Cyber-racism on online platforms

Global hyperconnection has merely evolved modes of racism from the local to the globalised. Social media platforms have come to be manifestations of community and society, reflecting thus the very same structures and practices of discrimination digitally, assisted, enabled, if not exacerbated by features unique to the internet: anonymity and pseudonymity, lack of accountability and direct repercussions, and the distancing of perpetrator and victim.


Defining cyber-racism

Cyber-racism goes beyond interpersonal racism, encompassing all forms of racism made digital, from institutional to structural to systemic. A detailed outline of each of these forms of racism necessitates a much more in-depth analysis.

For the purposes of this study, interpersonal cyber-racism refers to discrimination based on race occurring between persons, perpetrated either individually or via groups, e.g. online hate speech. Some analyses involve other grounds of discrimination under the umbrella of racism, including religion, nationality, and xenophobia or anti-migrant hatred. This study particularly looks at race or ethnicity, but briefly mentions the other grounds in light of the EU Online Code of Conduct. Institutional racism refers to the policies and practices within or across sociopolitical institutions that generate racial inequality. Such institutions do not exist in a similar manner digitally, and institutional cyber-racism thus (often) manifests itself through the private actors developing and moderating much of the Internet today.

With regards to social media platforms, we will focus on interpersonal racism, focusing in particular on the experiences of racism of racialised persons online. We will also focus to some degree on institutional racism when analysing the policies and practices of social media platforms in platform moderation, as well as policy responses of the EU and the impact thereof.


This analysis draws from, reviews, and visualises data collected from several conducted surveys, studies, and monitoring reports. The analysis will look at current data on the experiences of racialised persons online and assess this in light of the monitoring evaluations by the European Commission of the Online Code of Conduct. The analysis will look more specifically at current issues in platform moderation before closing on proposals for moderation reform.

Experiences of racialised persons online

Various studies point to a marked increase in instances of discrimination online, particularly since the COVID-19 pandemic. However, specific data on the experiences of persons based on racialised or ethnic identity is not surveyed in Europe as it is in the US – survey data on the latter will thus be used for this analysis.

In the US, a YouGov Survey found a significant year-on-year increase in harassment, including severe harassment, reported by Black Americans from 28% in 2021, to 40% in 2022, and 64% in 2023.

Of this 64% of Black Americans who experienced online harassment, slightly more than half (54%) attributed said harassment to their race or ethnicity. As such, around 1 in 3 Black Americans have faced online harassment because of their race.

Though the share of Asian Americans who experienced online harassment fell in 2023 to 44%, there was a significant increase in harassment from 23% in 2021, to 46% in 2022.

Of this 46% of Asian Americans who experienced online harassment, nearly half (47%) attributed said harassment to their race of ethnicity. Thus, nearly 1 in 4 Asian Americans have faced online harassment because of their race.

Ineffective moderation?

Despite the concerning rise of online harassment of racialised persons, regulatory efforts, both by regulatory authorities and the industry itself, have shown no significant impact.

In response to the rise of online discrimination, and as part of the EU’s anti-racism action plan, the European Commission adopted the Code of Conduct on Countering Illegal Hate Speech Online, a voluntary set of policies agreed with Facebook, Microsoft, Twitter and Youtube in May 2016, the implementation of which is evaluated through a yearly monitoring exercise and through a commonly agreed methodology allowing for comparison over time. Instagram, Snapchat and Dailymotion took part to the Code of Conduct in 2018; in 2019, TikTok in 2020, LinkedIn in 2021, and Rakuten Viber and Twitch in 2022.

Key figures from the 4th to 7th evaluation of the Code of Conduct shows little progress between 2019 and 2022. Notably, the assessment times of notifications based on the Code of Conduct’s target of assessment within 24 hours has been steadily decreasing relative to 2019, a fall from 89% on average in 2019 to 64,4% in 2022. Only TikTok had a better performance than in 2021, while all other platforms measured worse.

Two other key indicators of the Code of Conduct have shown minor improvement between 2021 and 2022, however with no significant improvement year-on-year and a decrease from their peaks in past years.

The removal rate of content notified has fallen from 71,7% on average in 2019 to 63,6% in 2022. Only YouTube had a better score than in 2021, while all other platforms measured worse.

Meanwhile, the percentage of notifications receiving feedback remains insufficient, with marginal improvement from 65,4% on average in 2019 to 66,4% in 2022. YouTube in particular scores significantly lower than other platforms, providing users feedback in 13,5% of notifications, a score more than 50% lower than the average.

