This research introduces a novel simulation framework that uses LLM-based agents to model and analyze rumor spreading dynamics across different social network structures.

• Creates diverse LLM agent types with varying behaviors to realistically simulate rumor propagation
• Tests rumor spread across four distinct network structures to identify vulnerable configurations
• Evaluates effectiveness of different network designs in limiting misinformation spread
• Provides a testbed for developing more resilient social platforms against rumor campaigns

For security professionals, this research offers valuable insights into how misinformation propagates through networks, enabling better detection systems and containment strategies for digital threats before they reach critical mass.

Simulating Rumor Spreading in Social Networks using LLM Agents