment to focus on the topic of
content moderation. This Coalition on Freedom of Expression and Content Moderation
could start by working with some key actors and organisations from the existing coalitions
while leaving the door open for the other members of those coalitions to join the new
coalition.
The coalition should be designed to unite and strengthen the various civil society
organisations and actors who have been working on, or are concerned with, content
moderation issues. In addition, the coalition should aim at engaging in a continuous and
meaningful dialogue with the more powerful stakeholder groups, state actors, and social
media companies.
In relation to civil society actors who have worked closely with social media platforms,
there is a clear power imbalance between them and the tech companies. Furthermore,
most of the civil society actors interviewed in this research also conveyed their concerns
in relation to the participation of state actors and the potential abuse of their authority to
pursue their political interests in the coalition. Some expressed a clear preference for the
coalition not to include state actors, to ensure that civil society groups can more freely
express their views in the coalition. They also observed that the bureaucratic processes of
governmental institutions would risk slowing down the working tempo of the coalition.
While there is a power imbalance between civil society groups and other more powerful
actors, there is also a culture of dialogue among state and non-state stakeholder groups in
58
Recommendations
the country. This can be seen in the existence of multi-stakeholder groups such as ID-IGF
and Siberkreasi (the national multi-stakeholder digital literacy movement in Indonesia), in
which state and non-state actors work together. Additionally, some civil society
interviewees also mentioned the need to involve governmental actors to increase and
nurture their relationships with the government, as well as to increase the legitimacy of the
coalition and its capacity to be heard by platforms. Some of the informants perceived that
it is beneficial to keep governmental actors informed on the projects of the coalition and
on platform governance debates at large. This would contribute to encouraging state
actors to gain a better understanding of the need to formulate content moderation and
digital platform related regulations that uphold international human rights standards.
The above findings and reasons have led this research to consider the importance for civil
society groups to build a critical but cooperative relationship with the more powerful
stakeholder groups in order to nurture a culture of dialogue, but without being co-opted or
captured by those stakeholder groups. A coalition that consists only of civil society groups
but engaged in close coordination with state actors and social media companies appears
to be the most appropriate approach.
As explained above, the coalition would seek to act as a bridge to develop sustainable
relations between various civil society groups with social media platforms and possibly
governmental actors. This would enable the coalition to play a role in conducting
dialogues to ensure that content moderation decisions will comply with international
standards on human rights while taking into consideration the multi-dimensional
complexity of the local context. The coalition could also seek to help users impacted by
the content moderation decisions of platforms.
In its approach to public actors, the coalition could first explore the development of
relationships with independent public authorities, such as Bawaslu, KPU, National
Commission on Human Rights (Komnas HAM), National Commission on Violence Against
Women (Komnas Perempuan), or from the ministries or state organisations that intersect
with the issue of content moderation (such as MCIT and BSSN). The coalition should seek
59
Recommendations
to stabilise its relationship with those public authorities to ensure the continuity of
institutional cooperation even when progressive individuals leave their posts.
The interviewees also mentioned that the role of the coalition could be expanded to:
• Cooperating with existing coalitions and networks to create a more holistic approach
and ecosystem of a healthy digital sphere in Indonesia. For example, the proposed
coalition could work together with digital literacy groups and initiatives in the country;
• Contributing to more responsible social media campaigns for the upcoming General
Election in 2024;
• Encouraging the adoption of laws and regulations related to content moderation and
social media in Indonesia that comply with international standards on freedom of
expression and other fundamental rights;
• Developing relations between Indonesian users and the Facebook Oversight Board and
its members to bring cases related to Indonesia to the Board;
• Strengthening exchanges and cooperation between Indonesian stakeholder groups and

