Meta Human Rights Report July 2022

JULY 2022
Meta Human
META HUMAN RIGHTS REPORT

Rights Report
Insights and Actions 2020 - 2021
INSIGHTS AND AC TIONS 2020 - 2021 1

About this Report
In March of 2021, Meta adopted its Human Rights the company’s salient human rights risks,
Policy in which we commit to reporting annually defined by the scale, scope, irremediable
on how we are addressing our human rights character and likelihood of impact. Our salience
impacts, including relevant insights arising from assessments are complemented by an additional
human rights due diligence, and the actions we materiality assessment included in Meta’s 2021
are taking in response. This is our first annual Sustainability Report.
report, covering our learnings and progress from
January 1, 2020 through December 31, 2021. This report is inspired by Principle 15 of the UN
Guiding Principles on Business and Human Rights
The scope of this report is Meta Platforms, which makes it clear that companies must “know
Inc. (formerly known as Facebook, Inc.) and and show” that they respect human rights.
our assessment of what we consider to be
Note on Terminology
On October 28, 2021, Facebook, Inc. changed its the application of a policy or procedure to
name to Meta Platforms, Inc. For consistency, this other products or entities. If a policy is labeled
report uses “Meta” to refer to the company both a “Facebook” policy, it does not apply to other

before and after October 28, 2021. References entities, e.g., Instagram or WhatsApp.
to “Facebook” apply only to the social media
platform, not the company as a whole. For example (and in contrast to other Meta
technologies), WhatsApp is an end-to-end
Meta’s Corporate Human Rights Policy applies encrypted messaging and calling application
enterprise-wide. Each Meta product and entity with unique human rights touchpoints. The
has its own policies and procedures with, at report considers the human rights impacts
times, varying human rights impacts. This report of end-to-end encryption. Likewise, this report’s
references a particular policy or procedure discussion of content moderation and related
that may not apply enterprise-wide (e.g., the issues on Facebook and Instagram do not apply to
Community Standards, which apply only to WhatsApp. Unless a policy is specified as applying
Facebook). Further, this report references actions to WhatsApp, it does not apply to WhatsApp.
taken by Meta as a company regarding a specific Further, while many content moderation decisions
entity. Such a statement is not intended to described in this report apply to Instagram
imply that Meta took that action regarding all and Facebook, there are intentional distinctions
entities. No statement in this report is intended in policies and procedures between the platforms.
to create — or should be construed as creating Unless a policy is specified as applying to
— new obligations (legal or otherwise) regarding Instagram, it does not apply to Instagram.
INSIGHTS AND AC TIONS 2020 - 2021 2 ABOUT THIS REPORT

Executive Summary
Background
Born out of the despair of war, international Meta recognizes its potential to impact human
human rights laws and treaties represent rights and has taken steps to demonstrate our
humanity’s best hopes for itself. commitment to human rights principles. We
joined the Global Network Initiative, and made
Meta’s mission is to give people the power to build commitments in our Corporate Human Rights
community and bring the world closer together. Policy to respect human rights as set out in the
We seek to advance our mission in a manner United Nations Guiding Principles on Business
consistent with the promotion of humanity’s and Human Rights (UNGP) and other international
fundamental rights: freedom of expression, human rights standards. This report provides an
privacy, non-discrimination, and more. update on our progress.
INSIGHTS AND AC TIONS 2020 - 2021 3 EXEC UTIVE SUMMARY

Theory of Change
Our human rights policy helps ensure that Meta human rights agenda. Our strategy leverages
incorporates consideration of human rights our mission and values. Simply put, we seek to
principles into our decisions and actions and translate human rights guidance into meaningful
supports the evolution of the business and action, every day.

Salient Risks
Under the UNGP framework, we conducted due company’s products. In this report, we identify
diligence to identify the most salient risks to those risks and share the steps we have taken to
human rights that could be associated with the help manage them.
Right to Freedom of Opinion and Expression
Meta’s platforms advance freedom of expression — which is interdependent with

all fundamental human rights. For example, freedom of expression requires security
of the body, protection from non-discrimination, and privacy. These rights are not
in tension; one secures the other. We manage risks related to balancing freedom of
expression and other fundamental rights on our social media platforms through our
Facebook Community Standards and Instagram Community Guidelines ( see, the
Community Standards Enforcement Reports).
During the reporting period, we updated our policies as part of our continuing effort
to balance these rights in an increasingly dynamic world: including, specifically

referencing human rights principles, clarifying our health misinformation policies,
enhancing our bullying and harassment policy to include stronger protections against
gender-based harassment for everyone including public figures, enacting prohibitions
against certain mass harassment or brigading, and expanding our policies that prohibit
veiled and implicit threats, among others. Meta further provides users with a path to
disagree with content policy decisions, as described in greater detail in this report
under Providing Access to Remedy.
Right to Privacy
The right to privacy is essential to realizing the right to freedom of expression.

On WhatsApp, and expanding to our other messaging apps, we offer end-to-end
encryption, an important privacy and security function to protect people’s private
messages and calls, including journalists and human rights defenders.
We protect users from unlawful or overbroad government data requests via our Meta
Data Policy, our WhatsApp Privacy Policy, our dedicated Law Enforcement Response
Team, and a UNGP-informed tool assessing our responsiveness to data requests from
law enforcement authorities (see, Government Requests for User Data Transparency
Reports). There are many legitimate and lawful reasons why governments request

information from social media companies: international criminal investigations,
anti-terrorism work, prevention of child sexual abuse and other efforts intended
to protect the lives, security and rights of people in their countries. We respond to
such requests for information that are consistent with internationally recognized
standards on human rights, including due process, and the rule of law. If we determine
that a government request is not consistent with applicable law or our policies, or
unlawful (for example, overly broad, or legally deficient in any way), we do not provide
governments with direct access or “back doors” to people’s information and we would
challenge any order that sought to have us redesign our systems to undermine the
encryption we provide to protect people’s data.
Right to Life, Liberty and Security of Person
Our Facebook Community Standards and Instagram Community Guidelines seek to

protect freedom of expression and the salient risks to freedom of life, liberty, and
security of the person, and include rules against Violence and Incitement, Dangerous
Individuals and Organizations, Coordinating Harm and Promoting Crime, Bullying and
Harassment, Human Exploitation, Hate Speech, Violent and Graphic Content, and

Misinformation. We manage risks of abuse of our products for human trafficking and
exploitation via in-product features to raise awareness of trafficking, deter violating
behavior, and offer support to victims. We are working to increase our ability to
identify the illicit actors, networks, organizations, and businesses that perpetrate
these activities and disrupt them accordingly.
We manage human trafficking and exploitation risks in our supply chain through our
Code of Conduct, our Responsible Supply Chain program, and our membership in the
Responsible Business Alliance and the Responsible Labor Initiative ( see, Anti-Slavery
and Human Trafficking Statement).
Moreover, as we detail in this report, protecting Human Rights Defenders is a core

pillar of our Human Rights Policy.
Rights to Equality and Non-Discrimination
Our Community Standards and Community Guidlelines address hate speech, and
we have advertising policies on non-discrimination. We have dedicated Civil Rights
and Human Rights teams; and we have cross-company initiatives for responsible
innovation and accessibility.

Best Interests of the Child
In developing youth-focused features, we are listening to leading experts on child

safety (including the National Center for Missing and Exploited Children).
We published a Parents Guide and other features to increase child and teen safety
on Instagram, and continued our work to fight child exploitation on WhatsApp and
Facebook and Instagram.
Right to Public Participation, to Vote, and to be Elected
Meta is committed to protecting the right to public participation, giving everyone a

voice through our products, and empowering people to vote. Over the last few years,
Meta has made significant investments in teams and technologies to better protect
free and fair elections, including dedicated teams focused on election integrity and
products that bring relevant and reliable voting information to people. We continue
to learn how to treat different types of speech that have the most meaningful impact
on public opinion and how people vote at the polls, such as ads that discuss, debate,
or advocate for or against important topics under our Ads about Social Issues,

Elections or Politics Policy. We fundamentally believe in freedom of expression, while
recognizing the need to set guidelines for ads about elections and to adopt measures
to reduce the spread of harmful misinformation.

Meta’s Human Rights
Work In Pratice
I. Applying Relevant Policies;
Meta’s respect for human rights — and their

underlying principles of equality, safety, dignity,
privacy and voice — is applied through existing
policies across our company, including the
Facebook Community Standards, Instagram
Community Guidelines, WhatsApp Terms
of Service, the Responsible Supply Chain
program, Facebook’s Privacy Principles, our
Law Enforcement Guidelines, our Data Policy,
Responsible AI pillars, and our Diversity and
Inclusion practices, among other policies.

II. Conducting Human Rights Due
Diligence and Disclosure;
We endeavor to assess human rights impacts that
may be linked directly to Meta, take appropriate
action and implement findings, monitor
implementation, and report annually on “relevant
insights” and “actions” under UNGP 21 as defined
in our Corporate Human Rights Policy. From
2020 — 2021, our UNGP-informed diligence
work included:
• Detailed human rights impact assessments,

e.g., the Philippines and End-to-End Encryption;
• Human rights due diligence on rapidly evolving
situations, e.g., Israel-Palestine (to be published
in Q3 2022 in response to the Oversight Board),
and product initiatives, e.g., Ray-Ban Stories;
• Integrated approaches, embedding UNGP
prioritization criteria, and well-defined sets
of human rights standards that allow us to
assess and to identify human rights concerns
for countries at risk and for particularly
vulnerable populations.
INSIGHTS AND AC TIONS 2020 - 2021 8 EXECUTIVE SUMMARY

III. Providing Remedies
for Human Rights Impacts
Users can challenge some decisions regarding

removal of their account or content via the help
center, and appeal up to the Oversight Board.
A detailed human rights report guided the
creation of the Oversight Board. The Oversight
Board consists of rights experts from a variety
of cultural and professional backgrounds. The
Board’s purpose is to: “[p]rotect free expression
by making principled, independent decisions about
important pieces of content and by issuing policy
advisory opinions on [Meta’s] content policies”
(See, Oversight Board Transparency Reports).
IV. Protecting Human
Rights Defenders
We are building on our existing work to protect
human rights defenders and their accounts.
Efforts include: combatting advanced threat
actors targeting them, protecting them from

incorrect content removals using Cross Check,
offering advanced security options, taking steps
to thwart unauthorized access to the accounts
of defenders who are arrested or detained, and
partnering with human rights organizations on
outreach and training. In November 2021, we
launched the Human Rights Defender Fund and
Journalist Safety Initiative, in partnership with
Civil Rights Defenders and International Center
for Journalists, which gives offline assistance
to human rights defenders facing critical threats
and supports new digital security efforts.
V. Governance, Oversight
and Accountability
Meta created a global Human Rights Policy Team

and tasked it with guiding the implementation of
the Human Rights Policy, training stakeholders
on the Policy, and applying the Policy throughout
the company by integrating into existing and
developing policies, programs, and products.
The Team is accountable to the Audit Risk and
Oversight Committee of the Board of Directors.
Meta’s Human Rights Policy and its commitments
to protect human rights are enterprise-wide.

The Road Ahead
Our Human Rights Policy laid a strong basis for Human and civil rights must be at the center
action. In the coming years, we intend to build the of our development of the metaverse.
maturity of our program, function, and reporting.
Protecting rights in the digital space is one of
We continue to strive to better protect the existential challenges for the human rights
rights in countries in conflict and crisis, and movement and for all who care about human
for particularly disadvantaged groups. We rights. In the face of this, we approach our work
understand that our work must address today’s with humility and determination.
technologies and look towards the future.

Table of Contents
Introduction 13
Human Rights and Policy Timeline 14
Part 1: Meta’s Human Rights Commitments 16
Background 17
The Global Network Initiative 18
United Nations Guiding Principles on Business and Human Rights 18
Our Human Rights Vision & Strategy: The Meta Human Rights Policy and Our 19
Theory of Change
From “Know and Show” to “Show, Not Tell” 20
Our Approach to Due Diligence 24
Meta’s Salient Risks 25

How Meta identifies its Salient Risks 26
Salient Risks 27
Part 2: Meta’s Human Rights Policy in Practice 28
Applying Relevant Policies 29

Content Policies 30
Actions with Regard to Salient Risks 31

Community Standards in Focus: Dangerous Organizations 34
Privacy 35
Privacy Program Overview 35
Salient Risks in Privacy 38
Protecting Users from Unlawful or Overbroad Government Data Requests 39
Properly Collecting, Using, Storing, and Deleting User Data 41
Civil Rights 42
Reponsible Innovation and Responsible AI 44
Accessibility 46
Anti-Slavery and Human Trafficking 47
Election Integrity 48
United States Presidential Election 2020 48
Myanmar Election 2020 49
Preparation for Philippines Presidential Election 2022 50
COVID-19 51
Emergency Due Diligence 52
Content Policy Due Diligence 52
Privacy Due Diligence and Data for Good 53
INSIGHTS AND AC TIONS 2020 - 2021 11 TABLE OF CONTENTS

Conducting Human Rights Due Diligence and Disclosure 54
Human Rights Impact Assessments 55
Philippines HRIA 56
Summary Disclosure - India Human Rights Impact Assessment 57
Product-Focused Human Rights Impact Assessments 61
End-to-End Encryption (“E2EE”) 61
Due Diligence Exercises 63
Ray-Ban Stories 64
Integrated Analytical Approaches 65
Countries At Risk 65
Our Program in Action: Case Studies 67
Myanmar 67
Ethiopia 68
Security Policies 69
Combating Emerging Threat Actors 70
Protecting Users Against Espionage 71
Providing Access to Remedy 72

The Oversight Board 73
Code of Conduct 74
Protecting Human Rights Defenders 75

Governance, Oversight, Accountability 77
Governance Model 77
Stakeholder Engagement 79
Program Spotlight: Trusted Partners 81
Engagement with International Organizations 82
A Final Note 83
INSIGHTS AND AC TIONS 2020 - 2021 12 TABLE OF CONTENTS

Introduction
While technology has changed much about This report is inspired by Principle 15 of the UN
our world, the principles of human rights remain Guiding Principles on Business and Human Rights
timeless. As the Universal Declaration of Human which makes it clear that companies must “know
Rights makes clear: “All human beings are and show” that they respect human rights.
born free and equal in dignity and rights” and
“everyone has the right to freedom of opinion It sums up our progress over the past two years,
and expression.” and explains how we think about our human
rights responsibilities and what we do to uphold
When first drafted, these principles were meant them. We seek to embed our commitments in a
to guide the behavior of governments. Today, governance model which supports integration of
they reach far beyond, influencing the policies our human rights work with ongoing activities and
and behaviors of global companies like Meta, policies on civil rights and Environmental, Social
whose decisions can make a real difference in and Corporate Governance (ESG) efforts, as part
how people’s rights are exercised, protected of the company’s culture, governance, decision-
and respected. making processes and communication strategies.