It is also worth noting that the most commonly reported grounds of hate speech are some form of hatred based on race, ethnicity, religion, or nationality – measuring at 75,2% of all notifications.

As it stands, the Code of Conduct, amongst other measures taken by major online platforms, shows little actual year-on-year change, if not a fall in aspects such as assessment times, while incidences of cyber-racism continue to rise year-on-year. It is also worth challenging the indicators in which ‘success’ tackling online hate is measured, as content removal for example, as will be highlighted below, does not necessarily entail better protection of marginalised communities.

Double standards and discrimination in platform moderation

Another shift since the COVID-19 pandemic has been the decrease of human reviewers, and increasing use of automated tools in the moderation of online platforms, and it is not clear if platforms have subjected these tools to any form of independent review for accuracy or efficacy.

Current biases within automated processes for reviewing and removing content contradictorily puts speech from marginalised communities at risk of over-removal. The automated flagging and removal of terrorist content, for example, has led to the restriction of Arabic-speaking users and journalists. Studies have further found that automated tools incorporating natural language processing to identify hate speech instead amplify racial biases. Models for automatic hate speech detection were 1.5 times more likely to flag tweets of Black people as offensive or hateful, with tweets in African American Vernacular English (AAVE) twice as likely to be labelled as ‘offensive’ or ‘abusive.’ Another study on five widely used data sets for studying hate speech found ‘consistent, systemic and substantial racial biases in classifiers trained on all five datasets.’ This could be attributed to an inability of such tools to understand contextual nuances in the use of language, such as terms which amount to slurs when directed against marginalised groups that are used neutrally or positively within that group; to the inadequacy of training data in other dialects or languages; to the translation of norms and notions of hate speech from one cultural and lingual context to another.

Human moderators can similarly amplify racial biases simply due to increasing workloads and pressure to finish their queues. This is further exacerbated by the externalisation of moderation to remote countries relative to the market moderated, and as such a lack of local and cultural knowledge or even appropriate language skills to accurately make determinations on content. Furthermore, the increase of regulatory corporate accountability for such hatred without sufficient safeguards or limits or even the threat of government regulation has led to a tendency towards risk-averseness and to remove content than not.

Platform rules governing terrorist content, hate speech, and harassment themselves are designed to give platforms broad discretion, with broad and imprecise application against marginalised communities, yet narrow against dominant groups. Coupled with overbroad tools for mass removals has resulted in ‘mistakes at scale that are decimating human rights content.’ Power dynamics are also not fully incorporated in rule construction, as seen in an internal Facebook training document from 2017 revealing that, in a set of female drivers, Black children, and White men, under its hate speech policy, only White men would be protected due to its distinction of protected and quasi- or nonprotected characteristics. This has resulted in inequitable enforcement practices that seemingly protect powerful groups and their leaders whilst subjecting marginalised groups to over-enforcement or -removal.

As such, current platform policies and practices, including an increasing shift away from human reviewers towards (purely) automated moderation, subject marginalised communities to heightened scrutiny and risk of over-enforcement, whilst insufficiently protecting them from harm.

A need for platform transparency and regulation

A crucial starting point towards more equitable and effective moderation of cyber-racism begins at more transparent practices. The Commission has stressed feedback and transparency to users as one of its key indicators in tackling online hate. With the DSA coming into force at the end of 2022 and subsequent obligations on transparency and feedback to users’ notifications, it remains to be seen how this impacts consequent monitoring assessments of online hate.

Moreover, platforms need to reform moderation to ensure that the voices of racialised communities are centred, and thus reorient moderation towards not just their protection but towards repairing, educating and sustaining communities. This requires the involvement of racialised communities in the formulation of moderation rules and policies, and a reassessment of the connection between speech, power, and marginalisation in the drafting of such policy, in the training of automated tools, and in the enforcement of moderation. As AI development continues, and thus its involvement in platform moderation, the inherent racial biases within it need to be challenged.

Comprehensive and systematic reform to platform moderation is essential to combatting cyber-racism, and where this cannot be expected from platforms themselves, legislators must intervene to ensure platform accountability, transparency, and fairness, and the equitable protection of the marginalised made digital.

My letter to FYEG: A politics of principle.

Dear FYEG,

I write this with a heavy heart.

When I first engaged in politics as a young teen in Malaysia nearly a decade ago, I made it my mission to fight for the oppressed and the voiceless. A politics of principle – of honesty, integrity, inclusivity, solidarity, and ultimately of humanity. A compass that has guided me ever since, and a promise I made to myself of principles I would never abandon.