influential international organisations and key actors working on platform governance
and content moderation related issues;
• In the longer term, the development of the coalition could lead to the creation of a
Social Media Council that would promote the development and enforcement of
community standards of social media platforms and uphold international standards on
freedom of expression while duly taking into consideration the voices of local actors
and the local context in Indonesia.
Sustainability of a civil society coalition
To enable the coalition to fulfil all or some of the above-mentioned potential roles, the
prerequisite is to develop and strengthen the internal structure, knowledge, and
coordination capacity of civil society groups to increase their deliberation and bargaining
60
Recommendations
capacity so they can coordinate their efforts and gain effective bargaining power to deal
with platforms and state actors.
The credibility of the civil society coalition among all stakeholder groups is key for its
success. As a kickstart, this research suggests bringing together all of the civil society
groups and actors identified in this research. ARTICLE 19 together with UNESCO and the
researcher will submit the final results of the study to a validation exercise by the
stakeholders who, by their participation and contribution, will reconfirm their interests and
commitment to join the coalition.
The members of the future coalition could then appoint a leader. Based on the analysis of
stakeholders, SAFEnet, Tifa Foundation, and Perludem appear as having the potential to
lead the coalition. They are all well-regarded local civil society organisations that have a
good track record of engaging with platforms and state actors. SAFEnet has been acting
as a civil society hub and engaging with issues of online freedom of expression with
various lay users, national and international partners. Tifa is progressing with its work on
intermediary liability and data governance. Of particular note, the Executive Director of
Tifa, Shita Laksmi, was an expert staff at MCIT. Perludem has been working on the issues
of election-related disinformation together with election providers and social media
platforms, and also initiated the Coalition for Social Media Ethics in Indonesia.
The coalition would also need constitutive documents, such as a Memorandum of

Understanding (MoU), charter, governance structure, and workplan, to be drafted and
adopted in a participative and transparent manner.
The coalition will need to have the capacity for internal coordination to manage the
diversity of actors, goals, and strategies. Along the way, the coalition should increase its
inclusiveness by involving more civil society actors in order to be perceived as legitimate
and effective. In other words, the internal legitimacy (inclusiveness and
representativeness) of the coalition will result in the perceived external legitimacy and
efficacy of this coalition to the intended audience. In order to manage the potential tension
between inclusiveness and efficacy of the coalition, the application of a new civil society
organisation should be supported by two endorsements from existing coalition members.
61
Recommendations
This is a proposed selection mechanism to ensure that future members have critical and
cooperative engagement with current members. Some interviewees mentioned that they
may decide to join the coalition after seeing the composition and reputation of the
coalition members. Some stated they would be reluctant to join and associate themselves
with the coalition if there are any civil society groups that prefer a more confrontative
strategy.
Subsequently, the capacity of the members of the coalition needs to be reinforced in terms
of knowledge on content moderation systems, governance of platforms, international
standards on freedom of expression, as well as the capacity to research the needs and
challenges of conducting content moderation in the various regions in Indonesia. As this
research has found, there is still limited awareness of content moderation issues in
Indonesia. While the leading civil society groups have knowledge on digital rights and
capacity to research the trend of problematic online content at the national level, they may
not have the expertise on the local context of content moderation issues. Meanwhile, civil
society organisations at the provincial level are familiar with the local patterns of
amplification of problematic speeches, but they may not have the capacity to turn their
tacit experiences into explicit knowledge that could reinforce their position in engagement
with platforms on content moderation.
The coalition and its constitutive process should look at developing the capacity of civil
society organisations that do not work directly on digital rights to monitor the
amplification trends of problematic content (‘grey-area’ content) in their local area, and to
engage with platforms in order to contribute to the adoption of content moderation
practices that respect freedom of expression. The coalition should also strive to link
leading and peripheral civil society groups and support them to develop research on the
progression of ‘grey-area’ content from legitimate (i.e. protected by freedom of
expression) to illegitimate online content, as well as the design of content moderation
practices that are compatible with international standards on freedom of expression.
Coalition members would also benefit from training on strategic communication,
engagement, and negotiation skills in presenting their research and joint positions, so that
62
Recommendations
they can have a stronger position and a more meaningful dialogue with platforms and
state actors.
Creating a coalition
The following sequence of steps can be used to create a Coalition on Freedom of