Indeed, the digital sphere represents a unique As a next step, we’ll build on preliminary work
challenge and opportunity for the human rights to carry out a comprehensive human rights risk
movement. Companies, the United Nations, (also known as a salient risk) assessment to help
governments, and civil society must work prioritize our approaches and align our resources
together to protect basic human rights online, so and commitments.
that as we continue to advance and innovate, we
ensure human rights are respected at the frontier Most of all, we strive for our efforts to make a
of new technologies. difference, and for our human rights commitments
to be reflected in today’s technologies and in the
To protect human rights online, there are rarely metaverse to come.
simple answers, only a careful balance to be
struck between competing values. With billions of
people using Meta’s apps and services across the
world every day, it is incumbent on us to have the
right policies in place, and the right processes for
acting on them.
Last year, we launched our first Corporate

Human Rights Policy, which formalizes the
company’s commitment to human rights and
explains how we apply its principles to our
platforms, products, policies, transparency,
and programming efforts.
Under this policy, we committed to publishing an

annual human rights report that would keep the
public informed of our progress.
INSIGHTS AND AC TIONS 2020 - 2021 13 INTRODUC TION

Human Rights and Policy Timeline
2011
Adoption of the UN Guiding Principles
2013
on Business and Human Rights by the
UN Human Rights Council Meta joins the Global
Network Initiative
2016
WhatsApp completes
its rollout of end-to-
end encryption
2018
Public release of Myanmar
independent human rights
impact assessment

2019
Creation of Meta Human
2020
Rights Policy Team
Voluntary disclosure of data to
the UN mechanism investigating
events in Myanmar begins
2020 May.
2020 May.
Announcement of
first 20 members of Public release of executive
the Oversight Board summaries and Meta responses
to human rights impact
assessments for Cambodia,
2020 July. Indonesia, and Sri Lanka
Release of Civil Rights

Audit Report
2020 Oct.
Oversight Board
begins operations
INSIGHTS AND AC TIONS 2020 - 2021 14 HUMAN RIGHTS AND POLICY TIMELINE
2021 Jan.
2021 March. Meta appoints first VP

of Civil Rights
Launch of the Meta
Human Rights Policy
2021 June.
Launch of revised
2021 Sep. Meta Code of Conduct
Meta joins the

UN Global Compact
2021 Oct.
Launch of revised content

policies for public figures and
human rights defenders

2021 Nov.
Publication of the
Civil Rights Audit
progress report
2021 Nov.
Launch of pilot human rights

defender fund and journalist
safety initiative in APAC
2021 Dec.
Launch of Philippines
Human Rights Impact
2022 April. Assessment
Publication of End-to-End
Encryption HRIA
INSIGHTS AND AC TIONS 2020 - 2021 15 HUMAN RIGHTS AND POLICY TIMELINE
PA RT 1:
Meta’s
Human Rights
Commitments
16
I.
Background
In April of 2004 — two months after Meta’s initial a way to encourage rights-respecting practices by
founding — Shi Tao, a Chinese journalist, logged telecommunication and technology companies.
into his Yahoo! account and emailed a document
about the 15th anniversary of the 1989 Tiananmen In 2008, this effort became the Global Network
Square protests to Democracy Forum, a US- Initiative (“GNI”). Directly interpreting the
hosted website. Intense controversy erupted relevant provisions of the International Covenant
when it was revealed that Yahoo!, a US-based on Civil and Political Rights, its members
internet company, acceded to requests by the committed to implementing certain human
Chinese government for Shi Tao’s personal data. rights protections in their decision making and
Tao was convicted of sharing state secrets, and processes. It was — and is — an important
sentenced to 10 years in prison. innovation in the field of business and human
rights. Meta became a GNI member in 2013.
Following this incident, information and
communications technology companies and Building on GNI frameworks, Meta adopted its
their human rights stakeholders aligned on Corporate Human Rights Policy, and created a

a framework to protect people’s rights to freedom policy team that works to integrate human rights
of expression and privacy against overbroad or principles across the company.
unlawful government requests; essentially, building
INSIGHTS AND AC TIONS 2020 - 2021 17 I. BACKGROUND

A. The Global Network Initiative
Global Network Initiative members must commit member companies are required to undergo
to its Principles on Freedom of Expression and periodic independent assessments conducted by
Privacy (“GNI Principles”). These Principles accredited third-party assessors. GNI members
are based on internationally recognized laws are accountable for identifying, preventing, and
and standards and are intended to “provide mitigating risks of overbroad or legally deficient
direction and guidance to the ICT [Information government requests to restrict freedom of
and Communications Technology] industry and expression or violate users’ right to privacy.
its stakeholders in protecting and advancing the
enjoyment of these human rights globally.” Since joining the GNI, Meta has cooperated
fully with two assessments, an assessment in
Meta joined the GNI in 2013, recognizing how 2015—2016, and an assessment in 2018—2019,
“advancing human rights, including freedom of the latter of which found Meta is making “good-
expression and the right to communicate freely, is faith efforts to implement the GNI Principles
core to our mission” and that by joining, we hoped with improvement over time” and commended
to “shed a spotlight on government practices Meta for “strengthen[ing] its systematic review
that threaten the economic, social and political of both privacy and freedom of expression.”
benefits the internet provides.” Meta’s next GNI assessment is ongoing as of this
report’s publication.
To assess whether GNI member companies are

“making good-faith efforts to implement the Further information regarding the GNI’s mission
GNI Principles with improvement over time,” and principles can be found on the GNI website.
United Nations Guiding Principles on Business and Human Rights
In 2011, the United Nations Human Rights Council unanimously adopted the United Nations
Guiding Principles on Business and Human Rights (“UNGP”).
Historically, governments were obliged to protect human rights. The UNGPs: changed this by
articulating a role for businesses in the rights framework. For example, under UNGP II(A)(13),
businesses have a responsibility to:
a. Avoid causing or contributing to adverse human rights impacts through their

own activities, and address such impacts when they occur; and
b. Seek to prevent or mitigate adverse human rights impacts that are directly
linked to their operations, products or services by their business relationships,
even if they have not contributed to those impacts.
Unlike treaty law, the UNGPs are voluntary and non-binding. Companies seeking to follow the
UNGPs signal that commitment in a clear and public policy statement.

B. Our Human Rights Vision & Strategy:
The Meta Human Rights Policy and Our Theory of Change
Meta’s mission — to give people the power to build Declaration of Human Rights, as well as the many
community and bring the world closer together — is treaties, practices, and norms that succeeded it. We
strongly aligned with human rights principles. Our strive for a world where businesses respect rights,
mission guides how we work, what we prioritize, and governments protect rights, and people prosper.
which principles we use to make decisions. In this
sense, our mission is inherently, closely associated Simply put — we seek to translate human rights
with the values and rights of the Universal guidance into meaningful action, every day.

From “Know and Show” to “Show, Not Tell”
UNGP 21 states that companies should have in place policies and processes by which they can
“know and show” their human rights responsibilities, including transparent and accountable
communication to people or groups whose rights may be impacted by a company’s operations.
With this in mind, and in pursuit of meaningful transparency, we worked to build an ethos of
human rights action that is grounded in “show, not tell.” e.g.:
• Newsroom Posts from the Global Human Rights Policy Team on human rights issues, including
Our Commitment to Human Rights, Our Approach to Countries at Risk.
• Our Transparency Center provides a hub for Facebook’s and Instagram’s integrity and
transparency work, acting as a central destination for all updates on how we enforce
Facebook’s Community Standards and how we respond to decisions, recommendations, and
case updates from the Oversight Board.
• We published an end-of-year threat report on six adversarial networks we found and removed
for Coordinated Inauthentic Behavior, Brigading and Mass Reporting.
• We publish a quarterly Community Standards Enforcement Report covering Facebook
and Instagram.
• We semi-annually publish data on how we enforce our Intellectual Property policies.
• We regularly publish data on Government Requests for User Data.

• We semi-annually publish data on Content Restrictions Based on Local Law.
• We semi-annually publish data on Internet Disruptions.
• We report quarterly on Widely Viewed Content Report: What People See on Facebook.
• We publish reports that apply human rights principles:
• Conflict Minerals Policy and Annual Report
• Annual Diversity Report
• Annual Sustainability Report
• Privacy Progress Update
• Progress on Civil Rights Audit Commitments
• Anti-Slavery and Human Trafficking Statement

Defining What Matters: Our Human Rights Policy
Building on our mission, we adopted our Corporate standards.1 The Policy guides teams to build rights-
Human Rights Policy on March 16, 2021, respecting products, respond to emerging crises,
committing ourselves to respecting human rights and work with speed and agility to embed human
as set out in the UNGPs and other human rights rights at scale. Also noteworthy:
We call the policy a policy — not a

“statement.” It is an enterprise-wide
commitment; part of our governance,
not just a reflection of a position.

• Our Code of Conduct requires personnel to
respect the Policy.
• Meta recognizes an expansive set of relevant
human rights standards and instruments — not
just freedom of expression — and other rights
contained in the International Covenant on Civil
and Political Rights (ICCPR).2 In particular, the
policy is framed to recognize the importance of
non-discrimination, equity, and equality — and
links to the work of the Meta Civil Rights team.
Section Two (“How We Implement”) describes

the Policy’s five main pillars and includes a
commitment to regularly reviewing and updating
the Policy.
01. Meta joins other technology companies in making human rights responsibilities
more explicit. For example, Apple published its human rights commitment in
August 2020, the same year that Google amended the charter of their Audit and
Compliance Committee to include oversight of civil and human rights issues.
02. E.g. the United Nations Guiding Principles on Business and Human Rights,
which encompasses internationally recognized human rights as defined by
the International Bill of Human Rights, including of the Universal Declaration
of Human Rights; the International Covenant on Civil and Political Rights; and
the International Covenant on Economic, Social and Cultural Rights — as well
as the International Labour Organization Declaration on Fundamental
Principles and Rights at Work.

Our commitments: In the Human Rights Policy, we recognize:
• The United Nations Guiding Principles on Business and Human Rights (UNGPs)
• The International Bill of Human Rights (the Universal Declaration of Human Rights;
the International Covenant on Civil and Political Rights; the International Covenant on Economic,
Social and Cultural Rights);
• The International Labour Organization Declaration on Fundamental Principles and Rights at Work;
• The International Convention on the Elimination of All Forms of Racial Discrimination;
• The Convention on the Elimination of All Forms of Discrimination Against Women;
• The Convention on the Rights of the Child;
• The Convention on the Rights of Persons with Disabilities;
• The Charter of Fundamental Rights of the European Union;
• The American Convention on Human Rights;
• The Global Network Initiative (GNI) Principles on Freedom of Expression and Privacy,
and their associated Implementation Guidelines;
• The OECD Principles on Artificial Intelligence; and
• The UN Declaration on Human Rights Defenders.
We recognize that the universal obligation of non-discrimination is a necessary — but not

sufficient — condition for real, lived, equality.
We strive to respect domestic laws. When faced with conflicts between such laws and our
human rights commitments, we seek to honor the principles of internationally recognized human
rights to the greatest extent possible. In these circumstances we seek to promote international
human rights standards by engaging with governments, and by collaborating with other
stakeholders and companies.

Meta is anchored in the value
“Live in the Future”
Bringing the Policy to Life: A Story of Culture and Progress
Meta is anchored in the value “Live in the Future,’’ Together, we create rights-respecting practices,
where we build the future of distributed work and decisions, approaches, and products.
where opportunity isn’t limited by geography; to
do this, we develop products using agile software To work at scale, we are guided by UNGP

development practices.3 prioritization criteria, and look for opportunities
to affect a broader system that touches multiple
Our human rights strategy is built from a well- products and countries. By way of example, the
known theory of change, the Knowledge, Meta-created Oversight Board’s role in content
Attitudes, and Practices Model (KAP). policy development and feedback advances
Researchers assume that knowledge, attitudes, human rights as well as other content moderation
and practice are related, and that knowledge and imperatives. Likewise, so does our workstream
attitudes directly influence practice.4 focused on mitigating risks of harm in countries
experiencing or at risk of violence.
We are a mission-driven company where
employees are typically aligned with human Using this theory of change, we are building
rights norms. In turn, this consensus leads to an impactful and durable human rights risk
a company-wide community that wants to management system; one that has greater strength
protect and advance human rights. Our Policy because it is closely aligned with our mission.
and integrated analytical approaches provide
guidance, expectations and guardrails for
employees to apply in their work. 03. According to the Agile Alliance, agile software development is “an umbrella
term for a set of frameworks and practices based on the values and
principles expressed in the Manifesto for Agile Software Development and
the 12 Principles behind it.” It refers to an environment where “Solutions
Human rights defenders, subject matter experts, evolve through collaboration between self-organizing cross-functional
external stakeholders, and our Human Rights teams utilizing the appropriate practices for their context.” For an overview
of agile and relevance across a broad range of industries, see Darrell Rigby,
Policy Team contribute to our knowledge. The Jeff Sutherland, and Hirotaka Takeuchi “Embracing Agile: How to Master
the Process that’s Transforming Management” Harvard Business Review,
Human Rights Policy Team shares that knowledge May 2016. See also, Nikola Jurisic, Michael Lurie, Philippine Risch, and Olli
by advising teams and decision-makers; Salo: Doing vs Being: Practical Lessons on Building an Agile Culture.
McKinsey and Company, August 4 2020
participating in external fora; developing role- 04. Originally pioneered in public health research, “K” stands for knowledge
of a problem or disease, “A” for attitude towards a problem or disease, and
specific and enterprise-wide training; and “P” for practice or preventive behavior to protect against the problem or
by performing useful and durable human rights disease. A relevant bibliography is cited in T.H. Wan et al., A Systematic Review
of KAP-O Framework for Diabetes, ESMED Medical Research Archives Vol 3
due diligence. No 9 (2016): April, Vol.3 Issue 9

C. Our Approach to Due Diligence
UNGP 17 states that companies should carry out during this reporting period. Our goal has been
human rights due diligence in order to “identify, to turn insights into actionable approaches that
prevent, mitigate and account for how they work for a company of Meta’s size, scale, and the
address their human rights impacts.” UNGP 24 culture of agile software development.
guides companies to prioritize their actions
to address actual and potential human rights In addition, we engaged in crisis support;
impacts according to severity. continuous due diligence processes, where we
focus on a key challenge for an ongoing period; as
We based our early approaches to human rights well as ongoing content policy feedback, and real
due diligence on our obligations as members of time decision-making support.
the GNI; specifically, we put into place strong
operational, policy, and legal measures designed We also sought to be an industry leader in
to scrutinize governmental takedown requests assessment and transparency. On an ongoing
and requests for personal data of users. basis, we shared insights and actions from our
human rights due diligence. Just prior to the
We expanded our approach with the November reporting period, Meta disclosed a detailed
2018 release of an independent human rights human rights report guiding the creation of the
impact assessment (“HRIA”) of Facebook’s impact Oversight Board. During the reporting period,
in Myanmar.5 Meta disclosed due diligence on Cambodia,

Indonesia, the Philippines, Sri Lanka, and —
Methodologies for human rights due diligence just after the reporting period — on its transition
are evolving. HRIAs are only one form of of Messenger and Instagram Direct Messages
diverse human rights due diligence offerings. to End-to-End Encryption.
Historically, many HRIAs stemmed from the
extractive or natural resource industries and
focused on a specific defined physical location,
or on supply chain issues. Today, initiatives
such as the UN B-Tech project seek to build on
these methodologies to better fit risks faced by
technology companies.
05. For more on HRIAs, see Part 02.II of this report. Meta’s reference in this Report
to third party diligence assessments cannot be construed as admission,
Global frameworks agree that the defining agreement with, or acceptance of any of the findings, conclusions, opinions or
element of a good faith HRIA is that it involves viewpoints identified in those assessments, or the methodology that was
employed to reach such findings, conclusions, opinions or viewpoints. Likewise,
direct consultation with affected rights-holders, while Meta references steps it has taken, or plans to take, which may correlate
to points assessors raised or recommendations they made, these also
as made clear in UNGP 21 which requires cannot be deemed an admission, agreement with, or acceptance of any
companies to have in place policies and processes findings, conclusions, opinions or viewpoints.
06. See also, UNGP 18 (“In order to gauge human rights risks, business enterprises
by which they can “know and show” their human should identify and assess any actual or potential adverse human rights
impacts with which they may be involved either through their own activities
rights responsibilities, including transparent and or as a result of their business relationships. This process should: (a) Draw on
accountable communication to people or groups internal and/or independent external human rights expertise; (b) Involve
meaningful consultation with potentially affected groups and other relevant
whose rights may be impacted by a company’s stakeholders, as appropriate to the size of the business enterprise and the
nature and context of the operation.”); OHCHR, The Corporate Responsibility
operations.6 Since 2018, Meta has rapidly to Respect Human Rights: An Interpretive Guide (2012), at 33 (“Human rights
experimented with ways to “show, not tell” our due diligence is about people. It reflects the entitlement of every human
being to be treated with dignity. It therefore involves relationships — between
commitment to human rights risk management an enterprise and those on whom it may have an impact. Hence, the key to
human rights due diligence is the need to understand the perspective of
through our due diligence program. Building on potentially affected individuals and groups. Where possible and appropriate
country-based HRIAs from 2018 and 2019, Meta to the enterprise’s size or human rights risk profile, this should involve direct
consultation with those who may be affected or their legitimate
tested a variety of due diligence approaches representatives, as discussed further under Guiding Principle 18.”).

