PulseAugur
EN
LIVE 08:12:48

LLM agents simulate hate speech cascades on Bluesky

Researchers have developed a multi-agent large language model (LLM) system to simulate hate speech propagation on online platforms, aiming to improve moderation strategies. The study analyzed three hateful cascades and one benign cascade from Bluesky, finding that hateful cascades exhibit a high percentage of hostile reposters and a star-like diffusion topology. The LLM simulator successfully reproduced the observed stance monoculture and toxicity-engagement homophily, with agent heterogeneity identified as a key factor in simulation fidelity. Targeting amplifiers within dense networks showed a potential reduction in hateful content spread. AI

IMPACT This research could lead to more effective AI-driven moderation tools for combating online hate speech.

RANK_REASON The cluster contains an academic paper detailing a new simulation methodology for studying online phenomena. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →

COVERAGE [1]

  1. arXiv cs.AI TIER_1 English(EN) · Fan Huang ·

    Simulating Hate Speech Cascades with Multi-LLM Agents: Empirical Grounding, Modeling Fidelity, and Intervention Strategies

    arXiv:2606.18264v1 Announce Type: cross Abstract: Faithful modeling of hateful content propagation on online platforms remains an open problem for moderation research. Classical cascade models that do not explicitly represent the profile, community, and content factors associated…