Upon moving to Europe in 2019, that flame burnt ever stronger. Fighting for the oppressed meant fighting for change in whatever way I could – organisational, societal, institutional, systemic. It meant making use of the opportunities I had to deepen my knowledge on racial justice, on decoloniality, on intersectionality, and so much more. It meant challenging the systems of oppression against the Global South and the racialised Other in the Western world.

It meant learning from, platforming the voices of, and fighting for the rights of the racialised Other – for a decolonial imaginary. It meant making my voice heard, and with it striving for a politics of the marginalised made central. A politics of principle.

It was this message of unapologetic, radical racial justice and inclusion with which I am ever grateful that our Member Organisations (MOs), all of you, elected me to the Executive Committee (EC) in 2022, and even more resoundingly so in 2023.

It has not been easy to fully strive for this over the past year, in an organisation that is unfortunately still so unfamiliar, even unwelcoming, to people like me. But the past two weeks has been especially challenging and hurtful.

As the crisis in Gaza unravelled and continues to unravel, as it became increasingly clear that Israel is inflicting a genocidal retaliation against the Palestinian peoples, the prioritisations of my fellow EC went against everything I stand for.

Instead of discussing the condemnation of genocide, apartheid, and settler-colonialism, amongst other crimes, of Israel – the EC was preoccupied with discussing the political optics and consequences criticising Israel would have on Grüne Jugend, the German Young Greens.

Instead of listening to and addressing the concerns of racialised peoples, especially of affected peoples – appeals from the Racial Justice Task Force, various young greens of colour, and myself were ignored and disregarded over the concerns of the German Young Greens.

Instead of amplifying the voices of the racialised and affected peoples and strongly standing in solidarity against all forms of colonialism, imperialism, and oppression – the EC chose to focus first on finding ways to not have a statement to prevent upsetting the German Young Greens, and then to repeatedly water down the statement until it would appease them, before allowing their exclusion from it altogether.

The procedure with which we have handled the statement on Gaza has led to the gross reproduction of power imbalances and unfair favour towards the German Young Greens – a double standard that we yet again only apply to them and not to any of our other MOs. One that reproduces the very same power imbalances in Europe we supposedly criticise. Despite the overwhelming consensus of all our MOs on acknowledging the past and ongoing oppression of Palestinians by Israel, we have de facto given the German Young Greens a veto on human rights, but also on various other issues in the past.

Our Federation is built on dialogue and functions on democratic majorities. We have managed to pursue strong stances and resolutions no matter how difficult or contentious, guided by our shared principles. To repeatedly and only allow one MO to disengage from these processes that shape the Federation, is to go against the spirit of this Federation. If the values of the German Young Greens do not align with these principles of justice, equality, inclusion, and respect for human rights at the core of this Federation and its MOs, it is not us that must compromise for them – it is them who must reform or leave this Federation. Ultimately and crucially, human rights are uncompromisable and non-negotiable: there can be no veto.

As it is, as a Federation that claims to be founded upon principles of equality and respect for human rights and dignity; that claims to stand for fairness, especially towards all our MOs; that claims to advocate for decoloniality, justice and the platforming of marginalised voices; that claims to fight for system change; we have failed in all these regards.

I stand here yet again feeling unheard, ignored, if not silenced in this Federation. I must thank my fellow EC member Benedetta Scuderi and the MOs that believe in me for their support through this, but much work must be done collectively to ensure inclusion, racial justice, and decoloniality are not simply buzzwords for us to campaign on. If we truly are antiracist, we must start internally. A politics of principle. At this time, I do not see that being possible within this EC or the Federation in its current state.

Effective immediately I will be distancing myself from the EC for the foreseeable future whilst still carrying out all my responsibilities to the fullest degree as we undergo external mediation, and until comprehensive structural changes are made into how we make decisions to ensure our core principles and the voices of affected peoples are truly platformed over unfair political interests and the reproduction of power imbalances. I urge the EC to pursue these changes or to otherwise make use of article 3.5.1. of the IRPs to dismiss me.

I will not be stepping down, but instead using my voice through the mandate you elected me upon to keep fighting for radical racial justice and inclusion in this Federation. Nothing about us, without us.

For you, for us, for FYEG,

Srishagon Abraham.

Solidarity with Palestine, now and forever.

TW: genocide, killings, dehumanisation, violence

I’ve been struggling to find the words to truly encapsulate my thoughts and feelings in a way that genuinely reflects them, but that is also worth adding to the mass right now regarding the unfolding crisis in Gaza.