II.
Meta’s Salient Risks
Every day, billions of people all over the world use

Meta’s social media platforms and messaging apps.
In this report, we aim to show the most salient

human rights issues identified by our due
diligence efforts to date, the actions taken across
product and policy teams to prevent or mitigate
these risks, and how we developed processes to
assess the effectiveness of those measures.

“Human rights risks are understood to
be the business enterprise’s potential
adverse human rights impacts.
Potential impacts should be addressed

through prevention or mitigation,
while actual impacts — those that have
already occurred — should be a subject
for remediation.”
C O M M EN TA RY TO U N G P 1 7
INSIGHTS AND AC TIONS 2020 - 2021 25 II. META’S SALIENT RISKS

A. How Meta Identifies its Salient Risks
As the Key Concepts section of the Interpretive Meta uses a thorough human rights assessment
Guide to the UNGPs clarifies: process to identify salient risks in a given or
particular context. Drawing on these assessments
The most salient human rights for a business — as well as our ongoing day-to-day human rights
enterprise are those that stand out as being due diligence — we established an initial list of
most at risk. This will typically vary according salient human rights issues.
to its sector and operating context. The Guiding
Principles make clear that an enterprise should We commit to conducting a company-wide salient
not focus exclusively on the most salient human risk analysis in the near future and reporting on it
rights issues and ignore others that might arise. in a forthcoming annual report.
But the most salient rights will logically be the
ones on which it concentrates its primary efforts.

B. Salient Risks
Based on due diligence efforts to date, the categories of salient risks for Meta can include, for example:
Rights Human Rights Standards
Universal Declaration Of Human Rights, Art. 19

International Covenant On Civil And Political Rights, Arts. 19, 20
Right to Freedom of
Convention On The Rights Of The Child, Arts. 13, 17
Opinion and Expression
International Convention On The Elimination Of All Forms Of
Racial Discrimination, Art. 4
Universal Declaration of Human Rights, Art. 12

Right to Privacy
International Covenant on Civil and Political Rights, Art. 17

Right to Life, Liberty and

International Covenant on Civil and Political Rights, Arts. 6, 9
Security of Person
Convention on the Rights of the Child, Art. 6
Universal Declaration of Human Rights, Arts. 1, 2, 7

International Covenant on Civil and Political Rights, Arts. 2, 3, 26
International Covenant on Economic, Social and Cultural
Rights to Equality and Rights, Arts. 2, 3
Non-Discrimination Convention on the Elimination of All Forms of Discrimination
Against Women, Art. 2
International Convention on the Elimination of All Forms of


International Covenant on Civil and Political Rights, Arts. 25
Rights to Public Participation, Convention on the Elimination of All Forms of Discrimination
to Vote, Be Elected Against Women, Art. 7, 8
International Convention on the Elimination of All Forms of

PA RT 2:
Meta’s Human
Rights Policy in
Practice
Our approach to human rights is grounded in the five pillars

of our Human Rights Policy:
I. Applying Relevant Policies;

II. Conducting Human Rights Due Diligence and Disclosure;
III. Providing Remedies for Human Rights Impacts;
IV. Protecting Human Rights Defenders; and
V. Governance, Oversight and Accountability
As we reflect on the Policy’s first year, we address our progress

and our room for growth across each area of commitment.
28
I.
Applying Relevant Policies
Under our Human Rights Policy, our respect for

human rights — and their underlying principles
of equality, safety, dignity, privacy and voice — is
applied through our existing policies, including
the Facebook Community Standards, Instagram
Community Guidelines, WhatsApp Terms of
Service, the Responsible Supply Chain program,
our Privacy Principles, our Law Enforcement
Guidelines, our Data Policy, our WhatsApp
Privacy Policy, our Code of Conduct, and our
Diversity and Inclusion practices, among other
policies so included.7
In the following section, we highlight key human-

rights related developments in the application of

our policies.
7. Moreover, as discussed infra, we extend our human rights commitments to

additional, internal product-focused decision-making approaches. These
efforts require coordination, dependencies and alignment with Products,
Operations, and Corporate partners.
INSIGHTS AND AC TIONS 2020 - 2021 29 I. APPLYING RELEVANT POLICIES

A. Content Policies
Freedom of expression is a fundamental right, one are — developed based on feedback from our
that is interdependent with other fundamental community and the advice of experts in fields
rights. We look to international human rights such as technology, public safety and human
experts when developing our standards for what rights. In seeking to ensure that everyone’s
content is and is not allowed on our social media voice is valued equally, we take care to create
platforms, and when deciding how to implement standards that include different views and beliefs,
these standards in practice. especially from people and communities that
might otherwise be overlooked or marginalized.
These rules are known as the Community For these reasons, when we limit expression, we
Standards for Facebook and Community do it in service of one or more of a set of values:
Guidelines for Instagram. They were — and Authenticity, Safety, Privacy, and Dignity.
Our Commitment to Voice:

The goal of our Community Standards is
to create a place for expression and give
people a voice. Meta wants people
to be able to talk openly about the issues
that matter to them, even if some
may disagree or find them objectionable.
T HE FAC E B O O K C O M M U N ITY S TA N DA R D S
The full set of Community Standards are Our work to infuse our Community Standards
available in 66 language options to-date, as are with human rights principles is ongoing and
our detection and enforcement policies. Details responsive to changing global salient risks,
on the effectiveness of our enforcement of consistent with Meta’s iterative approach to
these policies can be found in our quarterly policy development.
Community Standards Enforcement Reports.

01. Actions with Regard to Salient Risks
Right to Freedom of Opinion and Expression

Convention on the Rights of the Child, Arts. 13, 17
International Convention on the Elimination of All Forms of Racial Discrimination, Art. 4
Our Community Standards and Community Guidelines are highly informed by

international human rights standards (including the Rabat principles). Writing these
policies involves extensive stakeholder engagement, including with human rights
activists and academics. The implementation of these policies is informed by trusted
partners who provide important context, especially in high-risk environments.
For example, Meta drew on international human rights law in determining how to
address the complicated issue of bullying and harassment of public figures. For this
issue, the right to freedom of expression and legitimate public discourse around

people in the public eye must be carefully balanced against public figures’ rights to
security and non-discrimination, honor, and reputation.
To develop a consistent approach, Meta consulted with a diverse set of global

stakeholders including free speech advocates, human rights experts, women’s safety
groups, our Women’s Safety Expert Advisors, cartoonists and satirists, female
politicians and journalists, representatives of the LGBTQ+ community, content
creators and public figures.
Based on stakeholder feedback, and guided by international human rights principles,

the resultant Community Standards Bullying and Harassment policy maintains a
distinction between public figures and non-public figures, and requires removal of
attacks on public figures that encompass a wide range of harms to their fundamental
rights. During the reporting period, this policy was further updated to remove
certain severe sexualizing, degrading or derogatory content. The policy also provides
additional protections to public figures like journalists and human rights defenders
who have become famous involuntarily or because of their work.
In another, very important example: when a government sends Meta a take-down

request, we typically conduct three levels of review: against the Community Standards,
against applicable local law, and against international human rights standards and
our company principles. No matter which government makes a given request, we will
scrutinize it to make sure it is legally valid, proportionate and legitimate. In the interests
of full transparency, we publish reports that set out statistics on content we restricted
based on local law, including in response to government requests. We also publish case
studies of content we have restricted based on government takedown requests and
recently announced we may also share these requests with Lumen.

Right to Privacy

Privacy is one of the core values of the Community Standards. As we note: it “gives
people the freedom to be themselves, choose how and when to share on Facebook
and connect more easily.” In addition to prohibiting personally identifiable information
such as social security numbers, or other personal information directly identifying an
individual, the Community Standards also prohibit the posting of personal contact
information about oneself or others, including phone numbers or financial information
about businesses or organizations, except when publicly available. In 2021, Meta
requested a policy advisory opinion from the Oversight Board on the sharing of private
residential addresses and images. In 2022, the Board recommended that we disallow
this kind of sharing, and Meta acceded.
Right to Life, Liberty and Security of Person

We expanded our policies that prohibit veiled and implicit threats. We added a policy

to address hate speech attacks on concepts, ideas, practices, beliefs and institutions
related to protected characteristics when those attacks pose an imminent risk of
harm, intimidation or discrimination. We enhanced our bullying and harassment policy
to include stronger protections against gender-based harassment for everyone,
including public figures. We created new policies against certain mass harassment or
“brigading.” We also revisited our violence and incitement policies — specifically, the
allowance around the discussion of State use of force.
Rights to Equality and Non-Discrimination
Universal Declaration of Human Rights, Arts. 1, 2, 7

International Covenant on Civil and Political Rights, Arts. 2,3, 26
International Covenant on Economic, Social and Cultural Rights, Arts. 2, 3
Convention on the Elimination of All Forms of Discrimination Against Women, Art. 2
We make country-specific decisions informed by extensive stakeholder engagement

through our work with Trusted Partners and human rights activists and academics.
Using existing policies and relationships, we improved our abilities to enforce at scale.

In 2021, we provided an update on our work in Ethiopia identifying and removing a
number of persistent harmful false claims and out of context imagery that make false
allegations about the perpetrators, severity or targets of violence in Ethiopia. This
decision was based on guidance from over 50 local partners and independent experts.

The Philippines HRIA spurred improvements to our existing policy language

prohibiting multiple forms of human exploitation, and enforcement of our policy
that prohibits all sales of people, including children. As we described in Meta’s
Response, we updated policies around our proactive detection tools with the goal
to reduce discoverability of exploitative content and incorporate emerging trends
in consultation with the Philippines Department of Health and the National Kidney
and Transplant Institute, as well as the Inter-Agency Council Against Trafficking,
Overseas Workers Welfare Administration, Blas Ople Policy Center, Department of
Foreign Affairs and the Philippine Overseas Employment Administration. We have

ongoing coordination with the Philippines Inter-Agency Council Against Trafficking,
Department of Justice, Department of ICT and the Department of Education as well
as organizations advocating child safety, to quickly respond to reports and gather
local insights and trends, and help us improve our policies, tools, and resources to
better protect children online.
Rights to Public Participation, to Vote, Be Elected

International Covenant on Civil and Political Rights, Arts. 25
Convention on the Elimination of All Forms of Discrimination Against Women, Art. 7, 8
Our human rights-informed approach to election misinformation balances freedom

of expression against threats to the right to public participation. For example, in
preparation for the 2020 election in the United States, we took steps aimed at
preventing the spread of misinformation, (including clearer fact-checking labels),
fighting voter suppression and interference (including banning paid ads that suggest
voting is useless or advise people not to vote), and helping people better understand
the information they see online (including an initial investment of $2 million to support
a media literacy project). We also partnered with third-party fact-checkers
on WhatsApp.

02. Community Standards in Focus: Dangerous Organizations
One area that has drawn significant attention from as hate organizations because they engage in
human rights stakeholders has been our Dangerous coordinated violence against others based on
Individuals and Organizations Policy (“DOI” Policy), characteristics such as religion, race, ethnicity
which addresses networks of people who “proclaim or national origin and we routinely evaluate
a violent mission or are engaged in violence.” We groups and individuals to determine if they
do not want Meta to be a platform for hate. violate our policy.
• In addition, our policy covers Militarized Social
While terrorism and hate are global issues, there Movements, Violence-Inducing Conspiracy
is currently no globally recognized and accepted Networks, and individuals and groups banned
definition of terrorist or hate organizations. So for promoting hatred.
we developed the definitions in our DOI Policy to
assist our decision-making on enforcing against Our process for assessing organizations for
such organizations: possible designation includes structured review
of each case by a wide range of relevant teams,
• Our designations are divided into three tiers that including policy, legal, and security. As the
indicate the level of content enforcement, with DOI Policy demonstrates, our definitions and
Tier 1 resulting in the most extensive thresholds are agnostic to region or ideology.
enforcement, because we believe these entities Every designation goes through the process.
have the most direct ties to offline harm.
• We designate Dangerous Individuals and

14
Organizations under our policy after a rigorous
process that takes into account both online and
offline behavior. During this process, we work
to identify an organization’s goals and if it has a
track record of offline violence. Strategic network
• As for terrorist groups, we designate individuals disruptions to remove
and organizations based on their behavior
and whether they engage in, advocate, or lend
substantial support to purposive and planned
acts of violence with the intent to coerce,
23
intimidate and/or influence a civilian population, Different organizations
government, or international organization in
designated under our DOI policy
order to achieve a political, religious, or
ideological aim.
• In addition to the definition of terrorist groups,
and in recognition of the political complexity Finally, as part of our enforcement efforts
associated with the issue, we define as violent against DOIs, we undertake certain strategic
non-state actors those groups who primarily network disruptions, to help make sure that
direct violence toward state or military actors these groups cannot find ways back on our
(e.g. military targets). platforms and proliferate on our platforms. As
• As for hate groups: We ban groups that outlined in our Community Standards Reporting,
proclaim a hateful and violent mission from between October 2019 and August 2020, we
having a presence on Facebook and Instagram conducted 14 strategic network disruptions to
and we remove content that represents, remove 23 different organizations designated
praises or supports them. To date, we under our DOI Policy, including three U.S.-based
identified a range of groups across the globe white supremacist groups in October 2019.

B. Privacy
The right to privacy is a foundational right; an

essential requirement for the realization of
other rights, like the right to freedom of opinion
and expression. Particularly in this digital age,
interference with this right can both directly
and indirectly limit the free development and
exchange of ideas.
Privacy is one of the defining social issues of

our time and is central to Meta’s vision. We are
building a new privacy foundation to guide us
now and in the future. We are committed
to seeking feedback from and working with
stakeholders across industry, civil society, think
tanks and academia to improve our program.

Privacy Program Overview
Our Privacy program rests on a foundation of:
Governance: A cross-functional group of organizations across the company

provides engineering, legal, policy, compliance and product expertise that
enable the design and implementation of our privacy program. This includes an
independent committee of Meta’s Board of Directors that meets quarterly to
oversee our efforts to live up to our privacy commitments.
Privacy Education. To make privacy a core responsibility of every Meta employee,

we drive continuous privacy learning and education that spans training, internal
campaigns, regularly provided privacy content and other dynamic resources.
We do this through internal Workplace channels, lightning talks with privacy
leadership, internal Q&A sessions, a dedicated Privacy Week and an internal hub
of on-demand privacy content to help guide decisions and processes.

