Emotionally Charged, Logically Blurred: AI-driven Emotional Framing Impairs Human Fallacy Detection

2025-10-09

AI Summary

This study investigates how AI-generated emotional framing impairs human detection of logical fallacies. Using eight large language models, we injected emotionally charged language into logically identical fallacious arguments while preserving their underlying logical structure; the optimal model was selected to generate experimental stimuli, followed by rigorous human behavioral experiments. We systematically uncover, for the first time, the interaction mechanism between emotional framing and logical fallacies: emotional wording significantly reduces fallacy detection accuracy (mean F1 score decline of 14.5%) while increasing subjective persuasiveness, with heterogeneous effects across emotion types. Our contributions are twofold: (1) establishing emotional contagion as a critical enhancer of fallacy concealment, and (2) providing empirical evidence and methodological foundations for mitigating cognitive biases in AI-mediated public communication.

Technology Category

• Humans and AI: Emotional Intelligence
• Natural Language Processing: Discourse, Pragmatics & Argument Mining
• Philosophy and Ethics of AI: Safety, Robustness & Trustworthiness

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
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Logical fallacies are common in public communication and can mislead audiences; fallacious arguments may still appear convincing despite lacking soundness, because convincingness is inherently subjective. We present the first computational study of how emotional framing interacts with fallacies and convincingness, using large language models (LLMs) to systematically change emotional appeals in fallacious arguments. We benchmark eight LLMs on injecting emotional appeal into fallacious arguments while preserving their logical structures, then use the best models to generate stimuli for a human study. Our results show that LLM-driven emotional framing reduces human fallacy detection in F1 by 14.5% on average. Humans perform better in fallacy detection when perceiving enjoyment than fear or sadness, and these three emotions also correlate with significantly higher convincingness compared to neutral or other emotion states. Our work has implications for AI-driven emotional manipulation in the context of fallacious argumentation.

Problem

• AI emotional framing reduces human fallacy detection accuracy
• Emotional appeals in arguments increase perceived convincingness
• Different emotions affect human logical reasoning capabilities differently

Innovation

• LLMs inject emotional appeal into fallacious arguments
• Emotional framing reduces human fallacy detection performance
• Human detection varies with enjoyment, fear, and sadness emotions