Nima Veiseh & Matthias J. Becker
June 2025
This research is supported by the Laterman Family Foundation AddressHate Initiative, with contributions from Sarah Valente, Bruze Tizes, Julian Cohen.
This white paper explores the challenge of addressing digital antisemitism within the broader ecology of online hate. Rather than privileging technology alone, it argues that any meaningful response must begin with historically informed, interdisciplinary analysis—drawing on the humanities, social sciences, and discourse studies to interpret how antisemitic and hateful narratives are embedded, coded, and circulated online. From this foundation, the paper proposes a dialogue with technical disciplines, exploring how AI, graph-based modeling, and lexicon-informed detection tools can contribute—if carefully guided by contextual knowledge, interpretive frameworks and responsible engineering management.
Focusing on antisemitism as a historically dense and culturally resonant discourse, the paper underscores the need for nuanced classification beyond simple keyword detection or taxonomic models. It advocates for the integration of qualitative analysis, multimodal annotation, and culturally aware interpretation into any scalable mitigation strategy. The paper also highlights international policy comparisons and system design challenges, arguing that cross-border collaboration, open-source transparency, and public education must work in tandem to confront the shifting dynamics of hate in the AI era. It calls for structural partnerships between researchers, technologists, executives, educators, and policymakers to create both preventative and responsive tools capable of keeping pace with this evolving threat.
Keywords: antisemitism, digital hate (speech), social media, online extremism, interdisciplinary methodology, (critical) discourse analysis, multimodality, machine learning, LLMs, AI and AI ethics, content moderation, platform governance and hate speech policy, education, international policy.
An Interdisciplinary Tech-Driven Framework to Understand and Counter Online Antisemitism and Hate
This paper argues that confronting digital antisemitism—and the broader landscape of online hate—requires not only technological innovation, but its integration with interdisciplinary expertise. Advances in artificial intelligence, graph-based modeling, and linguistic classification offer powerful tools, but their effectiveness depends on how well they are informed by history, discourse analysis, cultural studies, and the social sciences. Hate is not merely a pattern to be detected or a signal to be flagged; it is a communicative and ideological practice, shaped by centuries of narrative tradition, political rhetoric, and shifting power relations. Any sustainable strategy must begin from this interpretive foundation.
Antisemitism provides a historically unique and analytically indispensable case study in this regard. As one of the world’s most persistent and adaptive forms of hate, it often serves as a barometer of broader political extremism and social fragmentation—a “canary in the coal mine” of societal breakdown. Its online resurgence is rarely spontaneous. Rather, it is often catalyzed by real-world flashpoints and rapidly mutates across platforms in response to unfolding events. The wave of antisemitic content following the October 7, 2023, Hamas attacks and the subsequent war in Gaza is a stark example of how digital hate surges in response to political crises (Becker et al., 2024).
Moreover, unlike many forms of decentralized online hate, antisemitism often benefits from deliberate funding structures and networked dissemination. From state-sponsored influence campaigns to ideologically driven soft-power strategies, significant resources have been invested in sustaining and scaling antisemitic narratives across digital ecosystems. These structural dynamics must be accounted for when designing any technical or policy-oriented response.
Moreover, unlike many forms of decentralized online hate, antisemitism often benefits from deliberate funding structures and networked dissemination (Networks of Hate: Qatari Paymasters, Soft Power and the Manipulation of Democracy, 2023). From state-sponsored influence campaigns to ideologically driven soft-power strategies, significant resources have been invested in sustaining and scaling antisemitic narratives across digital ecosystems. These structural dynamics must be accounted for when designing any technical or policy-oriented response.
While some aspects of online hate may appear “codified”—detectable through recurring linguistic forms or visual tropes—these patterns do not exist in isolation. Their meaning is always shaped by context, platform norms, and historical discourse. The goal, then, is not simply to “break the code,” but to understand how these codes operate socially, semiotically, and politically. This paper proposes a framework for doing so: not only through a tech-first approach, but through a rigorous, interdisciplinary model in which technology becomes an instrument of insight, not a substitute for it.
Antisemitism, in its complexity and intersection with other hate ideologies—racism, misogyny, conspiracy theories, anti-democratic resentment—offers a critical foundation for building broader detection, education, and policy systems. By starting from this case, we aim to develop methodologies that are not only scalable, but historically and discursively responsible. Thus, successfully countering antisemitism would provide a blueprint for addressing hate speech broadly.
We answer five core questions in this white paper:
Where does hate start?
What are the historical, cultural, and psychological roots of hate, particularly antisemitism?
How is hate shaped—verbally, visually, and discursively?
How do language, imagery, and rhetorical strategies encode and adapt hate in the digital age?