Accountability in Practice. In order to put our accountability foundation into
practice, we designed processes, escalation paths and technical mechanisms
that embed privacy across all facets of our company operations. This includes:
• A privacy risk assessment program that performs an annual assessment to
identify, assess, and address privacy risk across the company, as well as a
process to assess privacy risk after an incident occurs.
• Safeguards and controls, including operational activities, policies, and
technical systems.
• A privacy review process by which we assess privacy risks that involve
the collection, use, or sharing of people’s information and external
representations about our privacy and security practice; along with technical
requirements and tools to enhance accountability and operate the privacy
review process at scale, such as: a centralized tool that is used throughout
the project lifecycle for Privacy Review and a technical implementation
review that conducts review, verification, and documentation of the
technical implemental of privacy mitigations and commitments prior to
product launch.
• Incident Management, which includes: proactive identification of potential
privacy issues, a bug bounty program that has, over the past 10 years,

awarded bounties to around 1,500 researchers from 107 countries, and
transparency efforts, like our Incident Management program, that includes
steps to notify people where appropriate about issues impacting our
community, or working with law enforcement or other officials to address
what issues we find.
• Third Party Oversight, including a third-party privacy assessment process
for service providers to assess and mitigate privacy risk at Meta, as well
as a formal process for enforcing and offboarding third parties who violate
our privacy or security obligations, such as procedures and infrastructure
designed to ensure that third-party developers complete the Data Use
Checkup, and other technical and procedural mechanisms to monitor compliance.
• External Data Misuse. We also have technical mechanisms in place to mitigate
and prevent third parties from accessing data from Meta, through proactive
and reactive measures like prevention, deterrence, detection and enforcement.
Privacy Product Outcomes. The accountability processes, safeguards, and

technical mechanisms we built help ensure that new products and features
embed privacy by design. We see these updated processes enable us to
improve our privacy approach in new products and features, as we pivot to
respond to the world around us. This includes:

• Giving people greater access and control over their information
• Giving people access to their information
• Giving people control to manage their activity
• Offering an ephemeral messaging option
• Creating a secure, private messaging experience
• Providing age-appropriate experiences for youth
• Communicating data privacy practices to people
• Using privacy enhancing technologies for social good
Investments In A Technical Foundation That Supports Privacy, i.e., sustainable

technical solutions to meet evolving privacy expectations and ensure consistent
application of our privacy requirements across our products and systems, such as:
• A data deletion framework that helps alleviate the risk of potential error through
machine-learning automation; and
• Privacy-enhancing technologies based on advanced cryptographic and statistical
techniques that minimize the data we collect, process and share — to help

protect data at different stages of the data lifecycle.
Our privacy work is never finished, and we understand that our commitment means continuously
improving and focusing on this every day.

01. Salient Risks in Privacy
Two highlights of how Meta approaches its Meta’s approach to both challenges builds
salient privacy risks include: protecting users on our theory of change: developing systemic,
from unlawful or overbroad government demand scalable, integrated analytical approaches
for user data; and our responsibility to protect that provide guardrails informing product
user data and use it, share it, store it, and delete development enterprise-wide.
it pursuant to privacy and data protection
principles and applicable legal requirements.

The salient risk of government overreach
and abuse in demanding user
data from private technology companies
was the starting point of
Meta’s human rights journey.

I. Protecting Users from Unlawful or Overbroad Government Data Requests
The salient risk of government overreach and comply, we produce only the information that is
abuse in demanding user data from private narrowly tailored to that request. If we determine
technology companies was the starting point of that a government request is not consistent with
Meta’s human rights journey. The salience of this applicable law or our policies, or unlawful (for
risk has not diminished over time. In compliance example, overly broad, or legally deficient in any
with the GNI Principles, and our commitment way), we will push back.
to the UNGPs, we conducted human rights
due diligence to develop an approach for our We do not provide governments with “back doors”
responses to government demands for user data. to people’s information and we would challenge
any order that sought to have us redesign our
As set out in Facebook’s Data Policy, and systems to undermine the encryption we provide
guidelines for Law Enforcement, we scrutinize to protect people’s data, or any attempt to gag
every government request we receive, no us from disclosing the existence of such an order
matter which government makes the request. and our efforts to fight it. For example, Facebook
There are many legitimate and lawful reasons and WhatsApp paused — on human rights
why governments request information from grounds — all data disclosure to the Hong Kong
social media companies: international criminal government after the passage of the National
investigations, anti-terrorism work, prevention of Security Law in 2020.
child sexual abuse and other efforts intended to
protect the lives, security and rights of people in We “show, not tell,” by regularly publishing
their countries. We respond to such requests for transparency reports that set out, by country, the
information that are consistent with internationally number of government requests we have received
recognized standards on human rights, including and the proportion of requests for which some
due process, and the rule of law. When we do data was produced.

Reform Government Surveillance Coalition
Along with other major tech companies, Meta is a member of the Reform Government
Surveillance Coalition, which urges the world’s governments to adopt surveillance laws and
practices consistent with established norms of privacy and free expression and the rule of law.
The six core principles of the coalition are:
1. Limiting governments’ authority to collect users’ information;
2. Oversight and accountability;
3. Transparency about government demands;

4. Respecting the free flow of information;
5. Avoiding conflicts among governments; and
6. Ensuring security and privacy through strong encryption.

II. Properly Collecting, Using, Storing, and Deleting User Data
We made progress on our work to give people how their information is used and how to secure
more control over their data, and our broader their account.
mission to honor people’s privacy in everything • See “Why Am I Seeing This Ad?” This tool lets
we do by building processes, products, and people tap on posts from the friends, Pages,
technical mechanisms that laid the foundation for and Groups they follow as well as some of the
privacy and accountability across the company. posts Meta suggests to get more context on
The program includes internal education, due why they are appearing in News Feed.
diligence, safeguards and controls, and third- • Review Off-Facebook Activity. We created a
party oversight. For example, we provide people tool so people can control — or disconnect —
with ways to: the information businesses send to Meta about
their activity on other apps and websites.
• Manage their information. We make it easy for • Review Ads Interests & Preferences. This tool
people to access, manage, download or delete gives people more visibility into the actions
the personal data they have provided. they’ve taken to control the topics of the ads
• Transfer their data. We make it simple to move you see.
information, such as posts, photos and videos, • Access Cookie Controls. In the European
to another service. Region, we give people a more granular level
• Access the Accounts Center. People can control of control over their cookie choices and more
the experiences they connect across our apps information on how we use different kinds
by using our Accounts Center. of cookies.
• Access the Instagram Activity Center. People • Set up Two-Factor Authentication. We give
can access and manage content they’ve people this option in addition to their password to
commented on, shared or sent on Instagram. help protect their account from improper access.
• Perform a Privacy Checkup. This tool helps
people control who can see what they share,

C. Civil Rights
Civil rights are human rights, but they are also understood to be personal rights protected by
the Constitution of the United States and Federal and State laws. The Meta Civil Rights team
focuses on issues of discrimination and exclusion based upon race, ethnicity, religion, national
origin, gender, gender identity, sexual orientation and disability.
Meta’s commitment to human rights and civil This audit laid the groundwork for the Civil Rights
rights are closely interrelated. Our Human Rights Team that was created in November 2020 and
Policy lays a strong basis for continuing civil rights provided the company with an initial roadmap.
and racial justice work and standards.
In November of 2021, Meta published a
The initial framework for our Civil Rights work report that detailed Meta’s Progress on Civil
was the civil rights audit we commissioned at the Rights Audit Commitments, demonstrating
behest of the civil rights community to identify success in implementing 65 of the Auditors’

instances where the company could take a more recommendations and actions, as well as
comprehensive approach to respecting civil demonstrating ongoing or in-progress status of
rights. Meta published the findings in full: a total 42, and continuing evaluation of the feasibility of
of three reports, the last of which was published eight. This Progress Report explained that Meta’s
in July 2020. The extensive audit resulted in 117 civil rights work focuses on five key pillars:
recommendations and actions with input from over
100 civil rights and social justice organizations.
Law Enforcement & Hate: Addressing harm and creating accountability are critical to
enhancing protections for marginalized communities and victims of hate speech, hate
incidents and hate crimes are safe on our platforms. It also means working to prevent
law enforcement misconduct and surveillance of marginalized communities on our
platforms as these activities can violate our policies and chill free expression rights.
Voting & Civic Engagement: Grounding the work in efforts to inform and connect
people, communities, and civic institutions to come together and drive real
outcomes to improve lives, because when all people have an opportunity to
participate and are considered political equals, societies make decisions optimized
for the many, not the few.

Technology: It is Meta’s responsibility to help instill civil rights best practices
throughout the company with regard to artificial intelligence, machine learning,
virtual and augmented reality, and emerging technologies, and lead engagement
with external technology experts.
Product: Meta aims to advance our priority of doing good and causing no harm by
creating product review approaches and guidance tools aimed at prioritizing protected
classes and protecting systemically and historically marginalized communities.
Policy & Enforcement: Meta works to surface civil rights concerns on the front end
of policy development and enforcement decisions and ensure a more diverse and
inclusive engagement process.

Growing out of the civil rights audit, the Meta
Civil Rights Team has grown to include people
with expertise spanning hate crimes, voting
rights, counterterrorism, immigration, national
security, law enforcement, product inclusion,
algorithmic fairness and transparency.
As this team continues to advance civil rights

and liberties, it does so in partnership with the
Human Rights Policy Team. Meta is committed
to continuing to build with the values of justice,
equity, dignity, and safety in mind.

D. Responsible Innovation and Responsible AI
To help ensure that we build the future of the COVID-19 pandemic. Meta also developed a
internet responsibly, Meta developed human dedicated, cross-disciplinary Responsible AI (RAI)
rights-based integrated analytical approaches to team within its AI organization to help ensure
support our innovation and product development. that AI governance is based on foundational
values of respect for human rights, democracy,
Our Responsible Innovation Dimensions support and the rule of law.
product development. This framework is evolving
over time, but currently includes 10 dimensions: Those foundational values are at the root of
autonomy, civic engagement, constructive the wide range of principles statements that
discourse, economic security, environmental have been released around responsible AI
sustainability, fairness and inclusion, privacy and development, most especially the European
data protection, safety, voice, and well-being. These Commission’s High-Level Expert Group’s
dimensions in turn guide our analysis and practice. Ethics Guidelines for Trustworthy AI and the
Organization for Economic Cooperation and
Artificial Intelligence (AI) is a core component Development’s Principles on Artificial Intelligence,
of the technologies Meta uses to provide value which Meta helped develop.
to people and keep our communities safe, from
helping rank posts in News Feed to tackling Meta, in turn, organized its Responsible AI efforts
hate speech and misinformation related to the around five key pillars:

Privacy & Security
Fairness & Inclusion
Robustness & Safety
Transparency & Control
Accountability & Governance

We also invest in research and open-sourcing disseminate emerging best practices that are in
datasets and tools to help facilitate responsible use line with its AI Principles.
of AI, such as privacy-preserving machine learning,
AI explainability, and fairness. For example, in Through our partnership with Open Loop, we are
2021 we released our Casual Conversations data building innovative “policy prototyping” projects
set, composed of over 45,000 videos designed to for testing new potential AI policy requirements
similarly help researchers evaluate computer vision with regulators and startups before they become
and audio models for accuracy across a diverse set law, to ensure that they are both practical and
of ages, genders, apparent skin tones, and ambient impactful. We launched projects in Europe and the
lighting conditions. Asia Pacific region.
We are also improving transparency by piloting Finally, we are funding a global effort to solicit
simple, standardized documentation of our diverse academic research on AI ethics and
models and using interpretability software such governance topics, supporting the publication
as Captum. Although work in this area is still in its of academic papers in Asia, Africa, and Latin
infancy, our hope is that ultimately we will be able America and providing foundational support for
to build an integrated transparency solution that an independent Institute for Ethics in Artificial
can automatically feed information from internal Intelligence at the Technical University of Munich.
documentation efforts — like model cards8 — into
new transparency features and controls for the
people using our products.
In addition to our technical research and product- 08. Model Cards are short documents accompanying trained machine learning
models that provide benchmarked evaluation in a variety of conditions, such as
focused work, we are actively participating in across different cultural, demographic, or phenotypic groups ( e.g., race,
geographic location, sex, Fitzpatrick skin type) and intersectional groups ( e.g.,
efforts to establish clear AI principles and best age and race, or sex and Fitzpatrick skin type) that are relevant to the intended
practices, including collaborating with the OECD application domains. Model cards also disclose the context in which models are
intended to be used, details of the performance evaluation procedures, and
on the AI Observatory project to study and other relevant information.

E. Accessibility
The UN Convention on the Rights of People with • We introduced new accessibility-specific features
Disabilities sets out eight fundamental principles in Quest 2, including color filters to assist people
that underpin the protection of the rights of people with color-sensitivity and “adjust height” to
with disabilities. These include non-discrimination, enable better VR experiences for people who are

full and effective participation and inclusion in reclining, seated, or lying down.
society and accessibility. Given the importance of • Automatically generated captioning available
the digital world for people with disabilities to be for Facebook Live, Workplace Live, Instagram
able to participate fully in exercising their rights, TV, and Live Audio Rooms, as well as for
Meta sought to apply cutting-edge technologies Facebook ads, Pages, and Groups.
and make other investments to create more
accessible features and services for users with a In addition, we engaged in several initiatives
wide range of disabilities. Such measures include: to help promote accessibility beyond Meta.
We work collaboratively with leaders in the
• Applying the latest AI techniques to increase accessibility community globally, and are working
object and concept recognition in Automatic within our industry and through the W3C to
Alt Text or “AAT,” our award-winning automatic establish an interoperability standard for web and
photo-description technology that works with assistive technology.
screen readers to describe photos to people who
are blind or have low vision, by more than 10X. We also co-founded and remain active in
• Re-engineering facebook.com to integrate Teach Access, a collaboration of industry,
greater accessibility into the website, including academia, and disability advocates to advance the
technology that helps people using screen teaching of accessibility in education. Students
readers better understand what’s on a page and learn state-of-the-art best practices in accessible
find what interests them more quickly. software design, preparing them to make future
• WhatsApp has various accessibility features, technologies that are “born accessible.” We
including a voice-assisted camera feature, which are investing early in making the metaverse
makes it easier for people who are blind or accessible, including by creating XR Accessibility
have low vision to take photos on iOS devices, guidelines for developers in partnership with
and a new WhatsApp email accessibility support the XR Association and additional guidelines
channel for users to share suggestions, ask specifically for third-party developers of the
questions, or report problems. Quest VR platform.

F. Anti-Slavery and Human Trafficking
Meta is opposed to all forms of human trafficking, in accordance with the Responsible Business
slavery, servitude, forced or compulsory labor, Alliance (RBA) Code of Conduct, which includes
and all other trafficking-related activities. We standards related to indicators of modern slavery
are committed to respecting all applicable and human trafficking.
international human rights standards, labor and
employment laws, rules, and regulations, and to We verify supplier conformance with RSC policies
working to mitigate the risks of modern slavery and standards through continuous dialogue,
and human trafficking in our business operations independent audits and assessments, corrective
and supply chains. We manage labor, human action plans, worker surveys, and other forms of
rights, and environmental risks in our supply chain assurance. Any non-conformances identified are
through our Responsible Supply Chain (RSC) actively addressed through our corrective action
program. We collaborate and share ideas with the and key performance indicator programs. We
broader industry and global community to update regularly assess and track the effectiveness of our
and develop solutions that promote best practices actions taken in mitigating the associated risks of
for a responsible supply chain, and to that end we modern slavery and human trafficking. We strive
are members of the Responsible Business Alliance to improve year-on-year; every year we publish a
and the Responsible Labor Initiative. refreshed/new statement to demonstrate progress
and strengthening of our program efforts.
We require all personnel be trained on and comply
with our Code of Conduct, which affirms our As noted, Meta strongly opposes the abuse of

commitments to human rights and prohibits our family of apps and platforms to facilitate any
violations of law, including labor and employment form of human exploitation. At the core of our
laws. Meta requires that priority suppliers, which strategy, we continually update and refine our
are identified using a risk-based methodology, policies to ensure we are keeping up with salient
conform with all of the standards on labor, health human rights risks related to how abuse tactics
and safety, the environment, business ethics, change over time.
and the establishment of management systems
Strategic partners include:

National Center for Missing and Exploited Children, International Center for Missing
and Exploited Children, Internet Watch Foundation, ECPAT, Polaris, Stop the Traffik,
and Meta’s Expert Advisory Group
We implemented new in-product features to We collaborate with global NGOs to create

raise awareness of trafficking, deter violating educational campaigns, provide resources and
behavior, and offer support to victims. assist victims and survivors of human trafficking.
One example is our search redirection that We organize internal training, and events to raise
detects when people search for potentially awareness of human exploitation amongst
violating keywords and delivers a custom our employees.
deterrence message. We are increasing our
investments and our ability to identify the illicit We continually seek to expand and refine our
actors, networks, organizations, and businesses understanding of emerging salient risks related
that perpetrate these activities and disrupt to human exploitation as we amend and update
them accordingly. our enforcement guidelines.