Expression and Content Moderation:
1. Hold an inception meeting to conduct outreach and to build trust among the potential
members of the coalition:
a. ARTICLE 19, UNESCO, and the research consultant to submit the findings of the
research for validation by potential members of the coalition. By potential members
of the coalition, we refer to all of the identified civil society groups and actors in this
research, especially those who showed concerns on the topic of content moderation
during the interviews.
b. The potential members to introduce and share their concerns and works with regard
to content moderation issues in Indonesia and to confirm their interest and
commitment to join the coalition.
c. The potential members to discuss the vision, mission, and goals (short, medium,
and long-term), and the benefits they might gain from the coalition.
d. The potential members to appoint a specific individual and organisation to be a
focal point (a leader and secretariat) in order to get the coalition up and running. The
appointed leader and secretariat will be in charge of the coordination, engagement,
and the related administrative duties of the coalition (manage documentation of
meetings, manage members database, establish a mailing list or other
communication channels, prepare follow-up meetings, finance, etc).
e. The inception meeting is to conclude the agenda and action plans for the next
follow-up meeting. The leader and secretariat are responsible for the preparation of
the next meeting, such as preparing drafts of an MoU, governance structure,
coalition charter, and working plan of the coalition based on the discussion at this
meeting and then circulate the drafts to the mailing list to seek input, which will be
further discussed at the subsequent meeting.
63
Recommendations
f. The potential members may provide recommendations for other potential key
members that have not been identified in the research. The secretariat shall reach
out to those recommended potential members for the next meetings.
2. Convene follow-up meetings to confirm commitments from potential members and

further discuss and conclude an MoU, charter, governance structure, and workplan of
the coalition.
a. The potential members to develop and discuss the MoU for civil society
organisations or individuals joining the coalition. The MoU will provide the terms of
reference of the coalition and may include the expected objectives and deliverables,
the roles and responsibilities of members, the resource and financial plans (or how
contributions to the coalition can be made by members), work breakdown structure
and schedule, the codes of conduct, independence and conflicts of interests,
procedures for joining and departing, and dispute resolution.
b. The coalition charter may cover the names and branding of the coalition, purposes
and key principles, and vision, mission, and goals of the coalition.
c. The governance structure may include the structure of the coalition (the leader,
steering committee, secretariat, working groups); decision-making processes and
procedures; an dcriteria for leadership, membership, and appropriate rules.
d. The workplan: the participants may discuss a workplan and a roadmap for the
coalition, including a few key issues and objectives to be addressed by the coalition,
and a financial plan. Additionally, they may consider discussing working processes,
including communications and documentation, frequency, and location of meetings
for their workplan.
e. Those organisations or individuals who express an interest to join the coalition,
could then agree on and sign the MoU and other constitutive documents.
3. Announce the coalition: invite platforms and (progressive) state actors in the formal
launch of the coalition. Once a critical number of members have joined, the formation of
the coalition can be made public through press releases (including on its own website
64
Recommendations
and social media handles) and goals communicated widely. This could generate more
public interest and desire for more organisations to join the coalition.
4. After the formation process, the coalition is then ready to undertake capacity-building
and run their agreed working programmes (such as developing joint research and
engagement strategies), and then to monitor and evaluate the progress and impact of
their works.
65
Annex A: Risk analysis

The Coalition on Freedom of Expression Online and Content Moderation emerges as a
unique opportunity for participation and contribution by all the actors and as a mechanism
for meaningful change. The coalition offers a path to consensus on key content
moderation issues – and opportunities to address them. The following table provides an
overview of the potential risks related to the formation and functionality of the coalition,
identified by the respondents, including potential ways to overcome and mitigate them.
Risk type* Description of risk Likelihood** Impact*** Monitoring and mitigation

Finance The funding and Possible Major • The coalition members
sustainability of the to agree on the roadmap
coalition in the future of the coalition,
including plans on
funding sources
Political Government adopts Possible Major • Joint funding
new restrictive laws applications and
and regulations participation of coalition
members in donor
meetings and agenda-
setting process
• The coalition could play
a role in advocating
against such bill
Reputation The failure of the Unlikely Minor • The coalition to have
coalition would reflect appropriate membership
badly on the selection mechanisms
reputation of ARTICLE to ensure inclusiveness
19 in the country and effectiveness of the
coalition (critical yet
cooperative behaviour of
members)
Safeguarding Participants in the Possible Moderate • Liaise with community
coalition could be leaders to prevent
harassed by social harassment
groups • The coalition to propose
and advocate for legal
protection for human
rights defenders in
Indonesia
Stakeholder Some civil society Likely Moderate • Provide training
organisations may
lack knowledge of
freedom of expression
standards
66
Stakeholder Civil society actors are Likely Severe • Ensure that the coalition
already very busy and finds a balance between
spread thin over being a light-weight
multiple projects approach and
effectiveness
Stakeholder Civil society Almost certain Major • Discuss the shared
organisations may vision and mission,
become tired in the goals, working agenda
journey and lack and benefits for the
commitment and involved participants
interest to join the • Keep communication
coalition flow alive in the coalition
Stakeholder Distrust among civil Almost certain Major • Design the appropriate
society organisations governance structure
may prevent an that balances
effective development representation of all
of the coalition groups
Notes:
* The risk type is pre-classified in the following categories: Political, Safeguarding, Stakeholder, Finance,
Compliance, Reputation, Other, Covid-19.
** The risk likelihood is presented on the scale: Unlikely, Possible, Likely, and Almost certain.
*** The risk impact is presented on the scale: Minor, Moderate, Major, and Severe.
67
Annex B: Potential members of the coalition
Annex B: Potential members of the coalition

