How hatred spreads online: Revealing the hidden networks amplifying polarization in Italian politics

What is it about?

This research examines how hatred flows and spreads across social media platforms, specifically studying the case of Mario Draghi's resignation as Italian Prime Minister in July 2022. The study analyzed nearly 800,000 Italian-language messages on X (formerly Twitter) to understand how organized groups of users coordinate their actions to intensify political polarization through hateful content. The researchers discovered six major clusters of users, each aligned with different political movements—from Giorgia Meloni's far-right Brothers of Italy party to the Five Star Movement and other political factions. What makes this study particularly striking is that it identified a strategic pattern: accounts were deliberately coordinating to amplify hateful messages despite their moderate intensity, using what researchers call astroturfing—a tactic that creates the illusion of grassroots movements while actually being orchestrated by organized groups.

Featured Image

Photo by Arnaud Jaegers on Unsplash

Photo by Arnaud Jaegers on Unsplash

Why is it important?

This research is vital for understanding how democracies are under threat not primarily from overt violence, but from invisible manipulation of public opinion. The findings have several critical implications: For democracy and civil society: The study demonstrates that political actors strategically exploit social media platforms to delegitimize opponents and dehumanize social groups (migrants, minorities, political adversaries) rather than engaging in substantive policy debate. When hateful messaging becomes normalized in political discourse, it erodes trust in democratic institutions and weakens social cohesion. For understanding modern polarization: Rather than organic disagreement evolving naturally, the research shows that much digital polarization is deliberately engineered. Small networks of coordinated accounts—often with minimal followers—prove surprisingly effective at shaping what millions of people see and believe. This challenges the assumption that social media democratizes information; instead, it reveals how concentrated influence can be achieved through strategic coordination. For combating disinformation and hate: The study shows that effective hate speech often isn't obvious. It employs subtle emotional narratives, loaded language about national responsibility, and attacks on political figures rather than crude slurs. This makes it harder for automated moderation systems to detect, meaning much hateful content remains on platforms despite moderation efforts. For addressing the transnational dimension of influence: The presence of coordinated accounts from countries like Thailand and Albania suggests sophisticated, cross-border operations. This indicates that threats to national democracies aren't merely domestic—they involve international actors with geopolitical interests. For platform accountability and regulation: The research exposes gaps in how social media platforms moderate content and monitor coordinated behavior. With 80% of problematic messages remaining on X despite moderation policies, the study argues for enhanced detection tools that focus not just on individual hateful messages but on patterns of coordinated behavior that signal organized manipulation campaigns.

Perspectives