G. Election Integrity
We are committed to protecting elections,

to increasing authenticity, transparency and
accountability for advertisers, to giving everyone
a voice on our technologies and to empowering
people to vote. Between 2020 and 2021, Meta
made significant investments in teams and
technologies to better protect elections, empower
people to vote and create safe technologies to
share diverse beliefs.

01. United States Presidential Election 2020
For the United States Presidential Election in 2020, each of these takedowns publicly, and shared
we started preparations two years in advance. information with independent researchers so that
These enabled us to identify emerging threats and they could review and draw independent findings.
put systems in place to mitigate anticipated risks. • We removed more than 265,000 pieces of
content on Facebook and Instagram in the US
Among the threats we expected to see in between March 1 and Election Day for violating
the lead-up to the November election were: our voter interference policies.
perception hacking, the shift from larger-scale • We attached informational labels to content
operations to narrower campaigns that try to that discussed the legitimacy of the election or
slip under the radar and leverage unwitting claimed that lawful methods of voting like mail-
authentic people, and also blurring lines between in ballots would lead to fraud.
authentic public discourse and manipulation by • We ran the largest voting information campaign
co-opting domestic groups to amplify and join in American history, connecting people with
influence operations. Actions taken included: reliable information about voting from state and
local election authorities as well as nonpartisan
• We removed almost a dozen foreign operations civic partners.
backed by Russia, China and Iran that used fake • We worked with law enforcement in the days and
accounts to deceive users and undermine trust weeks after January 6, 2021 and continue to do so
in the United States under our policy against with the goal of ensuring that information linking
coordinated inauthentic behavior, announced the people responsible for their crimes is available.

02. Myanmar Election 2020
Having identified Myanmar as high priority • Implemented proactive detection to surface

and high risk in advance of its November 2020 content containing certain keywords/phrases
election, Meta created a multi-disciplinary associated with hate, misinformation, and
planning team with a human rights presence other harms for further review.
to help identify, prevent or mitigate key salient • Introduced a third-party fact-checking
human rights risks. Meta also took steps program in Myanmar as part of our ongoing
to prepare for the election, including: integrity efforts to reduce the spread of
misinformation and improve the quality of the
• Stood up a dedicated Integrity Product news people find online.
Operations Center with 24/7 coverage during • Directed people to authoritative sources of
the election to respond to risks and on/offline election information, where they could learn
developments in real-time. how to check voter lists, as well as voting
• Expanded our misinformation policy in times and locations, including Union Election
Myanmar in an effort to help combat voter Commission, Vote MM and First Time Youth
suppression, and help support the integrity Voters for 2020.
of the electoral process, including working • Trained civil society organizations and
with local partners to remove verifiable reporters on journalist safety, media and digital
misinformation and unverifiable rumors ( e.g., literacy, as well as Meta’s policies and third-
posts falsely claiming a candidate is a Bengali, party fact-checking programs. As part of this

not a Myanmar citizen, and thus ineligible). effort, we held a monthly television talk-show on
• Required all electoral and political ads in digital literacy called Tea Talks, that focused on
Myanmar have a “Paid for by” disclaimer issues like online bullying and account security.
attached to them to show the organization or • Introduced tools to newsrooms in Myanmar
person behind the ad. such as CrowdTangle, a public insights tool
• Worked with partners in Myanmar to verify that makes it easy to follow, analyze and report
official Facebook Pages of political parties on what’s happening with public content on
and candidates. social media.
• Worked with civil society partners to surface • Identified and disrupted six networks engaging
and rapidly review potentially harmful content, in Coordinated Inauthentic Behavior in
including misinformation that could lead to Myanmar. These networks of accounts, Pages
offline harm, veiled threats, and hate speech. and Groups were masking their identities to
• Introduced an Image Context reshare product mislead people about who they were and what
which provided people using the platform in they were doing by manipulating public
Myanmar with additional context before they discourse and misleading people about the
shared images that were more than a year old origins of content.
and could be potentially harmful or misleading.
• Introduced a new feature that limits the
number of times a message can be forwarded
to five. This safety feature was made available
on Messenger in Myanmar in 2020.
• Utilized hate speech classifiers to proactively
detect probable hate speech in Burmese
and either demote, route for human review,
or automatically delete depending on
confidence level.

03. Preparation for Philippines Presidential Election 2022
In line with our approach to other major program, Digital Tayo to reach over 6.5 million
global elections, Meta made it a high priority people in the Philippines. Digital Tayo covers
to invest in risk mitigation and preparation topics such as online safety, privacy, digital
efforts for the 2022 presidential election in the citizenship, news and media literacy, and
Philippines. These efforts were informed and launching civic campaigns.
guided by recommendations of the Philippines • We supported Internews Philippines to launch
Human Rights Impact Assessment, and were a the Philippine Fact Checker Incubator, to
demonstration of our commitment to follow up on increase the capacity of local organizations for
that work. We built new products and developed fact checking.
stronger policies in collaboration with the • To help people better understand who is
Commission on Elections, election watchdogs, behind the news they see on Facebook and
independent fact checkers and civil society Instagram, we applied labels to media outlets
organizations. In addition: that are determined to be wholly or partially
under the editorial control of their government,
• In partnership with the Philippine Commission including those in the Philippines.
on Elections and various civic organizations • As part of tackling other emerging harms, we
including the Legal Network for Truthful also removed a network of over 400 accounts,
Elections (LENTE), we launched a civic education Pages, and Groups in the Philippines that
campaign, conducted training on Meta’s policies, worked together to systematically violate our
and held roundtables with civil society. Community Standards and evade enforcement.

• We’ve expanded our flagship digital literacy

H. COVID-19
The world has suffered immense health and

human rights consequences since COVID-19
was declared by the WHO to be a Public Health
Emergency of International Concern (PHEIC) in
January of 2020.
Throughout the reporting period, Meta mobilized

to support public health, amplify authoritative
information, connect users to essential services,
as well as assist relevant agencies in their life-
saving work. Three areas of Meta’s COVID-19
response are particularly relevant to this report:
(1) emergency due diligence, (2) content policy
due diligence, and (3) additional actions to
support the right to health.9
In 2021, we also produced a detailed white paper

summarizing actions Meta took to support effective
COVID-19 response while protecting privacy.
09. Further details of Meta’s COVID response are available at our Responding to
COVID 19 site, and on the newsroom

01. Emergency Due Diligence
As the pandemic broke out, the Human Rights expression; as well as rapid review of
Policy Team worked on an emergency footing emergency product measures, adjustments to
with other internal teams to align emergency automated systems, adjustments to the
response actions with UNGP guidance and at-risk countries workstream, and frameworks
prioritization frameworks. Their actions included: for prioritizing content for human review by a
real time decision support and advice related to temporarily reduced pool of content moderators.
the rights to health, information, and freedom of
02. Content Policy Due Diligence
As the pandemic spread, Meta’s content policy Based on input from experts in health
teams worked to respect and incorporate communication and related fields, we took
authoritative guidance related to the right to additional steps to reduce the distribution of
health, the right to information, and permissible content that does not violate Meta policies
limits on freedom of expression. but may present misleading or sensationalized
information about vaccines in a way that would
Informed by this and other work, Meta be likely to discourage vaccinations.
took aggressive steps to limit the spread of

misinformation about the virus and its effective In an important step for enforcement at scale,
vaccines and treatments, as well as to connect Meta developed and deployed AI tools to scale
people with reliable health and service information. fact-checker’s work and detect copies of rated
We also updated other relevant content policies to content when someone tries to share them.
mitigate COVID-19-related risks, including (among We also built new computer vision classifiers
others) our hate speech policies; coordinating to enforce bans on violating ads and commerce
harm policies; restricted goods policies; and listings for certain medical products.
bullying and harassment policies.
The full set of our relevant policies are available

in: the Misinformation section of the Facebook of the time
Community Standards, the COVID-19 and
people who
Vaccine Policy Updates & Protections section
saw the
of the Facebook Help Center, by the Instagram
warning label
Community Guidelines, and the Instagram-
specific page relevant to COVID-19 and Vaccine 95% did not click
Policy Updates and Protections. past to view
Under these policies, we remove content

containing claims related to COVID-19 and
vaccines or treatments that, according to public These efforts were effective. In April 2020,
health authorities, are (a) false, and (b) likely to for example, we put warning labels on about
contribute to imminent physical harm. Imminent 50 million pieces of content based on around
physical harm examples include: increasing the 7,500 fact-checks from partners. During the
likelihood of exposure to or transmission of the month of March 2020, we displayed warnings on
virus, or having adverse effects on the public about 40 million posts related to COVID-19 on
health system’s ability to cope with the pandemic. Facebook, based on around 4,000 articles by our

independent fact-checking partners. 95% of the resources from health authorities through our
time people who saw the label did not click past COVID-19 Information Center and pop-ups on
to view. From the beginning of the pandemic Facebook and Instagram. We provided ad credits
to July 2021, we put these warning labels on to governments and health agencies while
167 million pieces of content. WhatsApp enabled national health ministries
around the world to communicate at scale
We also made significant efforts to serve public with their citizens by onboarding them to the
health organizations and emergency responder WhatsApp Business API, allowing them to provide
organizations, globally and locally. We connected timely information and respond automatically to
over 2 billion people from 189 countries to frequently asked questions.
03. Privacy Due Diligence and Data for Good
Our Data for Good Program has been sharing large scale COVID-19 Trends and Impact Survey,
privacy-protective tools and data with its 600+ to which over 100 million people in 130 countries
partners in more than 75 countries, many of and territories responded. We also hosted
whom are working to track the spread of the public visualizations of this publicly available
virus; measure the effectiveness of prevention information to help communicate to a wide range
measures; and better understand attitudes of audiences.
towards vaccines.

While Meta was developing its strengths-
Meta sought to address the urgent needs of public based approach, the human rights and privacy
health authorities while maintaining a strong teams undertook detailed human rights due
commitment to privacy and other fundamental diligence to guide Meta positions on responding
rights. After committing to be guided by public to governmental COVID-relevant data sharing
health experts in our response, Meta decided to requests, the development and promotion of
prioritize areas where our existing strengths made contact-tracing interventions, and the collection
our contributions particularly impactful. of sensitive data via chatbots in encrypted
messaging services. These complex frameworks
Meta understood from epidemiologists that confirmed the importance of the strength-based
surveys, in particular, could assist their efforts approach. Meta did not support the promotion
in three ways: (1) forecasting the spread of the of automated contact tracing apps and did
virus; (2) assessing peoples’ understanding of not seek to develop such technology, focusing
preventing behaviors, such as mask wearing; and its efforts instead on sharing population-level
(3) understanding attitudes and behaviors related insights on changes in mobility and how this
to vaccinations as vaccines became available. might affect transmission and the effectiveness
of stay at home orders around the world.
As a result, we launched a number of mobility
datasets that were used across dozens of More detailed information on these initiatives,
countries to better understand the effectiveness along with key discussion questions, is available
of lockdown orders and partnered with select in our April 2021 white paper: A Retrospective —
academic and public health partners to create a Protecting Privacy in Our Covid-19 Response.

II.
Conducting Human Rights

Due Diligence and Disclosure
Due diligence is the second core pillar of Meta’s

Human Rights Policy. Pursuant to the UNGPs, we
endeavor to (a) assess actual and potential human
rights impacts; (b) take appropriate action and
implement findings; (c) monitor implementation;
and (d) communicate how impacts are being
addressed. We seek to identify vulnerable or
marginalized groups and engage, meaningfully, to
listen to concerns, and create solutions. As part
of our “show, not tell” ethos, we have sought to
demonstrate meaningful transparency in our due
diligence exercises.
In this section, we detail examples of the

innovative approaches to diligence that we have
operationalized, as described earlier:
1. Human rights impact assessments of

specific countries, product initiatives, or
strategic initiatives;
2. Integrated analytical approaches embedding
UNGP prioritization criteria; and
3. Rapid salience analyses to support crisis
situations, real-time content policy feedback,
and product decision-making.
As the UNGPs were only approved by the UN

Human Rights Council in 2011, the field of human
rights due diligence is nascent, with rapidly
evolving methodologies and best practices.
For social media companies like Meta, whose
products, scale of operations and impact are
also evolving at a dramatic pace, the challenge of
iterating methodologies capable of identifying
salient risks and developing meaningful
recommendations is considerable.
INSIGHTS AND AC TIONS 2020 - 2021 54 II. CONDUC TING HUMAN RIGHTS DUE DILIGENCE AND DISCLOSURE
A. Human Rights Impact Assessments
A human rights impact assessment (HRIA) is Rights-holders, in this context, included not only
a detailed, direct form of human rights due users of our platforms and services, but the many
diligence that allows companies like Meta to others whose rights were potentially impacted by
identify potential human rights risks and impacts, online activity and conduct.
to promote human rights and seek to prevent and
mitigate risks. This work is in line with both our Countries targeted for human rights
commitments as members of the GNI, and our due diligence exercises have stemmed
responsibility under UNGP 18.10 from Oversight Board recommendations,
recommendations from prior diligence exercises,
In 2018, Meta commissioned and published the and other ad-hoc determinations based on
full text of an independent HRIA for Myanmar: stakeholder feedback. We are moving to a more
An Independent Assessment of the Human Rights systematic approach informed by decision
Impact of Facebook in Myanmar. frameworks and stakeholder engagement.
In May 2020, we published executive summaries

of and recommendations from three more HRIAs
for: Indonesia, Sri Lanka and Cambodia, along with
our responses (Indonesia, Sri Lanka, Cambodia),
at which time we also directly briefed interested 10. UNGP 18 requires businesses, “in order to gauge human rights risks” to
“identify and assess any actual or potential adverse human rights impacts
and affected rights-holders, with the exception with which they may be involved either through their own activities or as
of Cambodia where COVID-19 and the security a result of their business relationships” while drawing on “internal or external
expertise and “meaningful consultation with potentially affected groups and
environment prevented us from doing so. other stakeholders.”
01. Philippines HRIA

Salient Risks:
The Philippines HRIA was conducted by an economic opportunities, giving voice to people
independent third party from February to July and being essential tools for monitoring and
2020 and published in December 2021. The defending human rights during the COVID-19
assessment found that Meta technologies are pandemic. It also highlighted salient human
widely used and have positive human rights rights risks, including concerns about the misuse
impacts in the Philippines, as well as salient of our technologies for misinformation and
human rights risks. For example, the assessment disinformation, online harassment, incitement of
found that Meta technologies play an important violence, surveillance, online sexual exploitation,
and positive role in providing access to human organs trafficking, and extremist activities.
Recommendations and Meta’s Response:
The HRIA made 40 recommendations. At the we made substantial commitments to prepare

end of 2021, Meta committed to implement for the 2022 Philippine presidential election and
24 recommendations, partly implement seven strengthened our partnerships and programs
recommendations, and was assessing the with Philippine civil society organizations and
feasibility of another nine. We will be tracking government entities. Meta’s full commitments are
our implementation regularly through 2022 and available at Meta Response: Philippines Human
beyond. In response to the HRIA recommendations, Rights Impact Assessment.
02. Summary Disclosure - India Human Rights Impact Assessment11

Purpose
In late 2019, Meta commissioned an independent human rights impact assessment

