This piece is based on a Lightning Talk titled “Digital freedoms censorship” at the 2024 APC Community Gathering.
On 27 October 2023, the United Nations Committee on the Elimination of Racial Discrimination expressed grave concern about “the sharp increase in racist hate speech and dehumanization directed at Palestinians since October 7, particularly on the Internet and social media.” This concern arises amidst the widespread dissemination of violent and inflammatory content in Hebrew against Palestinians across social media platforms, particularly after October 2023.
The surge in violent Hebrew content on social media has been substantial, though systematic violent content and hate speech against Palestinians predates October. From January to September 2023, 7amleh detected approximately 7 million instances of violent Hebrew content across various social media platforms, including Facebook which accounted for 24.57% of the violent content online.
Facebook, operated by Meta, is one of the primary platforms where this violent content proliferates. Despite the vast spread of such content, Meta has yet to implement substantial measures or policy changes to address the rampant online violence against Palestinians, especially during the ongoing genocidal war on Gaza and its people.
Persistent failures: Meta’s inadequate Hebrew content moderation
In its report, titled Human Rights Due Diligence of Meta’s Impacts in Israel and Palestine in May 2021, Business for Social Responsibility (BSR) concluded that Meta lacked a functional “Hebrew hostile speech classifier” despite having one for the Arabic language. As a result, Hebrew content was subject to under-enforcement of policies. Among other recommendations, the report advised Meta to “continue work on having functioning Hebrew classifiers”.
In a subsequent update in September 2023, Meta announced that it had adopted BSR’s recommendations and updated the Hebrew classifier, which was initially launched in 2022. These updates were intended to enhance the detection and moderation of harmful content, particularly violent speech in Hebrew. However, our research and documentation at 7amleh tells a different story. Despite Meta’s claims, there has been a significant surge in violent Hebrew content targeting Palestinians detected by 7amleh’s Violence Indicator since October 2023. This indicator has identified approximately 7 million pieces of such content, 20.8% of them on Facebook platform, suggesting a severe ongoing issue with Meta’s content moderation capabilities.
At the same time, Meta lowered its trust threshold from 80% to just 25% for Palestinian/Arabic content. Meaning that any content that has a chance of 25% of being violative of community standards is removed immediately. This has led to the over-moderation of Palestinian content while allowing a lot of Hebrew violative content on its platforms.
Not only have organisations and average individuals engaged in violent rhetoric, but Israeli officials and policymakers have also been involved such activity. Since the commencement of investigative efforts on 7 October 2023, Law for Palestine has meticulously gathered a compelling body of evidence exposing the widespread incitement to violence and genocidal intent perpetrated by Israeli officials and public figures against Palestinians. 7amleh’s documentation includes over 500 instances of violent and genocidal incitement appearing in social media posts, television interviews and official statements from Israeli politicians, military personnel, journalists and other influential figures.
This is also was noted by the International Court of Justice. In its ruling, the ICJ, directed Israel, in response to the genocide case brought by South Africa, to “take all measures within its power to prevent and punish the direct and public incitement to commit genocide in relation to members of the Palestinian group in the Gaza Strip.” The specific mention of incitement by public figures against Gaza, among other measures ordered, underscores the grave nature of the violent digital content targeting Palestinians.
Israel, as a state, is complicit by failing to take action or hold perpetrators accountable for the violent content against Palestinians. As an occupying power, it has a mandate to protect all people under its control and to hold its own citizens accountable for digital crimes.
The Israeli government is not only sponsoring violent content online but also supporting the spread of misinformation against Palestinians in the USA and other Western countries. Over recent months, organisations have identified possible Israeli government-sponsored influence activities related to the Gaza war. In January, the Israeli newspaper Haaretz reported that the Israeli government had purchased technology for online influence campaigns. By February, researchers and DFRLab uncovered fake social media accounts attacking UN staff working with Palestinian refugees, and targeting Black Democratic members of Congress. In March, DFRLab found another network spreading anti-Muslim narratives in Canada.