Why must Palestinians and those in support of Palestine now condemn the actions of Hamas,

while the Israeli Defence Force has been committing these very acts for decades?

Why must Palestinians and those in support of Palestine now condemn the dehumanisation and desecration of Israelis,

while the Israeli government calls for the ‘extermination’ and ‘annihilation’ of ‘human animals’?

Why must Palestinians and those in support of Palestine justify and qualify their right to resist,

while Israel’s right to defend its security includes the ongoing carpet bombing of civilian residences and infrastructure, the use of white phosphorous shells against civilians, the cutting off of water, food, electricity, and fuel, and the bombing of the singular path out of Gaza?

Why must Palestinians be pushed towards ‘peace talks’ and an unequal compromise in order to exist,

while the side with the power and capability to ensure a peaceful resolution refuses to do so, and has continually suppressed and killed Palestinians in peaceful resistance?

Why must we now call for the application of the Western international legal order for the breaches of human rights and war crimes by Hamas,

while Israel has refused to be bound by or held accountable under international law for its gross violations of the most fundamental human rights and peremptory norms?

Why do we not care about Palestinian suffering and Palestinian lives?

Any and all deaths are horrible. The killing of civilians, the desecration of lives and human dignity, of men, women, children by Hamas and the Israeli Defence Force – it is deplorable and unacceptable. But those critical of the historical and present violence by Israel against the Palestinian people, those standing with the Palestinian people, and the voices of the oppressed – the Palestinian people themselves – should not have to denounce Hamas every single time in order to justify their calls for solidarity with Palestine. It is yet again a vile double standard we apply to the oppressed but not to the oppressor.

The overwhelming majority of Europe and the Global North raise the incomprehensible ‘right of Israel to defend its security’ yet have continually closed a blind eye to, if not enabled, the decades of persecution, subjugation, oppression, and subordination of Palestine and the Palestinian people by Israel.

The unfolding crisis is not a conflict between two parties. It is indefensible to term this a conflict when one party is an imperialistic global power and occupier, and the other an oppressed and occupied de facto stateless peoples. It is the culmination of decades of oppression of Palestine by Israel – a colonial, apartheid regime engaged in aggressive settler-colonialism and decades of desecration, torture, and mass murder of Palestinian civilians. Decades of oppression that has increasingly pushed a minority of Palestinians towards militant extremism as their resistance.

While we stress that this minority of Palestinians, mobilising through Hamas, PFLP, DFLP, Lions’ Den and other militant groups, do not represent the Palestinian people, this distinction falls on deaf ears to the Israeli government. The militant resistance has brought with it a senseless and disproportionate retaliation by Israel towards all of the people in Gaza that is now leading towards the settler-colonial state-sanctioned genocide of Palestinians, funded and supported by the Global North, and forcing the unthinkable question each day:

Will Gazans wake up to another tomorrow or be forced to the remnants of history?

Yet again, as it always is, ‘human rights for me, not for thee’ goes the Global North modus operandi. The core preaches of its values of human rights, of human dignity, of democracy, yet watches with glee as the periphery perishes. The liberal Western legal order is but a façade for an oppressive, racist, imperialist core.

A status quo that remains unchallenged by even the most progressive politicians, activists, and civil society in Europe. After years of alleged prioritisation of intersectionality, racial justice, and antiracism; of anti-imperialism, decoloniality, and Global South solidarity; of platforming voices and listening to the marginalised; of unpacking and addressing historical violence and injustice; of system change – has this all merely been performative?

Palestinian history is erased; Palestinian voices are ignored or silenced; Palestinian liberation is conditioned or derided; Palestinian livelihood, environments, and communities are destroyed; and Palestinian lives are terminated.

There are many vocal and steadfast anti-Zionist Israelis and progressive Western civil society standing up against the roots of this crisis, we applaud and need their voices too, yet they remain but a minority. To truly strive for a free Palestine, for a solidary global decolonial project, the Global North needs to look inwards and challenge its values and beliefs to truly speak truth to power.

So to the Israelis supporting the apartheid regime or choosing to remain silent on Zionism and on the injustices committed by the Israeli state and its apparatus,

to the Western politician stressing Israel’s ‘right to defend its security’ and fuelling Israel’s murderous retaliation,

to the liberal activist’s reductive ‘both sides’ argument equating Palestine with Israel and calling for ‘ceasefires’ and ‘peace talks’ while Israel continually disregards the agreements made in them,

I ask again,

Why do we not care about Palestinian suffering and Palestinian lives?

From the river to the sea, now and forever, I hope Palestine to be free.