(HRIA) on potential human rights risks in India related to its platforms. The project
was undertaken by Foley Hoag LLP12. The project launched in March 2020. It
experienced limitations caused by COVID-19, with a research and content end date
of June 30, 2021.
The assessment was conducted independently of Meta. Meta provides this

summary of the report’s insights and actions consistent with commitments made
in its Corporate Human Rights Policy, and the United Nations Guiding Principles on
Business and Human Rights, including Principle 21(c).
This summary cannot be construed to endorse, accept or adopt the independent

assessor’s conclusions, recommendations or analysis.13
11. This summary was developed after consulting The Danish Institute for Human employed to reach such findings, conclusions, opinions or viewpoints. Likewise,
Rights Guidance on HRIA of Digital Activities Reporting and Evaluation while Meta in its response references steps it has taken, or plans to take, which
12. Foley Hoag LLP is a US law firm with a human rights practice. may correlate to points Foley Hoag raised or recommendations it made, these
13. Meta’s publication of this summary, and its response thereto, cannot be construed also cannot be deemed an admission, agreement with, or acceptance of any
as admission, agreement with, or acceptance of any of the findings, conclusions, findings, conclusions, opinions or viewpoints.
opinions or viewpoints identified by Foley Hoag, or the methodology that was
Methodology
The HRIA involved interviews with 40 civil society stakeholders, academics, and
journalists. Stakeholder identities were kept anonymous and not shared with
Facebook/Meta. The due diligence also involved the review of content policies, certain
relevant content, as well as a survey, designed across multiple dimensions of diversity,
of over 2000 individual rights holders.
The report refers extensively to international standards. These include the United
Nations Guiding Principles on Business and Human Rights; the U.N. International
Bill of Human Rights (including the Universal Declaration of Human Rights
(“UDHR”), the International Covenant on Civil & Political Rights (“ICCPR”), and the
International Covenant on Economic, Social and Cultural Rights (“ICESCR”); the U.N.
Rabat Plan of Action, including freedom of expression, information, and opinion; the
right to non-discrimination; and standards relating to advocacy of national, racial, or
religious hatred.
Limitations

The project experienced certain limitations caused by COVID-19. HRIAs conducted
on the operations and activities of digital products exist in an emerging and
constantly adapting space.
Based on the guidance of human rights experts, we have produced this synthesis in
order to mitigate security risks as per UNGP 21(c).
While commissioned by Meta, the independent Assessment is not adopted by Meta

or a statement on its behalf. Instead, the Assessment will ideally serve as a baseline
to help identify and allow Meta to address the most salient platform-related human
rights risks, and help guide additional human rights due diligence.
This document is intended to be a summary of the Assessment.
Context
In 2019, civil society groups published several reports criticizing Facebook content
policy rules and content moderation processes in India. Using the guidance of the
UNGPs, Meta initiated a human rights due diligence project to identify and mitigate
potential human rights risks.
Insights
The HRIA was an independent report by a third party.
It found Meta platforms had provided an invaluable space for civil society to
organize and gain momentum, provided users with essential information and facts
on voting, and also enabled important public health updates. It also noted that
Meta had committed to platform safety and integrity in India, and had devoted
considerable energy in recent years to promoting respect for its users, and to
protecting user safety.
These efforts included: implementing an industry-leading set of content rules,

developed in close consultation with the human rights community; substantial
investments in new resources to detect and mitigate hateful and discriminatory
speech; new content moderation resources, and recruitment of expert global human
rights personnel.
Risks

The HRIA noted the potential for Meta’s platforms to be connected to salient human
rights risks caused by third parties, including: restrictions of freedom of expression
and information; third party advocacy of hatred that incites hostility, discrimination,
or violence; rights to non-discrimination; as well as violations of rights to privacy and
security of person. It stated Meta faced criticism and potential reputational risks
related to risks of hateful or discriminatory speech by end users.
The Assessment also noted a difference between company and external stakeholder
understandings of content policies. It noted persistent challenges relating to user
education; difficulties of reporting and reviewing content; and challenges of enforcing
content policies across different languages. In addition, the assessors noted that civil
society stakeholders raised several allegations of bias in content moderation. The
assessors did not assess or reach conclusions about whether such bias existed.
Actions
The HRIA developed recommendations covering implementation and oversight;

content moderation; and product interventions; and other areas.
As previously stated, the conclusions and recommendations made in the

Assessment are not statements made on behalf of or adopted by Meta. Meta is
studying these findings, however, and will consider them as a baseline to identify
and guide related actions. We continue to work towards providing a safe and open
platform to all users.
On implementation and oversight, Meta has taken steps to expand its human rights
team; aims to ensure there is appropriate regional expertise; and will seek to build
awareness of its diverse slate recruitment approach globally. We’ll deepen our
stakeholder engagement in a comprehensive manner.
On content moderation, Meta already actively seeks to align our content moderation
and enforcement policies with international human rights law.
In 2021—22, we made content policy changes related to a) groups at risk of outing

under our coordinating harm policies and b) context-specific content attacking
concepts and practices associated with protected characteristics. We will also
develop a prototype test for application of the Rabat Principles, and will implement
it for use as a policy tool of last resort.
In addition, in 2020—2022 Meta significantly increased its India-related content

moderation workforce and language support. As of the time of writing, Meta had
reviewers across 21 Indian languages, with both language and cultural expertise,
as well as multiple classifiers. Translating widely used languages has substantial
practical benefit for our users.
With regard to transparency on content trends, Meta began publishing detailed India
transparency reporting in June 2021. The India monthly report is available at our

Transparency Center and covers multiple policy areas for Facebook, Instagram, and
WhatsApp. In addition, we are actively seeking ways to improve our transparency
reporting around government requests to remove or restrict content, as reported to
the Oversight Board.
We will expand participation in our Resiliency Initiative, which empowers local

communities with digital tools to combat hate, violence, and conflict within and
beyond their networks. We will also seek to expand our counterspeech Search
Redirect Program, which redirects users who search using hate and violence-related
search terms towards resources, education, and outreach groups that can help.
With regard to product Interventions and other measures:
In 2021—2022, Meta designed and deployed multiple product features to try to prevent or mitigate acts
of abuse on WhatsApp. Meta also disclosed a detailed independent human rights impact assessment of
its expansion of End-to-End Encryption in April 2022. The recommendations will help guide our approach
to safer private messaging: helping to prevent abuse and to safeguard people’s privacy, and giving people
controls to help them stay safe.
We are developing consultation processes around our annual human rights disclosures at the time of
writing. We are implementing mergers and acquisitions human rights due diligence, to be incorporated
into our standardized due diligence processes. We are launching our human rights website in July 2022.
As per our corporate Human Rights Policy, we are sharing insights and actions from this due diligence in
our annual human rights report.
B. Product-Focused Human Rights Impact Assessments
Meta commissions detailed human rights impact current platforms. Product-focused HRIAs seek
assessments in instances of ground-breaking to provide guidance and identify considerations
new products, or systemic product changes to for product development.
01. End-to-End Encryption (“E2EE”)
E2EE keeps people and their personal In October 2019, Meta commissioned Business
communications safe from hackers, criminals and for Social Responsibility (BSR) to undertake an
authoritarian regimes. In 2016, we implemented HRIA of the expansion of E2EE. The subsequent
E2EE by default on WhatsApp and as an option on two years of work took place during the reporting
Messenger. In March 2019, we announced plans period. The HRIA was published in full in April
to extend this protection by default across our 2022, accompanied by a Meta response.
messaging apps.
Salient Risks:
BSR noted a number of positive opportunities and espionage. BSR also explained how human
that E2EE would enable, including further rights risks — such as child sexual abuse
protecting a diverse range of human rights such and exploitation, mis/disinformation, violent
as: privacy, freedom of expression, protection extremism, human trafficking — all operate
against cybercrime threats, physical safety, independently of E2EE, and should be addressed
freedom of belief and religious practices, and separately, without undermining E2EE and focus
freedom from state-sponsored surveillance on key marginalized communities.
The report made a number of recommendations

to address risks and maximize the benefits of
E2EE, divided among four categories of action:
product, process, product policy and public
policy; to wit:
• Recommendations about specific products

and features: user reporting; UX and user

testing; user education; languages; friction;
opt-in account linking.
• Recommendations for how Meta can detect
and mitigate human rights risk: process harm
prevention strategies, such as metadata
analysis; Machine Learning (ML) techniques for
proactive detection; holistic child rights
strategy; investigate client-side scanning
techniques; assess impacts of cross-app
communications.
• Recommendations for product policy changes:
consistent privacy policies and improved
transparency; define content standards; ML
explainability; improved user reporting; user
appeals transparency; grievance mechanisms.
• Recommendations for how Meta should engage
external stakeholders: advocate for end-to-
end encryption; engage policy makers; engage The report concludes that client-side scanning
stakeholders; collaborate with researchers; technologies as they exist today would undermine
quantify harms; collaborate with industry; train the integrity of E2EE and disproportionately
law enforcement. restrict people’s privacy and other human rights.
Meta’s response details our commitment to Meta is committed to implementing the vast
implementing 34 of the recommendations, partly majority of the recommendations and is working
implementing four, assessing the feasibility of diligently toward our plans for expanding E2EE as
another six and taking no further action on one — a means to help protect people and support their
the pursuit of client-side scanning technologies. human rights.
C. Due Diligence Exercises
Meta uses rapid, decision-useful due diligence run simultaneously to the crisis or product
exercises to provide salient risk assessments development to provide immediate, real-time
for moments of crisis, immediate product feedback to inform decision-making. These
development needs, and ongoing content assessments can be conducted in-house, or
policy decisions. with a third-party. They are designed to run
for shorter periods of time than our formal
This work can build on the diligence exercise HRIAs ( i.e. , weeks/months compared to years,
designating a country high priority, or be on average). While their form can differ, their
responsive to emerging needs. It can look function is uniform: actionable, decision-
back at moments of crisis to provide insights useful, and immediate insights to guide the
on improved processes going forward; or management of salient risks.
01. Ray-Ban Stories
Salient Risks:
In September 2021, in partnership with of the possible human rights implications of

EssilorLuxottica, Meta introduced Ray-Ban this product. Salient risks identified included:
Stories, a first-generation pair of smart glasses informed consent from bystanders, the safety of
that have cameras in the frames that can take people wearing the device, as well as its effect on
photos and videos, which you can download to a vulnerable groups that might be at increased risk
companion app and then upload to social media. of adverse effects from its operation, like women
and children, human rights defenders, or minority
Prior to launch, to mitigate human rights groups. It also examined privacy risks and future,
issues posed by people who may misuse such ongoing risks posed by any new technology, from
a device, Meta undertook an internal analysis visual search to the cloud storage of data.
The saliency assessment began by noting some

positive human rights impact of hands-free and

voice-controlled photo and video capabilities,
including help for people with disabilities, or
aiding accountability for human rights violators.
It then noted if and where use of the device
could implicate various human rights treaties,
including things like: whether it could be seen
by people with visual impairments, and if not
whether that could violate the rights to dignity
and privacy, or non-discrimination.
Its recommendations included other acts

of human rights due diligence, including
expanding user research to focus on vulnerable
bystanders, like homeless communities, minority
groups and human rights defenders, as well as
recommendations to conduct further field-level
focus groups to understand the product’s impact
in situ.
To mitigate these issues, the due diligence

exercise recommended exploring several
courses of action in consideration of bystanders,
such as: developing an acceptable use policy,
a do-not-disturb function, other signals for
bystanders or the option to tag content as being
from the device when it is shared.
D. Integrated Analytical Approaches
As part of our broader human rights due diligence

efforts, we put into practice various integrated,
analytical approaches that help us comprehend
our most salient risks.
01. Countries At Risk
As we note in Section 02 of our Human Rights Policy,

“We pay particular attention to
the rights and needs of users from
groups or populations that
may be at heightened risk of becoming
vulnerable or marginalized.”
In identified countries (and priority elections), prioritizing countries with the highest risk of
we take a more customized comprehensive offline harm and violence, every six months.
due diligence, risk mitigation and response To assess long-term conditions on the ground,
approach — acting quickly to initiate the process a cross-functional team of data scientists,
to remove content that violates our policies and political scientists and regional experts review
taking protective measures, including deploying quantitative and qualitative data twice a year
country-specific support. Since 2018, this work drawn from over 60 sources including Varieties
has been led by teams with expertise in human of Democracy (V-Dem), Uppsala Conflict Data
rights, hate speech, women and child protection, Program, the United States Holocaust Memorial
misinformation, and polarization. Many have Museum’s Early Warning Project, the Armed
lived or worked in high-risk countries and speak Conflict Location & Event Data Project, and the
relevant languages. World Bank. This data can include information
regarding civic participation and human rights,
Beginning in 2019, we developed a human societal tensions and violence, and the quality of
rights due diligence process for reviewing and relevant information ecosystems.
UNGP-Informed Factors to Determining Countries at Risk:
• Long-term conditions and historical context: We rely on regional experts, platform data and data
from more than 60 sources to assess the long-term conditions on the ground.
• How much the use of our products could potentially impact a context: We prioritize countries
based on a number of factors, including: where our apps have become most central to society,
such as in countries where a larger share of people use our products; where there is potential for an
increase in offline harms; and where social media adoption has grown.
• Current events on the ground: We give special consideration to discrete events that might magnify
current societal problems, such as local risk or occurrence of atrocity crimes, polarizing elections,
episodes of violence, and COVID-19 vaccination and transmission rates.
Guided by the above, we provide specialized evaluating and refining our policies to address
support and focus teams on countries most at evolving nuances of hate speech, identify
risk by working to drive human rights to the front groups at heightened risk of violence and human
of our product, policy, operations, and corporate rights abusers, or the potential for rumors and
decision-making. Our focus is threefold: remove misinformation to contribute to physical harm,
content that violates our policies, respect particularly in countries where ethnic and
people’s right to free expression, and help to keep religious tensions are present.

people safe both online and offline. • Rapidly enhance our technology and
enforcement to help keep our community safe
As part of our work, we took the following actions: when the risk of impact is greater, i.e, in advance
of elections and during periods of heightened
• Understand and engage with local contexts unrest. For example, we may reduce or restrict
and communities to ensure our enforcement distribution of content from accounts previously
accounts for local context. We expanded our found to post content violating our policies,
global network of third-party fact-checkers, while our teams investigate. Also, we may
invested significant resources in more than 30 offer additional features, e.g. providing people
countries with active conflict or societal unrest, in Afghanistan following the Taliban takeover,
and together with UN partners and dozens the ability to lock their profile as an extra layer
of local and global NGOs, we developed of privacy, security and protection.
programming to make online engagement safer. • Improve our moderation across languages
• Develop and revise policies to prohibit by adding more expertise.
harmful content and behavior by constantly
In 2021, we hired content moderators in

12 new languages, including Haitian Creole,
Kirundi, Tswana and Kinyarwanda.
i. Our Program In Action: Case Studies
Recent examples of our work in this area experienced a military coup in early 2021; and
include our efforts in Ethiopia, Mexico, Myanmar, (2) Ethiopia which, since November 2020,
Afghanistan, and Haiti. We highlight two case has experienced a brutal civil war characterized
studies which exemplify this work in action: by severe human rights violations and massive
(1) Myanmar which held elections in 2020 and humanitarian needs.
a. Myanmar
Evolving Challenge: military-linked commercial entities, and removed

