This report is available in Arabic.

Over two years of research across X and Instagram, ISD Jordan has built an evidence-based picture of online hate speech in Jordan, documenting who is targeted, how harmful content spreads, what drives it and what needs to change. This report brings together an overview of key trends and dynamics of online hate speech in Jordan, summarising the findings of a research series, conducted between September 2024 and April 2026 as part of the EU-funded project “Strengthening Prevention and Response Mechanisms to Intersectional Online Hate Speech in Jordan,” implemented by ISD Jordan in partnership with I-DARE Sustainable Development and Saba Hamlet for Gender Equality.

The project was designed to address a set of clearly identified gaps: the absence of a nationally agreed definition of hate speech, policy frameworks, a lack of evidence-based research on hate speech in the Jordanian context, limited Arabic-language resources for prevention and response and institutional awareness of the nature and scale of the problem. Two years on, this research series represents a direct and substantive response to each of those gaps by providing an evidence-based national framework developed through a multi-stakeholder process and generating recommendations grounded in data.

The research had four core objectives: to identify and map key trends and narratives related to online hate speech across social media platforms; to examine spikes linked to political, social or cultural events; to assess which groups are targeted and how; and to generate actionable, data-driven recommendations for policymakers, platforms, civil society and media actors. Across each of these objectives, the research produced concrete findings.

As with other countries in the Middle East and North Africa (MENA) region, Jordan is heavily affected by events in its neighboring states, including conflicts, as well as political and economic instability. Regional trigger events -such as the ongoing Gaza conflict and the events in Sweida, Syria generated spikes of sectarian and nationalist content online. These conflicts profoundly shape Jordanian society across ideological, economic, political and social dimensions. The cumulative effect is that hate speech becomes not merely reactive but also embedded in how Jordanians process regional instability—reflecting anxieties about resources, sovereignty and survival that transcend the immediate conflict.

Across the dataset, ISD identified three persistent typologies of harmful content on Jordan’s social media platforms: gender-based abuse targeting women in public life, religious hate speech directed disproportionately at Shia communities and other minorities and exclusionary nationalist discourse. These are not peripheral phenomena. They are structural features of Jordan’s online environment, present across platforms, persistent across time and concentrated against specific communities.

These findings point to something important: Jordan’s online environment cannot be understood, or addressed, as a purely domestic problem. ISD’s research consistently documented how pan-Arab narratives, sectarian framings originating elsewhere in the Arab world, and content from regional influencers circulate freely on Jordanian social media platforms, where they are adapted and amplified within local contexts. The country’s information ecosystem is inseparable from broader regional flows. Any response that treats online hate speech in Jordan as a domestic issue alone will be incomplete.

ISD’s research also identified two structural conditions that make harmful content harder to detect, limit and address. The first is the Arabic-language moderation gap. Major platforms have systematically underinvested in Arabic-language tools, human review and accessible redress mechanisms. ISD found direct evidence that this gap is being actively exploited: Arabizi—Arabic written in Latin characters and numerals—is used as a deliberate and widespread workaround to evade automated detection. Users posting hateful content have adapted their behaviour to the known limitations of platform moderation systems, and those systems have not kept pace. The second structural condition is algorithmic amplification. Content that provokes strong emotional responses such as identity-polarising, divisive content targeted at specific communities, receives disproportionate reach on platforms optimised for engagement. ISD’s monitoring found this dynamic operating across all three typologies, with harmful narratives gaining visibility.