Op-ed: Countering Hate Speech on Social Media

The availability of social media platforms have provided the ability for people to share their opinions and views easily and quickly. The ease of online content creation and the anonymity of social media has contributed to an increase in offensive and hate speech content, and it has also fueled political polarization.

In recent years, social media companies have faced accusations about their role in amplifying hate speech, fake news, and division in society.  There is actually no international legal definition of hate speech, and it is contentious and contested to clearly define it. For instance, the European Court of Human rights defines hate speech as “all forms of expression verbal or written which spread, incite, promote or justify hatred based on intolerance also on grounds of religion”.  

On the other hand, the United Nations defined hate speech as "any kind of communication in speech, writing or behavior, that attacks or uses pejorative or discriminatory language with reference to a person or a group on the basis of who they are, in other words, based on their religion, ethnicity, nationality race, color, descent, gender or another identity factor."

As a matter of fact, hate speech holds different meanings for different people in different cultures. However, there is a consensus that hate speech is harmful and toxic, and freedom of speech does not give the right to offend others.

While some speech can be categorized as incorrect, or somehow rude or not acceptable, hate speech is totally wrong and cannot be tolerated. Indeed, there is a type of discourse that society cannot tolerate: for example, racism is intolerable and totally wrong given its lasting negative impact, and the serious harm inflicted to the targeted individuals or groups. And the widespread sharing of hate speech posts can worsen the damage.

In fact, offensive speech online has, in the past, led to several acts of violence such as the Christchurch terror attack in New Zealand in 2019, as it was reported that they were inspired by racist messages shared online. Hate speech also fuels prejudice, poisons debates, and creates a toxic atmosphere that may lead to polarization and division within societies. This is especially true during election campaigns, as accusations, propaganda, and disinformation online can increase and this situation may create pre-conditions for riots and violence.

Therefore, it is evident that hate speech is a major threat to peace that should be eradicated today to prevent horrific crimes that could happen tomorrow. Reducing hate speech will reduce the great deal of harm that can be caused by social media.

How can we address the issue of hate speech online? Is there a single way to solve this major concern? What’s the best response to hate speech?  From a legal perspective, as hate speech can lead to violence with calls for hostility and discrimination, it should be clearly categorized and prohibited under international law, especially as online hate speech has no borders and cannot be solely addressed only by national laws in each country.

Social media companies are also concerned by the extent of the problem, as they have been trying to reduce hate speech using advanced technologies such as Artificial Intelligence (AI) and Natural Language Processing (NLP) techniques. However, despite some advances, these AI-based methods have been blamed for inaccurate flagging of hate speech and eventually suspending incorrectly the related accounts. The main problem with the automatic detection of hate speech online is the fact that the algorithms used are based on current technologies which are far from being perfect, even if noticeable improvements were made in recent years.

AI detection models may fail to detect real-world data which is often different from the data used to train the model. For example, in the case of Arabic hate speech detection, if an AI model is only trained to detect hate speech terms written in Modern Standard Arabic, it will eventually fail to detect hate speech terms written in dialectal Arabic varieties. So an effort still has to be made in this area to address low-resourced languages and dialectal varieties frequently used in social media. In the future, with more investment in language technologies targeting Arabic and its different dialectal varieties, and other low-resourced languages, AI-based systems are expected to improve and achieve better detection performance.

However, the problem of hate speech demands a collective and holistic approach, and cannot be solved solely by building more accurate and advanced AI tools for hate speech detection. While using such advanced technologies is important and can help to reduce the massive spread of such posts, there are also other means to help eradicate hate speech online, such as adapted legal instruments and regulations which can be updated to cover hate speech.

Finally, education and media literacy are powerful ways of preventing and potentially ending hate speech. Indeed, it is important to raise awareness and encourage individuals to behave in a respectful manner online, and to also provide training on legal resources so that online hate speech offenders can be reported and prosecuted, wherever they are.

Countries around the world can be inspired by the Finland approach to tackling fake news through an ambitious digital media literacy program targeting schools, which led to the country being top-ranked in the media literacy index according to the Open Society Index. In fact, as stated by the authors of the Media Literacy index: “High-quality education and having more and more educated people is a prerequisite for tackling the negative effects of fake news and post-truth….While some regulation is necessary, education seems to be the best all-round solution.”.

The average social media user need to have a clear understanding of hate speech and its consequences and should be made aware of the ways to report it, exposing such behavior and communicating it to the relevant authorities. Such awareness can be spread through proper education and campaigns to learn how to respond to hate speech, as well as writing against it. The approach used to teach children how to understand and react to fake news can also be adapted and used to teach them how to react to hate speech.

Certainly, it is our duty as educators and society at large to teach children digital media literacy and critical thinking skills that can help to reduce hate speech content online.