In spite of the researcher’s best efforts, interviews could not be arranged within the
timeframe of this research.
Organisation Category
Center for Cyber Law and Digital Transformation of Padjajaran University in Academia and think tanks
Bandung, West Java
Center for Digital Society of Gadjah Mada University Academia and think tanks
Center for Human Rights Law Studies of Airlangga University in Surabaya, Academia and think tanks
East Java
House of Pancasila Nationality\Rumah Kebangsaan Pancasila Civil society organisation
ICT Volunteers (RTIK) Civil society organisation
Indonesian Cyber Media Association (AMSI) Media organisation
Indonesian Institute Academia and think tanks
Indonesian Medical Association (IDI) Civil society organisation
Indonesian Santri YouTuber Association (AYSI) Content creators
Indonesian Women Coalition\Koalisi Perempuan Indonesia Civil society organisation
Indonesian YouTuber Association (AYI) Content creators
Institute for Policy Analysis of Conflict (IPAC) Academia and think tanks
Institute for Policy Research and Advocacy (ELSAM) Civil society organisation
Legal Aid Foundation of the Indonesian Women’s Association for Justice Civil society organisation
(LBH APIK)
Papua-Fransiscan Justice, Peace, and Integrity of Creation (JPIC) Civil society organisation
Protection Desk Indonesia\Yayasan Perlindungan Insani Indonesia (YPII) Civil society organisation
Remotivi Civil society organisation
Wahid Institute Civil society organisation
Women’s Crisis Center in Jombang Civil society organisation
Women’s solidarity for humanity and human rights (Spekham) Civil society organisation
Yayasan Cinta Anak Bangsa (YCAB) Foundation Civil society organisation
68
Annex C: Interview sheet
Annex C: Interview sheet

The researcher held interviews with representatives from the following organisations:
Organisation Category
Amnesty International – Indonesia Civil society organisation
Arus Pelangi Civil society organisation
Association for Elections and Democracy (Perludem) Civil society organisation
Democracy and Electoral and Empowerment Partnership (DEEP) Civil society organisation
ECPAT Indonesia Civil society organisation
Engage Media – Indonesia Civil society organisation
Gaya Nusantara Civil society organisation
Human Rights Watch – Indonesia Civil society organisation
ICT Watch Civil society organisation
Independent Committee Aware of Elections (KISP) Civil society organisation
Indonesian Anti-Slander Society (Mafindo) Civil society organisation
Institute for International Peace Building\Yayasan Prasasti Perdamaian Civil society organisation
Maarif Institute for Culture and Humanity Civil society organisation
Paparisa Ambon Bergerak Civil society organisation
Partisipasi Muda\Generasi Melek Politik Civil society organisation
Southeast Asia Freedom of Expression Network (SAFEnet) Civil society organisation
Tifa Foundation Civil society organisation
Warga Muda Civil society organisation
Alliance of Independent Journalists (AJI) Media organisation
Center for Indonesian Policy Studies (CIPS) in Jakarta Academia and think tanks
Centre for Strategic and International Studies (CSIS) in Jakarta Academia and think tanks
In addition, the researcher had the opportunity to present and discuss preliminary findings
with stakeholders:
• At two focus group discussions organised by Perludem on 29 December 2021 and 18

