Content Moderation and Freedom of Expression Handbook

"With a handful of major companies wielding immense power over what billions of users can see and share online, concerns about freedom of expression, media diversity, and privacy breaches have grown."

This handbook offers an overview of the current state of content moderation on the largest social media platforms and explores some of the main issues current content moderation practices raise from a freedom of expression perspective. The handbook is designed as a resource for civil society actors and other stakeholders who are involved in content moderation issues or seeking to engage in the topic, with the hope that it will support their engagement with social media platforms and policymakers in their efforts to protect freedom of expression.

As explained in the handbook, "In their early days, social media platforms were widely seen as a powerful force for good, liberating free expression, enabling connections between people, and spearheading a democratic revolution across the world. This perception has now changed. Today, a very small number of large social media companies act as gatekeepers, controlling what a huge number of people get to see or say online. They have a direct impact on the dynamic of content distribution, as well as on online media diversity and freedom of expression. This significant power and influence is coupled with the fact that the business models of the largest social media companies are often based on the collection of vast amounts of data about their users and their online habits (behavioural data) and the monetisation of this data through online (targeted) advertising. This significantly interferes with users' right to privacy and can have a negative effect on freedom of expression. Of particular concern is the spread of 'hate speech' and 'disinformation' on online platforms. Social media companies have been accused of prioritising profit over user safety by using algorithms that promote the consumption of harmful content, including 'hate speech' and 'disinformation'." For this reason, there are increased calls for social media companies to step up their content moderation efforts and tackle such problematic content.

This handbook is part of a project that is seeking to address this issue - the Social Media 4 Peace project, which is being implemented by the United Nations Educational, Scientific and Cultural Organization (UNESCO) and ARTICLE 19 in Bosnia and Herzegovina, Kenya, Indonesia, and Colombia with the support of the European Union. The overall objective of the project is to strengthen the resilience of societies to potentially harmful content spread online, in particular "hate speech" and "disinformation", while protecting freedom of expression and contributing to the promotion of peace narratives through digital technologies - notably, social media. ARTICLE 19's contribution to the project focuses on concerns raised by current content moderation practices on the largest social media platforms in the four target countries, and this handbook is designed to support that work. (See Related Summaries below for more information on the Social Media 4 Peace project.)

The handbook includes numerous concrete examples and cases to illustrate the questions raised by different standards, practices, and policies pertinent to content moderation. It builds upon ARTICLE 19's policies and expertise in content moderation and platform regulation and reflects ARTICLE 19's longstanding calls that measures responding to problematic content, including "disinformation" and "hate speech" must always conform with international standards on freedom of expression and other human rights.

The handbook includes the following two main sections:

Section 1: Applicable International Human Rights Standards

This section focuses on three main issues:

• the applicable standards for protecting freedom of expression online that should guide any measures adopted by states and social media companies in content moderation;
• how international freedom of expression standards apply to "hate speech" and "disinformation": two categories of speech that capture a wide range of expression but lack a uniform definition under international human rights law; and
• the human rights responsibilities of social media companies, how they differ from those of states, and what this means in practice for content moderation.

Section 2: Content Moderation in Practice

This section looks at three main issues:

• the contractual relationship between users and the largest social media companies through terms of service and community standards that govern online speech, and the issues raised by regulating speech by contract;
• the role of regulatory frameworks in content moderation - in particular, the concept of intermediary liability and the recent trend towards greater regulation of social media companies;
• content moderation processes typically applied by the largest social media companies - including automated systems, human reviewers, and user and third-party reporting - with a special focus on the shortcomings of automated content moderation systems.

The handbook complements research conducted by ARTICLE 19 under the Social Media 4 Peace project, which focuses on Bosnia and Herzegovina, Indonesia, and Kenya. The research revealed the disconnect between the practices of these global companies and the local communities affected by their content moderation decisions. Social media companies routinely overlook the voices of local communities and fail to consider cultural, social, historical, economic, and political context when moderating content. The findings also show how flawed content moderation practices can transform social media platforms into hotbeds of "disinformation", "hate speech", and discrimination, with significant implications for communities and for peace and stability in post-conflict countries. (See Related Summaries below for more information about the project and these reports.)