Ahead of the November 2020 elections in Pages, Groups and accounts representing
Myanmar, and given the Myanmar Military’s (the military-controlled businesses. We based our
“Tatmadaw”) history of atrocities against the actions on:
Rohingya people, Myanmar was identified as • Extensive documentation by the international
high priority and at high risk for human rights community of these entities’ direct role in
violations. Though Meta’s efforts to protect funding the Tatmadaw’s ongoing violence and
platform integrity during the election were human rights abuses, which also formed the
successful, the political climate substantially basis for escalating sanctions imposed by the
deteriorated, culminating in the 2021 coup d’etat US, EU, and other governments.
by the Tatmadaw that unleashed a wave of • The UN Fact-Finding Mission on Myanmar’s

repression, attacks on peaceful demonstrations, 2019 report on the economic interests of the
and the silencing of dissenting voices. Tatmadaw, in line with the UN Guiding
Principles on Business and Human Rights.
Diligence: • We reaffirmed Meta’s ban on 20 military-
A cross-functional, multi-disciplinary team linked individuals and organizations in 2018,
came together to prepare for the elections; later, including Commander-in-Chief Min Aung
worked to identify the salient risks due to the Hlaing, for their role in severe human rights
coup, and took steps to help prevent and mitigate violations; and the decision to remove at least
those risks. six Coordinated Inauthentic Behavior networks
run by the Tatmadaw from 2018 to 2020.
Action: • Meta implemented a Myanmar-specific policy
Following the coup, the military government removing praise, support and advocacy of
unleashed a wave of human rights violations violence by Myanmar security forces against
against the people of Myanmar, protestors, civilians and/or arrest of civilians by the military
human rights defenders, journalists, and others. and security forces on our platform, under our
We treated the situation as an emergency. Our coordinating harm and publicizing crime policy.
Integrity Operations Center ran around the clock, • We offered a “Lock Your Profile” feature for
bringing together subject matter experts from Facebook, providing additional safety, privacy
across the company, including Myanmar nationals and security for our on-the-ground users.
with native language skills, to monitor and • Since the coup, we disabled the Tatmadaw
respond to threats in real time. True News Information Team Page, and MRTV
and MRTV Live Pages for continuing to violate
As a result, Meta took a series of strong actions: our policies. We also reduced the distribution
of content on at least 23 pages and profiles
• Meta banned the Tatmadaw and military- controlled and/or operated by the Tatmadaw so
controlled state and media entities from fewer people see them.
Facebook and Instagram, as well as ads from
b. Ethiopia
Evolving Challenge:
As of this report’s publication, Ethiopia is
undergoing a conflict that began in November
2020 between the government of Ethiopia and
forces in the northern Tigray region. As the
turmoil spreads, the situation remains volatile.
Ethiopia is an especially challenging environment
to address these issues, in part because there
are multiple languages spoken in the country.
Meta and its platforms are key avenues for
communication in Ethiopia.
Diligence:
Since early 2020, human rights due diligence for
Ethiopia has included field-based information
gathering and research; focused stakeholder
engagement; and formal due diligence. In 2021,
Meta commissioned forward-looking human
rights due diligence to help identify existing and

prospective human rights risks and inform internal
decision-making. We engaged activists, human
rights defenders, journalists, civil society groups
throughout Ethiopia and the diaspora, adding to
the insights gained regularly from human rights • Partnered with the Center for African
groups, NGOs, local civil society organizations, Leadership Studies to implement “My Digital
and other regional experts. World,” a series of live webinars on topics such
as online safety, privacy, digital citizenship,
Action: news and media literacy.
Using learnings from our due diligence, Meta • Established dedicated reporting channels for
increased efforts to tackle harmful content specialized international and local human
(e.g., violence and incitement, hate speech, rights and civil society organizations to quickly
coordinated inauthentic behavior, misinformation review problematic content they identify for
and harm), and addressed elections-related risks possible violations.
and content involving dangerous organizations • Launched a Lock Your Profile feature (as in
and individuals. We also: Myanmar) for Facebook, providing additional
safety, privacy and security for our on-the-
• Improved and simplified our reporting tools ground users.
to make it easier for Ethiopian users to report • Increased our investment in proactive
possible violating content, so we can investigate. detection technology that helps identify
• Rolled out a media literacy campaign, aimed at violating content such as hate speech before
educating and informing people on how to it is reported, including adding in Amharic and
detect potential false news, and ran billboard Oromo capabilities.
advertising campaigns across Addis Ababa, • Hired more Ethiopian nationals with native
the first of its kind in Africa, focused on language skills to monitor and respond to threats
informing and educating people on how to stay in real time and onboarded more Trusted Partners
safe online and use social media responsibly. to provide contextualized advice.
02. Security Policies
Meta strives to combat coordinated inauthentic accounts, we remove both the inauthentic and
behavior, defined as coordinated efforts to authentic accounts, Pages, and Groups directly
manipulate public debate for a strategic goal involved. Using both automated and manual
where fake accounts are central to the operation. detection, Meta continues to remove accounts
Such content on Meta’s platforms could lead and Pages connected to networks that we took
to salient human rights risks advanced through down in the past.
coordinated posts that:
Under our UNGP and GNI commitments, this
• attack people on the basis of their opinion, work is informed by stakeholders and performed
beliefs, or protected characteristics; with transparency. Since 2017, Meta has
• are connected to harm; reported on over 150 influence operations and
• depict violence; published information on threats we see from
• bully or harass users; nation states, commercial firms or unattributed
• encourage self-injury, misinformation or groups. We share information across the world
disinformation; or with investigative journalists, government
• exacerbate conflict, corruption, and instability officials, independent researchers, and industry
in conflict affected and high risk areas. peers to better understand and expose internet-
wide security risks, including ahead of critical
Meta performs ongoing diligence to review, elections. In 2021 alone, we removed 52
assess, and ultimately help remove deceptive networks that originated in over 30 countries

campaigns around the world. When we find and expanded our beta research platform
campaigns that seek to mislead people about (containing over 100 data sets) to researchers
who they are and what they are doing using fake studying influence operations.
i. Combating Emerging Threat Actors
The global threats we tackle have significantly reporting by coordinating false reports against
evolved since we started sharing our findings activists and others who publicly criticized the
about Coordinated Inauthentic Behavior in 2017. Vietnamese government in an attempt to have
Adversarial networks are always looking for new these users removed from Facebook.
ways to evade our policies. To account for this
constantly shifting environment, we build our
defenses to adapt and incorporate new layers
of defense. Our goal over time is to make these
behaviors more costly and difficult to hide, and
less effective.
• New Defenses against Coordinated Mass Mass Reporting:

Harassment (“Brigading”), Abusive We remove any adversarial
Reporting and Coordinated Violations networks we find where people
of our Community Standards work together to mass-report an
account or content to get
In December 2021, we shared our first report that it incorrectly taken down from
brought together multiple network disruptions our platform.
for distinct violations of our security policies:
Coordinated Inauthentic Behavior and two new

protocols — Brigading and Mass Reporting.
We are starting to remove coordinated violating
networks when we find people — whether they use
authentic or fake accounts — working together to
violate or evade our Community Standards.
Brigading:
We remove any adversarial
networks we find where
people work together to mass
comment, mass post or engage
in other types of repetitive mass
behaviors to harass others
or silence them.
As detailed in the December, 2021 Adversarial

Threat Report, we removed a network of accounts
that originated in Italy and France and targeted
medical professionals, journalists, and elected
officials with mass harassment. We also removed
a network of accounts in Vietnam for violating
our Inauthentic Behavior policy against mass
ii. Protecting Users Against Espionage
Cyber espionage actors typically target people we disrupt these operations, we take down their
across the internet to collect intelligence, accounts, block their domains from being shared
manipulate them into revealing information, and on our platform, and notify people who we believe
compromise their devices and accounts. When were targeted by these malicious groups.
Our Campaign Against Surveillance for Hire
In December 2021, we announced our action against surveillance-for-hire

entities operating on our platform, as part of a broader campaign effort that also
included a call for further regulation and industry collaboration to counter this
threat. Surveillance-for-hire entails commercial players duplicitously engaging in
reconnaissance, engagement and exploitation of users on social media platforms,
which can lead to salient human rights risks. During the reporting period, we took
action against seven different surveillance-for-hire entities who indiscriminately
targeted people in over 100 countries on behalf of their clients. These providers are
based in China, Israel, India, and North Macedonia, among others.

Broader Response to Abuse by Surveillance-For-Hire Groups
While cyber mercenaries claim that their services and surveillanceware are meant to
focus only on criminals and terrorists, our own investigation, independent researchers,
our industry peers and governments have demonstrated that targeting includes
journalists, dissidents, critics of authoritarian regimes, families of opposition and
human rights activists. In fact, for platforms like ours, there is no scalable way to
discern the purpose or legitimacy of such targeting. This is why we focus on enforcing
against this behavior, regardless of who is behind it or who the target might be.
To support the work of law enforcement, we already have authorized channels where
government agencies can submit lawful requests for information, rather than resorting
to the surveillance-for-hire industry. These channels are designed to safeguard due
process and we report the number and the origin of these requests publicly.
To protect people against cyber mercenaries operating across many platforms and
national boundaries, we engage in a collective effort with other platforms, policymakers
and civil society to counter the underlying market and its incentive structure.
III.
Providing Access to Remedy
The third pillar of Meta’s Human Rights Policy Help Center on platforms and apps, including an
concerns the provision of remedies for human appeals process to the first-of-its-kind Oversight
rights impacts. Board. Anonymous reporting for Meta personnel
is available through the Whistleblower and
Access to Remedy is a key component of the Complaint Policy.
UNGPs, with regard to both State and non-State-
based grievance mechanisms. Affected rights- We seek to provide meaningful transparency, with
holders should be able to claim remedy without routine publication of policies with human rights
fear of victimization. implications, information related to how content
determinations can be appealed, and decisions
At Meta, we strive to provide pathways for of, and responses to, the Oversight Board. When
stakeholders to report concerns, for Meta to content is removed, we notify users and clearly
review them, and for Meta to provide remedy identify pathways for content removal decisions
and remediation consistent with UNGP 31. to be appealed. Data regarding removal decisions
Meta maintains multiple grievance pathways, and other aspects of policy enforcement are

identified in the Code of Conduct, in the published through the Transparency Center.
INSIGHTS AND AC TIONS 2020 - 2021 72 III. PROVIDING ACCESS TO REMEDY

A. The Oversight Board
After extensive global consultation, Meta founded enterprises to “establish or participate in
the Oversight Board in 2020. The Oversight effective operational-level grievance mechanisms
Board was born, in part, to respond to concerns for individuals and communities who may be
that Meta needed an independent third party to adversely impacted [by their operations].” This
adjudicate violations of its rules, provide policy not only makes it possible to “identify adverse
recommendations and address access to remedy. human rights impacts as a part of an enterprise’s
ongoing human rights due diligence,” it also
The Board consists of experts from a variety “makes it possible for grievances, once identified,
of cultural and professional backgrounds, chosen to be addressed and for adverse impacts to be
because they are “experienced at deliberating remediated early and directly by the business
thoughtfully and collegially, skilled at making and enterprise, thereby preventing harms from
explaining decisions based on a set of policies compounding and grievances from escalating.”
or principles, and familiar with digital content
and governance.”14 For its part, the Meta Human Rights Policy Team
supported this project by working with colleagues
As its charter makes clear, the purpose of on initial due diligence, the Charter, member
the board is to: “[p]rotect free expression by selection, initial referrals, board education,
making principled, independent decisions about training of Meta staff for interacting with the
important pieces of content and by issuing policy board, due diligence on transparency and follow

advisory opinions on [Meta’s] content policies.” up to board recommendations.
The independence of the Oversight Board is Following Oversight Board decisions, Meta
strongly protected by its Charter, its Bylaws, its started the process of reinstating identical
members, and its funding. content with parallel context in the following
cases: Uyghur Muslims, Hydroxychloroquine,
Prior to its launch, Business for Social Azithromycin and COVID-19, and a Nazi Quote.
Responsibility completed a Human Rights Review These actions affect not only content previously
of the Oversight Board. This review made a series posted on Facebook and Instagram but also future
of recommendations, including prioritizing cases content. For cases where the board upholds our
that present the most severe human rights harms, judgment, we continue to ensure identical content
using the UNGP’s scope, scale and remediability with parallel context remains either up or down, in
criteria. The review also recommended that the line with the board’s decision.
human rights impacts of each case be reviewed,
and that the human rights principles of legitimacy, The Oversight Board has directly called for
necessity, and proportionality be used. One transparency when circumstances require it,
year later, a follow up human rights review by and brought in for discussion people critical of
Business for Social Responsibility found that the company.
“the Oversight Board’s governing documents
provide an increasingly valuable framework for
[taking a] human rights-based approach to
content decisions.”
The Oversight Board plays a critical, industry-

leading role in the fulfillment of Meta’s
responsibility under the UNGPs to provide
“Access to Remedy,” which requires business 14. https://oversightboard.com/meet-the-board/

In another example of its focus on transparency, The Board, via its case decisions and policy
when Meta referred to the Board a question recommendations, is defining what a human
regarding its restriction of then-US President rights-based approach to content means in
Donald Trump’s access to posting content on his practice, to a level of detail, practical application
Facebook page and Instagram account, the Board and intellectual rigor, that did not previously exist.
observed that Meta failed to mention a “cross-
check” system, and that, “Given that the referral The Oversight Board publishes a quarterly
included a specific policy question about account- transparency report highlighting its extensive
level enforcement for political leaders, many of efforts adjudicating the most difficult and pressing
whom the Board believes were covered by cross- speech issues facing Meta. Cases before the board
check, this omission is not acceptable.” include those submitted by Meta, users or others.
B. Code of Conduct
Meta’s Code of Conduct sets out guidelines as a core part of our expectations for all
for business conduct required of all Meta Meta personnel.
personnel, and is accompanied by multiple
channels to report concerns or violations, The ability to make anonymous reports is covered
and potential remedies. It helps embed by Meta’s Whistleblower and Complaint Policy.
human rights into our overall governance Retaliation or intimidation against individuals
by including respect for human rights reporting issues in good faith is strictly forbidden.

I V.
Protecting Human Rights Defenders
We’re building on our existing work to protect

defenders’ accounts — efforts that include
combatting advanced threat actors targeting
them, protecting them from incorrect content
removals using Cross-Check, offering advanced
security options, taking steps to thwart
unauthorized access to the accounts of defenders
who are arrested or detained, and partnering
with human rights organizations on outreach
and training.
One important workstream has been our work

on civic actors, who may face higher risks due
to their contributions to civic affairs, either
online or offline. These risks include threats to

privacy, reputation, well-being or physical safety.
Interventions applied for civic actors include
harassment monitoring, enhanced account
security and impersonation takedowns.
WhatsApp’s end-to-end encryption is widely

recognized to be essential for the protection
The fourth pillar of our Human Rights Policy of human rights defenders and journalists — as
is dedicated to protecting Human Rights well for racial and religious minorities, LGBTQ+
Defenders, as defined by the UN Declaration communities and other disadvantaged groups.
on Human Rights Defenders, recognizing them
as a priority group for support and protection.
We commissioned internal due diligence in

2019–20, the goal of which was to take a first Engagement with human rights
step toward designing a human rights defender defenders (and related UN actors)
program, mapping existing work, identifying has increased significantly. We are
gaps and opportunities, and suggesting possible driving regular engagement at the
next steps. We then established an internal local and global levels to inform
Human Rights Defender working group, which our understanding of related risks,
has significantly improved communication mitigations, and challenges. Our
and information-sharing across teams and engagements have included digital
helped to identify gaps, priorities, and potential security training for activists
interventions to better support human rights in a number of countries.
defenders. These interventions include steps
related to safety, security, research, policy
development, and product or operational support.
INSIGHTS AND AC TIONS 2020 - 2021 75 IV. PROTEC TING HUMAN RIGHTS DEFENDERS
A good example of innovative work in the
Africa region has been the creation of the Sub-
Saharan Africa (SSA) Women’s Working Group,
which was convened in 2021 for women activists
across SSA working on issues that intersect
with gender equality and digital spaces.
Participants represented grassroots, regional
and international organisations covering issues
ranging from online gender-based-violence,

feminist rapid-response services, online safety
for the LGBTQ+ community, tech-feminism
and tech law & governance.
This fund was formally launched in November
The working group helped us improve policies 2021, as the Human Rights Defender Fund and
and products affecting women users on Facebook Journalist Safety Initiative, in partnership with
platforms and include these voices in ongoing Civil Rights Defenders and International Center
work that requires engagement for content for Journalists and is being piloted in the Asia-
policies and other platform developments and Pacific region.
products affecting this constituency.
The fund supports devices and security
technologies, temporary relocation, and
emergency legal and safety assistance, as well
as mental health support to human rights
defenders and journalists. It also enables the
When Meta launched our development of peer-to-peer support programs
Human Rights Policy in March and relevant mental health practitioner networks.
of 2021, we also committed The fund is being distributed over a two-year
to create a fund that will give period, and managed independently.
offline assistance to human
rights defenders facing critical We have made good progress in integrating the
threats and support new protection of human rights defenders into our
digital security efforts. responses in countries at-risk and will continue
to look to develop and roll out new policies and
products in more regions — especially those
identified as most critical for the well-being of
human rights defenders and other civic actors.
INSIGHTS AND AC TIONS 2020 - 2021 76 IV. PROTEC TING HUMAN RIGHTS DEFENDERS
V.
Governance, Oversight, Accountability
The fifth core pillar of our Human Rights Policy

reflects our commitment to UNGP 16 to
embed respect for human rights throughout
our enterprise.

