Spiral of Silence in Large Language Model Agents

2025-09-28

AI Summary

This study investigates whether large language models (LLMs), when deployed as autonomous agent populations, spontaneously exhibit “spiral of silence”-like opinion convergence under purely statistical generative mechanisms. Using controlled experiments, we systematically manipulate two classes of signals—historical dialogue context and role-based identity—and analyze opinion dynamics via Mann-Kendall trend tests, Spearman rank correlation, kurtosis, and interquartile range. Results demonstrate that statistically significant majority dominance and spiral-of-silence patterns emerge *only* when historical and role signals jointly co-occur; isolated manipulation of either signal leads instead to strong anchoring or opinion dispersion. This work challenges the conventional Spiral of Silence (SoS) theory—which presumes human-specific psychological mechanisms—by establishing that social consensus formation can arise computationally from purely statistical model interactions. It thus introduces a novel paradigm for modeling collective AI behavior grounded in measurable, mechanistic social dynamics.

Technology Category

• Cognitive Modeling & Cognitive Systems: Social Cognition And Interaction
• Multiagent Systems: Mechanism Design
• Machine Learning: Large Multimodal Models (LMMs)

Application Category

• Social Networks and Social Media: Generative AI / large language models and their impact on social systems
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation

Abstract

The Spiral of Silence (SoS) theory holds that individuals with minority views often refrain from speaking out for fear of social isolation, enabling majority positions to dominate public discourse. When the 'agents' are large language models (LLMs), however, the classical psychological explanation is not directly applicable, since SoS was developed for human societies. This raises a central question: can SoS-like dynamics nevertheless emerge from purely statistical language generation in LLM collectives? We propose an evaluation framework for examining SoS in LLM agents. Specifically, we consider four controlled conditions that systematically vary the availability of 'History' and 'Persona' signals. Opinion dynamics are assessed using trend tests such as Mann-Kendall and Spearman's rank, along with concentration measures including kurtosis and interquartile range. Experiments across open-source and closed-source models show that history and persona together produce strong majority dominance and replicate SoS patterns; history signals alone induce strong anchoring; and persona signals alone foster diverse but uncorrelated opinions, indicating that without historical anchoring, SoS dynamics cannot emerge. The work bridges computational sociology and responsible AI design, highlighting the need to monitor and mitigate emergent conformity in LLM-agent systems.

Problem

• Investigating Spiral of Silence dynamics in LLM agent collectives
• Evaluating how history and persona signals affect opinion formation
• Assessing emergent conformity patterns in large language model systems

Innovation

• Evaluates SoS dynamics using history and persona signals
• Assesses opinion trends with statistical trend tests
• Identifies conditions for majority dominance in LLMs