February 2022 as part of a series of discussions to prepare a roadmap for secure
elections in 2024.
• At the public launch of SAFEnet’s 2021 Digital Rights in Indonesia Situation Report on 2
March 2022.
69
Bibliography
Bibliography
AJI, The Alliance of Independent Journalists 2018 Year-End Note: Persecution and
Violence Threaten Journalists, 2018.
Alexandra, L., Satria, A., Suryahudaya, E. G., and Krisetya, B., CSIS National Hate Speech
Dashboard, 2021.
Amalia, M., Esti, K., and Camil, M. R., The industry of political buzzing in Indonesia and its
impact on social media governance: Examining viral tweets, 21 June 2020.
APJII, APJII’s Internet Survey Report 2019 – 2020 (Q2) [Indonesian], 2020.
ARTICLE 19, Hate Speech Explained: A Toolkit, 2019.
ARTICLE 19, Indonesia: Ministerial Regulation 5 will exacerbate freedom of expression

restrictions, 29 September 2021.
ARTICLE 19, Online Harassment And Abuse Against Women Journalists And Major Social
Media Platforms, 2020.
ARTICLE 19, Side-stepping Rights: Regulating Speech by Contract, Policy Brief, 2018.
ARTICLE 19, Social Media Councils: One Piece in the Puzzle of Content Moderation, 2021.
ARTICLE 19, Watching the Watchmen: Content Moderation, Governance, and Freedom of
Expression, Policy Brief, 2021.
Bailey, H. and Howard, P. N., Country Case Studies Industrialized Disinformation: 2020
Global Inventory of Organized Social Media Manipulation, 2020.
Bawaslu, Bawaslu, KPU, and MCIT sign MoA to tackle hoaxes and negative content at
regional election, [Indonesian], 31 January 2018.
Bawaslu, The 2020 online election campaign will be supervised more strictly, [Indonesian],
2020.
70
Bibliography
Bawaslu, Together with Facebook and Google, Bawaslu undertook these efforts to tackle
hoaxes, [Indonesian], 29 March 2019.
Beatty, A., Berkhout, E., Bima, L. Pradhan, M. and Suryadama, D., ‘Schooling progress,
learning reversal: Indonesia’s learning profiles between 2000 and 2014’, International
Journal of Educational Development, 85, 2021.
Bhinneka Kultura, Parenting narratives to promote tolerance in families in Indonesia

[Indonesian], 4 December 2021.
Bradshaw, S. and Howard, P. N., The Global Disinformation Order: 2019 Global Inventory of
Organised Social Media Manipulation, Oxford Internet Institute/University of Oxford,
Computational Propaganda Research Project, 26 September 2019.
CSIS Indonesia, Fire in the Husk: The Phenomenon of Hate Speech in Indonesia, 18 August
2021.
ELSAM, Internet Content Governance in Indonesia: Policies, Practices and Problems, 2017.
Fealy, G., ‘Bigger than Ahok: Explaining the 2 December mass rally’, Indonesia at
Melbourne, 7 December 2016.
Fitriani, A., Satria, P. P., Nirmalasari, and Adriana, R. The Current State of Terrorism in
Indonesia: Vulnerable Groups, Networks, and Responses, Centre for Strategic and
International Studies, Jakarta, 2018.
Freedom on the Net 2021. Indonesia, 2021.
Gorwa, R., Binns, R., and Katzenbach, C., ‘Algorithmic content moderation: Technical and
political challenges in the automation of platform governance’, Big Data & Society, 7(1),
2020.
Haristya, S. ‘The efficacy of civil society in global Internet governance’, Internet Histories,
4(3), 2020.
71
Bibliography
Harsono, A. and McMinn, T. I Wanted to Run Away: Abusive Dress Codes for Women and
Girls in Indonesia, Human Rights Watch, 2021.
IPAC, Indonesia and the tech giants vs. ISIS supporters: Combating violent extremism
online, [YouTube], [Indonesian], July 2018.
Johansson, A. C. Social Media and Politics in Indonesia, Stockholm School of Economics

Asia Working Paper Series 2016-42, Stockholm School of Economics, Stockholm China
Economic Research Institute.
Kartikawangi, D., ‘Focus group-based evaluation of social media usage in Indonesia’s

digital government’, Asian Journal for Public Opinion Research, 8(1), 2020.
Khoirina, M. M. and Sisprasodjo, N. R., ‘Media social – Instagram usage and performance
benefit (Case study on housewives online seller in Indonesia)’, International Journal of
Entrepreneurship and Business Development, 2(1), 2018.
Komnas Perempuan, Churches against sexual violence, [YouTube], [Indonesian], 3 July