A. Governance Model
Meta’s Human Rights Policy and its commitments In a company of tens of thousands of employees
to human rights are enterprise-wide. Nick and contingent workers, and with billions of
Clegg (President, Global Affairs) and Jennifer people using our apps and services, a key
Newstead (Chief Legal Officer) oversee policy challenge for the Human Rights Policy program
implementation. Salient risks are reported is to ensure that we leverage our expertise and
periodically to the Audit and Risk Oversight resources for maximum impact.
Committee of the Meta Board of Directors.
We seek to ensure that salient human rights
The global Human Rights Policy Team drives issues are directed to the most relevant decision-
application of the Policy, setting the company’s making forum. The Human Rights Policy Team
human rights agenda, identifying priorities, functions in an umbrella governance structure:
shaping strategy, inspiring and informing consulting with and convening the functional
colleagues, and engaging with human rights units implementing human rights-informed policy
activists and experts. This team was created and product changes.
in 2019, with the appointment of Miranda Sissons
as its inaugural director and by the end of
2021 had four full-time staff, with several more
15. As of March 2022 the global Meta Human Rights Policy Team consists of 9 full
positions onboarding in 2022.15 time employees.
INSIGHTS AND AC TIONS 2020 - 2021 77 V. GOVERNANCE, OVERSIGHT, ACCOUNTABILITY

Teams working on community operations, law • Driving human rights due diligence processes
enforcement and security, responsible innovation, with key product and public policy colleagues;
product integrity and public policy (among others) • Participating actively in the at-risk country
regularly identify, prevent, or mitigate salient work, ensuring that human rights considerations
human rights risks. Similarly, teams focused on are effectively prioritized;
civil rights, Africa, the Middle East and Turkey, • Coordinating work on human rights
sustainability, International Organizations, and defender protection;
Stakeholder Engagement also work to implement • Representing Meta at various human rights
policy goals. fora including the GNI, B-Tech, and with key UN
and NGO human rights agencies.
Key activities to ensure proper governance over
human rights responsibilities include: Meta’s Board of Directors’ Audit & Risk Oversight
Committee’s charter provides for the committee’s
• Convening and driving the work of various cross- oversight of programs, policies and risks related
functional, inter-departmental working groups to social responsibility, which includes oversight
dedicated to the implementation of the Human of human rights-related risks. Beginning in 2021,
Rights Policy; the Director of the Human Rights Policy Team
• Joining cross-functional, inter-departmental updates the Audit and Risk Oversight Committee
working groups to provide human rights at least annually.
perspective and principles to their work;

In the course of the last two years, Meta’s teams have worked across
a wide range of human rights issues, including but not limited to:
• Business and Human Rights; • Child rights;

• Freedom of Expression (particularly • Regional human rights practices
proportionality); and instruments;
• Privacy; • Modern slavery;
• Freedom of Assembly; • Lawful use of lethal force and law
• Political participation; enforcement principles;
• Civic engagement; • Right to health;
• Non-discrimination; • Machine learning techniques;
• International humanitarian law; • The San Remo Principles;
• International criminal law; • Open source evidence investigation;
• Transitional justice processes, particularly • Quantitative and qualitative research
process design for truth seeking and processes and indicators;
reparations initiatives; • Comparative law;
• Agile software development processes; • Know your customer (KYC) processes; and
• Atrocity crime prevention; • Adult learning practices.

B. Stakeholder Engagement
Our Human Rights Policy is founded on deep well as develop appropriate mitigation strategies,

engagement with rights-holders and other including fair and transparent policies. During
stakeholders.16 Stakeholder Engagement is the reporting period, we also sought to pioneer
thus relevant to every pillar of the policy. new methodologies of direct rights holder
consultation. For example, we combined a large
Stakeholder engagement is a company-wide scale quantitative survey of rights holders with
effort, integrated into several company processes. traditional methods of stakeholder consultation in
When developing the Human Rights Policy, our human rights due diligence in the Philippines.
for example, the Human Rights Policy Team
engaged extensively with human rights experts In another important example, the Content Policy
and activists, including activist investors. Team Stakeholder Engagement team strengthens our
members are in regular contact with human rights content policies by bringing global knowledge and
organizations, activists, academics and others feedback into the policy development process. This
to hear concerns and criticisms, seek advice and team builds relationships with NGOs, academics
brief on ongoing developments. and other thought leaders, guided by our core
principles of inclusivity, expertise, and transparency.
As described in earlier sections of this report,
such engagement includes consultations and co- In all, the team covers the following policy areas:
design with human rights defenders on measures Community Standards, advertising policies,
to protect account security, engagement on product policies, misinformation, content ranking,
coordinated inauthentic behavior and espionage, and news integrity.
roundtables with regional civil society actors
ahead of elections, and much more. Work with stakeholders crafting our policies
includes:
Stakeholder engagement is also an important part
of Meta’s human rights due diligence process. It
enables us to identify and understand the human
16. By “stakeholders” we mean rights-holders, all organizations, and individuals
rights impact of our products and services as who are impacted by, and therefore have a stake in, Meta’s policies

Understanding State Media
To inform our policy definition and criteria
Engagement on Stereotypes for state media, we consulted with
We created a policy on harmful more than 65 global experts specializing
stereotypes. As part of our approach in media, governance and human rights.
to hate speech and harmful stereotypes, This input was crucial to helping us
we convened several academic understand the different ways and degrees
roundtables and connected with external to which governments exert editorial
stakeholders who helped us identify control over certain media entities.
the risks associated with certain
stereotypes in their communities.
Recommendation Guidelines
In developing our Recommendations
Guidance on “Public Figures” Guidelines for the Facebook app
We adopted a definition of “public figure” and Instagram, we consulted with
for our bullying and harassment policies 50 leading experts who specialize in
based on engagement with academics, recommender systems, expression,
civil society organizations, free expression safety and digital rights.
experts, human rights defenders,
journalists and creators/influencers.

C. Program Spotlight: Trusted Partners
Meta’s Trusted Partner (TP) program is an In the past, TPs have alerted us to developments
escalation channel made up of over 400 key non- and risks that our internal monitoring efforts had
governmental, not-for-profit, humanitarian, and missed. In Myanmar, for example, TPs provided
international organizations in 113 countries who us with additional context about the recent coup
report content, accounts, and behavior that we and online risks that emerged in its aftermath.
review in context. In Ethiopia, our TPs helped us combat harmful
misinformation and hate speech stoking inter-
Trusted Partners are critical allies in our efforts ethnic violence. In Afghanistan, our civil society
to understand the local impact of our services, partners helped us to mitigate the targeting of
stay abreast of emerging trends, correct mistakes international aid workers and other vulnerable
made in scaled review, and improve policies and persons in our platforms.
enforcement in specific policy areas such as
misinformation and harm. Their reports provide The Trusted Partner program also provides a
valuable insights, data points, and edge cases, vehicle for amplifying the voices of marginalized
which help inform policy development processes communities, incorporating relevant information
and help keep our users safe both on and offline. about the risks they face on our platforms,
and providing support to under-represented
and disproportionately-impacted groups.

D. Engagement with International Organizations
Part of our accountability process under our development processes; and have briefed and
governance pillar is transparent, collaborative, shared insights from our at risk countries with
and reciprocal engagement with the UN. Our multiple relevant groups.
International Institutions and Relations team • Meta briefed the UN Security Council’s October
supports engagement across the UN and 28, 2021 Arria Formula Meeting on hate speech
intergovernmental organization system, including and social media.
dialogue with UN human rights actors, which • A key part of our engagement with the UN
continues to broaden and deepen both at global system and other multilateral actors has been
and country level. to encourage a strong defense of an open
and unified internet, with an emphasis on the
As long-running human rights issues continue many negative human rights impacts of internet
to manifest online, we see this collaborative fragmentation. Meta is also part of the
discourse with the UN human rights system as Internet Fragmentation group.
particularly valuable — this includes thematically • Meta has been represented in several human
focused actors such as UN Human Rights rights and hate speech related panels, including
Council-mandated Special Rapporteurs, but by senior executives, to further discourse
also broader dialogue on tech and human rights and thinking around tackling hate speech and
across the UN Secretariat and with specialized misinformation.
agencies such as UNESCO. • As part of UNESCO’s Social Media 4 Peace

project, we have worked with UN representatives
• The International Institutions and Relations and in Bosnia and Kenya to discuss with civil society
Human Rights Policy Teams have met regularly and government stakeholders, ideas of how
with the UN Office of the High Commissioner social media platforms can promote peace.
for Human Rights since 2018, and we have been • Since June 2021, we have voluntarily disclosed
active members of the B-Tech Project. 16 productions with organic content from
• In addition, we meet with various entities of hundreds of Pages, and Basic Subscriber
the UN system including the Office of the information for hundreds of Pages and Profiles,
Special Advisor for the Prevention of Genocide, to the Independent Investigative Mechanism
as well as UN Special Rapporteurs on: Freedom of for Myanmar (IIMM) which is seeking
Expression, Human Rights Defenders, Protection accountability for the most serious international
of Minorities, Myanmar, Racism and Racial crimes and violations of international law
Discrimination and Freedom of Religion and committed in Myanmar.
Belief to discuss the intersection of human rights • In September 2021, Meta joined the UN
and technology, and its relevance to the actions of Global Compact, the world’s largest corporate
the UN system across peace and security, human sustainability initiative. This marked an
rights, and humanitarian agendas. important milestone in Meta’s commitment
• We also work with UN entities as trusted partners to human rights, labor, environment and
at both global and country level — including anti-corruption. To take actions that advance
Haiti, Bosnia, and Ethiopia. We work with them societal goals, Meta has committed to the
to mitigate content risks and to better understand Global Compact’s 10 Principles on Human
the links between on-platform conduct and off- Rights, Labor, Environment, and Anti-Corruption.
line harms. We are collaborating on issues of racial
justice, the implementation of the UNGPs and the
critical importance of end-to-end encryption. We
are contributing our expertise and experience to a
variety of human rights reporting and norm

A Final Note
“The saddest aspect of life right now

is that science gathers knowledge faster
than society gathers wisdom.”
IS A AC A S IM OV
Isaac Asimov once wrote, “The saddest aspect of This conflict, along with other crises in countries
life right now is that science gathers knowledge across the world, underscores how important it is
faster than society gathers wisdom.” for Meta and other companies to live up to their
human rights commitments — for wisdom to
While this report was being written, a war in always accompany knowledge.
Ukraine was fiercely waged, precipitated by a
Russian invasion. As with the Arab Spring of 2011, We hope this report gives detailed insight into the
content on social media is seen to have a profound work Meta has done to identify, mitigate, and prevent
role in shaping how this conflict is perceived, and human rights risks from 2020 — 2021. We look
potentially, its dynamics. forward to more work, and more reporting, to come.
INSIGHTS AND AC TIONS 2020 - 2021 83 A FINAL NOTE

Meta Human Rights Report July 2022

Uploaded by

Copyright:

Available Formats

Meta Human Rights Report July 2022

Uploaded by

Document Information

Original Description:

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

Meta Human Rights Report July 2022

Uploaded by

Copyright:

Available Formats

JULY 2022

META HUMAN RIGHTS REPORT

INSIGHTS AND AC TIONS 2020 - 2021 1

META HUMAN RIGHTS REPORT

INSIGHTS AND AC TIONS 2020 - 2021 2 ABOUT THIS REPORT

META HUMAN RIGHTS REPORT

INSIGHTS AND AC TIONS 2020 - 2021 3 EXEC UTIVE SUMMARY

META HUMAN RIGHTS REPORT

INSIGHTS AND AC TIONS 2020 - 2021 4 EXEC UTIVE SUMMARY

Right to Freedom of Opinion and Expression

Meta’s platforms advance freedom of expression — which is interdependent with

META HUMAN RIGHTS REPORT

The right to privacy is essential to realizing the right to freedom of expression.

INSIGHTS AND AC TIONS 2020 - 2021 5 EXEC UTIVE SUMMARY

Right to Life, Liberty and Security of Person

Our Facebook Community Standards and Instagram Community Guidelines seek to

META HUMAN RIGHTS REPORT

Moreover, as we detail in this report, protecting Human Rights Defenders is a core

Rights to Equality and Non-Discrimination

INSIGHTS AND AC TIONS 2020 - 2021 6 EXEC UTIVE SUMMARY

In developing youth-focused features, we are listening to leading experts on child

Right to Public Participation, to Vote, and to be Elected

Meta is committed to protecting the right to public participation, giving everyone a

META HUMAN RIGHTS REPORT

INSIGHTS AND AC TIONS 2020 - 2021 7 EXEC UTIVE SUMMARY

I. Applying Relevant Policies;

Meta’s respect for human rights — and their

META HUMAN RIGHTS REPORT

• Detailed human rights impact assessments,

INSIGHTS AND AC TIONS 2020 - 2021 8 EXECUTIVE SUMMARY

Users can challenge some decisions regarding

META HUMAN RIGHTS REPORT

Meta created a global Human Rights Policy Team

INSIGHTS AND AC TIONS 2020 - 2021 9 EXEC UTIVE SUMMARY

META HUMAN RIGHTS REPORT

INSIGHTS AND AC TIONS 2020 - 2021 10 EXEC UTIVE SUMMARY

Human Rights and Policy Timeline 14

Part 1: Meta’s Human Rights Commitments 16

Meta’s Salient Risks 25

META HUMAN RIGHTS REPORT

Part 2: Meta’s Human Rights Policy in Practice 28

Applying Relevant Policies 29

Actions with Regard to Salient Risks 31

INSIGHTS AND AC TIONS 2020 - 2021 11 TABLE OF CONTENTS

Human Rights Impact Assessments 55

Combating Emerging Threat Actors 70

Protecting Users Against Espionage 71

Providing Access to Remedy 72

Protecting Human Rights Defenders 75

META HUMAN RIGHTS REPORT

Program Spotlight: Trusted Partners 81

Engagement with International Organizations 82

INSIGHTS AND AC TIONS 2020 - 2021 12 TABLE OF CONTENTS

META HUMAN RIGHTS REPORT

Last year, we launched our first Corporate

Under this policy, we committed to publishing an

INSIGHTS AND AC TIONS 2020 - 2021 13 INTRODUC TION

META HUMAN RIGHTS REPORT

Release of Civil Rights

We expanded our policies that prohibit veiled and implicit threats. We added a policy