At the Zera Institute, our research examines how hate, extremism, and antisemitism spread through digital ecosystems—and how they erode the foundations of liberal democracy. We uncover what often goes unnoticed: the linguistic, visual, and algorithmic patterns through which radical ideologies become normalized, shared, and socially embedded.

A central pillar of our research is the collaboration with Dr. Matthias J. Becker, one of Europe’s leading experts on digital antisemitism and coded hate speech, who is affiliated with Zera as a research affiliate. Building on his methodology, developed over many years, our interdisciplinary team—spanning linguistics, political science, sociology, media analysis, and AI—develops advanced approaches to identifying and decoding antisemitic, conspiratorial, and extremist narratives across platforms such as YouTube, X, and TikTok. Their work goes far beyond keyword tracking, capturing irony, meme culture, multilingual nuance, and interaction dynamics in comment sections. This enables us to map not only how hate speech is expressed, but how it spreads, mutates, and incites.

What makes Zera unique is its synthesis of digital precision and cultural and historical depth. Prof. Benjamin Folit-Weinberg, Director of Research at Zera, is a classical philologist and expert on ancient Greek thought. His scholarship examines how language, metaphor, and narrative shape our understanding of the world around us. At Zera, he brings this philosophical and cultural lens to contemporary debates, enriching our understanding of the values—liberal democracy, pluralism, tolerance—that are increasingly under threat. His insights help frame the digital dynamics we observe within a broader intellectual and cultural context.

Liberal democracy is struggling, with violence, especially political violence and violence against minorities, growing at an alarming rate, while trust in institutions plummets. Online extremism and radicalization are major drivers of this trend. And antisemitism is a crucial component of online radicalization and extremism.

To date, the interaction between these three phenomena – threats to liberal democracy, online extremism, and antisemitism – has received limited systematic analysis, particularly in influencer-centered social media environments. Our primary report addresses this gap through an empirical study of how influencers and their audiences jointly produce and intensify illiberal discourse. The analysis focuses on three interrelated processes that pose acute challenges to contemporary liberal democracy: the normalization of political violence; the legitimation of violence against minorities, especially Jewish people; and the progressive delegitimization of democratic institutions.

The research project underlying the following reports was completed at the end of January 2026.

Primary report at a glance:

•  Analyzes more than 11,000 user comments from Twitter/X, Instagram, YouTube, TikTok, and other social media platforms
• Undertakes a cross-national comparative study of over two dozen German and American influencer accounts and their audiences, spanning the political spectrum
•  Builds on the long-standing methodology of the Decoding Antisemitism project, led by Dr. Matthias J. Becker
• Sheds light on the role played by online discourse in normalizing violence, especially political violence and violence against minorities, and the erosion of trust in institutions
• Provides insights into context-specific manifestations of illiberalism across German and American digital ecosystems
• Takes foundational steps, translating qualitative findings into scalable, AI-assisted analytical prototypes

Second report

Our second report builds on the empirical findings of our first study by formulating hypotheses for future research and identifying four possible policy-level interventions and policy-relevant intervention pathways. We pay particular attention to the role of comment sections in amplifying, radicalizing, and legitimizing discourse beyond the initial framing provided by influencers.

• Social Media Policy & Regulation: Evidence-based criteria for platform governance; identification of amplification dynamics warranting intervention
• Security/National Security & Risk Assessment: Early-warning system indicators; improved detection of coordinated or foreign influence operations
• Education & Prevention: Empirically grounded insights into how radical frames normalize violence and foundations for counter-radicalization and resilience-oriented educational content
• Applied AI Development: Translation of qualitative discourse insights into monitoring and diagnostic systems, with an emphasis on prevention rather than purely reactive enforcement