Reports from Meta and OpenAI confirmed that STOIC, an Israel-based company, used AI tools to manipulate political narratives online regarding Palestine and Israel. By leveraging ChatGPT, STOIC generated misleading content, which was then disseminated by hundreds of fake accounts on Meta’s platforms, aiming to misinform and bias audiences against Palestinians.
Silencing Palestinian voices and fuelling violence in Gaza
The spread of violent and defamatory content online against Palestinians has a chilling effect on freedom of expression and speech, undermines access to information, and is correlated with the genocidal war on Gaza.
In a report published by Adalah, data reveals that since the Gaza war began in October 2023, Palestinian students in Israel have faced severe disciplinary actions that target their freedom of speech and Palestinian identity. Right-wing campus groups and academic institutions have scrutinised students’ social media posts expressing solidarity with Gaza or featuring Quranic verses, labelling them as “support for terrorism”. As a result, at least 36 universities have taken action against 124 Palestinian students, with Adalah representing 95 of them. Nearly half of these cases have led to suspensions or expulsions, disproportionately affecting female students (79%). These actions, endorsed by the Israeli minister of education, stand in stark contrast to the lack of repercussions for Jewish Israeli students who post explicit incitement.
Adalah also published a report detailing the implications of smear and inflammatory campaigns against Palestinian citizens of Israel. The report outlined the repressive measures taken by Israeli authorities, companies and academic institutions against Palestinians expressing sympathy for the bloodshed in Gaza. These measures included the arrest of hundreds of individuals, demotions at work, and termination of employment.
7amleh has demonstrated in previous cases the correlation between the spread of violent online content and atrocities committed on the ground. In a report on the violent digital content targeting Huwara town in occupied West Bank in early 2023, 7amleh showed that the inflammatory online campaign culminated in a large-scale offline attack, resulting in the killing of a Palestinian, injuries to others, burning of hundreds of cars, and extensive damage to properties and crops. This same fertile environment for online violence persists during the ongoing war on Gaza, but at a significantly greater magnitude.
Meta is allowing more genocidal rhetoric in against Gaza than ever. Human rights defenders and organisations are drawing parallels between Meta’s role in the Myanmar genocide and the current atrocities in Gaza, highlighting the platform’s tolerance of violent rhetoric against Palestinians. By prioritising profit over the safety and lives of populations, Meta is perpetuating harm on an alarming scale instead of learning from past experiences.
The implications of unchecked online violence
The surge in violent Hebrew content against Palestinians on social media platforms, particularly since October 2023, is a profound concern highlighted by the United Nations Committee on the Elimination of Racial Discrimination. Despite Meta’s claims of improved content moderation systems, the widespread dissemination of hate speech and incitement to violence continues unchecked. This persistent issue is exacerbated by Israeli government-sponsored influence campaigns and the lack of accountability for the violent content shared by Israeli citizens and officials.
The implications of this unchecked online violence are severe. It not only undermines freedom of expression and access to information but also correlates directly with the physical violence experienced against Palestinians.
In conclusion, the proliferation of violent and smearing content against Palestinians on social media reflects and fuels the broader context of genocidal violence and systemic discrimination. It underscores the urgent need for robust and effective content moderation policies, accountability for perpetrators of online hate, and a commitment to protecting the rights and freedoms of all individuals, especially those under occupation. Addressing these issues is critical to mitigating harm and fostering a more just and equitable digital and real-world environment.
Image: Emergency Rally for Palestine by Matt Hrkac via Flickr (CC BY 2.0 DEED)
Ahmad Qadi is a dedicated human rights defender, researcher and activist. He currently holds the position of monitoring and documentation manager at 7amleh – The Arab Center for the Advancement of Social Media, where he is responsible for documenting digital rights violations and ensuring the protection of online users from digital threats and censorship.