Mind What You Ask For: Emotional and Rational Faces of Persuasion by Large Language Models

2025-02-13

AI Summary

Large language models (LLMs) increasingly mediate public discourse, yet their susceptibility to deploying emotionally charged rather than evidence-based persuasive strategies poses significant risks for AI-driven collective misinformation. Method: We conduct a multidisciplinary investigation integrating psycholinguistic analysis, social influence theory modeling, and cross-model prompting experiments across 12 state-of-the-art LLMs to systematically deconstruct pragmatic features and persuasion pathways in persuasive responses. Contribution/Results: We introduce the first human-centered, interdisciplinary risk assessment framework for AI persuasion. Results reveal a pervasive emotional bias in LLM-generated persuasive content; critically, rationality-oriented prompting significantly enhances response credibility and resistance to misinformation (p < 0.01). This work provides empirically grounded, actionable interventions for explainable AI governance and fills a critical gap in the joint analysis of LLM persuasion mechanisms and societal impact.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: (Large) Language Models
• Humans and AI: Learning Human Values and Preferences

Application Category

• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• 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

Abstract

Be careful what you ask for, you just might get it. This saying fits with the way large language models (LLMs) are trained, which, instead of being rewarded for correctness, are increasingly rewarded for pleasing the recipient. So, they are increasingly effective at persuading us that their answers are valuable. But what tricks do they use in this persuasion? In this study, we examine what are the psycholinguistic features of the responses used by twelve different language models. By grouping response content according to rational or emotional prompts and exploring social influence principles employed by LLMs, we ask whether and how we can mitigate the risks of LLM-driven mass misinformation. We position this study within the broader discourse on human-centred AI, emphasizing the need for interdisciplinary approaches to mitigate cognitive and societal risks posed by persuasive AI responses.

Problem

• Analyze psycholinguistic features in LLM responses
• Explore rational vs. emotional persuasion tactics
• Mitigate risks of LLM-driven misinformation

Innovation

• Analyzes psycholinguistic features of LLMs
• Explores rational and emotional prompts
• Mitigates risks of AI-driven misinformation