2020.
Komnas Perempuan, Records of Violence against Women in 2020, 5 March 2021.
Lim, M., ‘Freedom to hate: social media, algorithmic enclaves, and the rise of tribal
nationalism in Indonesia’, Critical Asian Studies, 49(3), 2017.
Maharddhika and Salabi, N. A. Interference to vote rights: Phenomenon and responsibility

[Indonesian], 21 September 2021.
Nugroho, Y, Siregar, M. F., and Laksmi, S., Mapping Media Policy In Indonesia, CIPG, HIVOS,
University of Manchester Business School, Ford Foundation, 2012.
Nugroho, Y., Citizens in @action: Collaboration, participatory democracy and freedom of

information – Mapping contemporary civic activism and the use of new social media in
Indonesia, University of Manchester and HIVOS Regional Office Southeast Asia, 2011.
Organisation for Economic Co-operation and Development, Education at a glance, 2019.
72
Bibliography
Perludem, Political party finance reform in Southeast-Asia, 12 September 2021.
Roeslie, C. K., Data stolen and online gender based violence: The voices of victims that
were underestimated by the police, 5 November 2021.
SAFEnet, 2021 Digital Rights in Indonesia Situation Report, February 2022.
SAFEnet, Public discussion and launch of Digital Rights in Indonesia Situation Report,
[YouTube], [Indonesian], 2 March 2022.
Sastramidjaja, Y., Berenschot, W., Wijayanto, and Fahmi, I., The threat of cyber troops,
Inside Indonesia, 2021.
Sholeh, B., ‘The dynamics of Muslim and Christian relations in Ambon, Eastern Indonesia’,
International Journal of Business and Social Science, 4(3), 2013.
Sinpeng, A., Martin, F., Gelber, K. and Shields, K., Facebook: Regulating Hate Speech in the
Asia Pacific, Department of Media and Communications, University of Sydney, 2021.
Siregar, F. E., ‘The role of the elections supervisory agency to contend hoax and hate
speech in the course of 2019 Indonesian general election’, Padjajaran Journal of Law, 7(2),
2020.
Slama, M., ‘Practising Islam through social media in Indonesia’, Indonesia and the Malay
World, 46, 2018.
Suwana, F., Pramiyanti, A., Mayangsari, I. D., Nuraeni, R., and Firdaus, Y., ‘Digital media use
of Generation Z during Covid-19 pandemic’, Jurnal Sosioteknologi 19(3), 2020.
United Nations Educational, Scientific and Cultural Organization, UNESCO initiates global
dialogue to enhance the transparency of Internet companies, with release of illustrative
high-level principles, 3 May 2020.
van Bruinessen, M., Post-Suharto Muslim engagements with civil society and
democratization, Yogyakarta, Pustaka Pelajar, 2004.
73
Bibliography
Widyaningsih, R. and Kuntarto, K., ‘Family suicide bombing: A psychological analysis of

contemporary terrorism’, Journal of Social Religious Research, 26(2), 2018.
Zuiderveen Borgesius, F. J. et al., Online political microtargeting: Promises and threats for
democracy. Utrecht Law Review, 14(1), 2018.
74
Endnotes
Endnotes
1
An initial list was elaborated through desk research and in consultation with UNESCO with the aim of
speaking with a broad and representative range of relevant stakeholders, including representatives from civil
society, the private sector, public actors, and social media companies. However, in spite of the researcher’s
best efforts, it was not possible within the timeframe of this research to arrange interviews with all the
organisations identified originally.
2
On 30 September 1965, there was a coup attempt; six senior generals and a lieutenant were kidnapped and
murdered. The coup was countered under the direction of General Suharto. There was then widespread
antipathy towards the rise of communism, especially toward the Indonesian Communist Party (PKI). As a
result of this sentiment, then-President Sukarno was marginalised and Suharto came to power. After these
violent events, there was a perception of mass conversions to Christianity and thus fears of Western efforts
to roll back communism and also to combat the political strength of Islam in Indonesia. The events also
deepened the divide between Islam and Communism in Indonesia. The government promoted an official
version of the 1965 events, that PKI was the main actor and a traitor to the country that needed to be
eliminated.
3
However, research shows that Indonesia has failed to regulate the media, in particular in relation to
mitigating the profit-driven logic of the industry. See Y. Nugroho, M. F. Siregar, and S. Laksmi, Mapping media
policy in Indonesia, 2012.
4
It is a YouTube channel that regularly conducts a talk show on various popular topics by inviting guests
ranging from lay people, artists, experts in various fields to high profile guests like Ministers Sri Mulyani, Budi
Sadikin, Luhut Binsar (the last is known to be selective in accepting interview requests from the press).
5
Chat and messenging apps are not covered in this research.
6
M. K. Alfarizi, Internet in Papua is off, causing a digital immigrants phenomenon, who are they?, 8 June
2021.
7
The Santa Clara Principles on Transparency and Accountability in Content Moderation, an initiative by
global civil society organisations, provide a helpful reference to assess the content rules and content
moderation practices of social media companies in light of international standards.
8
Meta, Misinformation. Accessed 4 March 2022.
9
Twitter, Hateful conduct policy. Accessed 23 December 2021.
10
TikTok. Community guidelines. Accessed 23 December 2021. Note: there are four other prohibited
contents, but this study quotes only this type of prohibited content to compare and contrast the rules related
to violence threats in different platforms.
11
YouTube. Hate speech policy. Accessed 23 December 2021.
12
Instagram, Community guidelines. Accessed 23 December 2021.
13
Meta, Violence and criminal behaviour. Accessed 23 December 2021.
75
Endnotes
14
For an elaboration, see the 2018 analysis by ARTICLE 19 of the terms of service and community guidelines
of Facebook, Twitter, and YouTube in light of international standards on freedom of expression and other
fundamental rights.
15
Meta, Violence and criminal behaviour. Accessed 23 December 2021.
16
For a definition of hate speech under international standards on freedom of expression, see ARTICLE 19,
Hate speech explained: A toolkit, 2019.
17
In the toolkit on hate speech, ARTICLE 19 provides guidance on what policy measures state and non-state
actors can undertake to create an enabling environment for freedom of expression and equality that
addresses the underlying causes of ‘hate speech’ while maximising opportunities to counter it.
18
For an analysis of international standards on freedom of expression in relation to national security, see the
Johannesburg Principles on National Security, Freedom of Expression and Access to Information, adopted in
1995, and the Tschwane Principles on National Security and the Right to Information, adopted in 2013. In
particular, these Principles state that restrictions to freedom of expression on the ground of national security
can only be justified if their genuine purpose and demonstrable effect is to protect the country’s existence or
its territorial integrity against the use or threat of force, or its capacity to respond to the use or threat of
force. The restrictions should never serve as a pretext for protecting the government from embarrassment
or exposure of wrongdoing, to conceal information about the functioning of its public institutions, or to
entrench a particular ideology.
19
See also ARTICLE 19’s briefing papers and recommendations on tackling online abuse and harassment
against women journalists.
20
See ARTICLE 19’s analysis of such regulations at Blog: Indonesia’s intermediary regulation imperils
Internet freedom; and Indonesia: Ministerial Regulation 5 will exacerbate freedom of expression restrictions
21
ECPAT (End Child Prostitution in Asian Tourism) is a global network and campaign against the sexual
exploitation of children.
22
The right to freedom of expression is guaranteed in Article 19 of the International Covenant on Civil and
Political Rights.
23
Article 27 prohibits users to distribute and/or transmit electronic information with contents of
pornography, gambling, defamation, and extortion or threats. Article 28 forbids people to disseminate false
information that disadvantage consumers in electronic transaction and also to spread hate speech. Article
29 disallows people to distribute violence threats or scares aimed personally.
24
International standards, the Manila Principles on Intermediary Liability and the Updated Santa Clara
Principles, states that intermediaries, including social media platforms, should not be held liable over the
content produced by their uses. However, they are responsible for undertaking content moderation practices
that uphold international standards on freedom of expression with sufficient understanding of the related
local contexts and culture.
25
I. Arsyam, Who is Ade Armando? US graduated University of Indonesia’s Lecturer but often accused of
being more like a buzzer, 30 June 2021; M. Amalia, K. Esti and M.R. Camil, The Industry of Political Buzzing in
Indonesia and Its Impact on Social Media Governance: Examining Viral Tweets, 21 June 2020.
76
Endnotes
26
T. Gobel, MCIT’s cyberdrone now detects illegal selling in e-commerce, [Indonesian], 24 February 2020; A.
Yuliani, Get to know cyber drone 9, Indonesia’s Internet police, [Indonesian], 5 January 2018.
27
See the Santa Clara Principles (version 2.0).
28
ARTICLE 19’s views on the regulation of platforms are presented in: Side-stepping Rights: Regulating
Speech by Contract, 2018; Watching the Watchmen: Content Moderation, Governance and Freedom of
Expression, 2021; and Taming Big Tech, 2021. See also the Santa Clara Principles (version 2.0).
29
Internetsehat, The role of YouTube Trusted Flagger in cleaning YouTube (part 1 of 2), [Indonesian], 5
November 2018; The role of YouTube Trusted Flagger in cleaning YouTube (part 2 of 2), [Indonesian], 5
November 2018;
30
AWPI, Profile, [Indonesian]; Press Council, Journalists organisations, [Indonesia].
31
See the public comments (in Indonesian language) in this video – S. Alrenzha, 3 reasons and threads of
blocked TikTok accounts ? Tik Tok Block, [YouTube], [Indonesian], accessed 20 January 2022.
32
Santri is a term for someone who attends Islamic education in Islamic boarding schools.
33
See the public comments (in Indonesian language) in this video – S. Alrenzha, 3 reasons and threads of
blocked TikTok accounts ? Tik Tok Block, [YouTube], [Indonesian], accessed 20 January 2022.
77

Indonesia Country Report

Uploaded by

Copyright:

Available Formats

Indonesia Country Report

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Indonesia Country Report

Uploaded by

Copyright:

Available Formats

In partnership with UNESCO

Content Moderation and

© ARTICLE 19, 2022

This work is provided under the Creative Commons Attribution-Non-Commercial-ShareAlike 3.0

1) give credit to ARTICLE 19;

2) do not use this work for commercial purposes;

To access the full legal text of this licence, please visit:

A Data Protection Officer certified by the Association of Data Privacy Professionals

The state of content moderation in Indonesia 15

Annex A: Risk analysis 66

About the project

For the purposes of this report, we rely on the following definitions:

Methodology and structure of the report

The structure of the report is as follows:

The second chapter, Analysis of stakeholders, provides an analysis of the relevant

The population of Indonesia is characterised by social, cultural, and economic diversity.

1. Belief in the one and only God

These principles are often generalised in terms of religious devotion, humanitarianism,

As Indonesia was established as a sovereign nation in 17 August 1945, Pancasila became

The state of content moderation in Indonesia

Overview: Impact of content moderation on peace and stability

This chapter discusses categories of problematic content (disinformation, hate speech,

An outspoken figure with a double minority background (Chinese Indonesian and

Social media platforms operate globally

Figure 1: Screenshot of Facebook’s misinformation guidelines in Indonesia language.8

‘Do not post, upload, stream, or share:

• Hateful content related to an individual or group, including:

– calling for or justifying violence against them.’10

• Encourage violence against individuals or groups based on any of the attributes

In addition to difficulties related to the definition of categories of prohibited content,14 it

In response to a question on whether it is possible for community guidelines to specify

Problematic content in Indonesia

Observing such a critical situation, a representative of an election authority argued with

have enabled the distribution of disinformation as paid political advertisement targeted at

Terrorism and radicalism

Online gender-based violence

Content moderation dynamics in Indonesia

Content moderation decisions in Indonesia

However, in its requests to platforms, the government made no difference between

Generally, international standards stipulate that any restriction must be:

• Prescribed by law: any restriction must be formulated with sufficient precision.

• Necessary and proportionate: restrictions must have a direct and immediate

The overly broad definition of prohibited online content in Indonesia’s Internet-related

MR 5\2020 in particular requires digital intermediaries, such as social media platforms, to

Furthermore, as previously presented in the case studies, some interviewees conveyed

Another case happened to the fact-checkers of the Indonesian fact-checking organisation

A lack of understanding about Indonesia is apparent in the content moderation decisions

Actors who initiate requests for content moderation

coordinated in reporting problematic content.26 Upon receiving requests and

Google also trained Bawaslu’s officials to do reporting on problematic content on

Although a Meta representative claimed in the public discussion conducted by SAFEnet

Access for individual users to an internal complaint mechanism

As previously explained, even an established public body such as the National

Facebook’s Oversight Board

Statistics and data on content moderation

However, there is limited to no data or research by Indonesian or international actors that

Level of information in the transparency reports

Contrary to the recommendations from the international civil society,27 transparency

Transparency reports typically provide large amounts of information on community

Risk type* Description of risk Likelihood Impact* Monitoring and mitigation