Are Large Language Models Sensitive to the Motives Behind Communication?

2025-10-22

AI Summary

This study investigates large language models’ (LLMs) ability to detect and rationally discount human communicative motives—particularly bias- or interest-driven testimony. Adopting a cognitive science–informed controlled experimental paradigm, we integrate rational learning modeling with naturalistic evaluation in sponsored advertising contexts to systematically assess LLMs’ motive-sensitive information discounting—a first-of-its-kind validation. We propose “intention salience enhancement,” a lightweight prompting intervention that heightens model awareness of source motives. Results demonstrate that LLMs exhibit human-like motive sensitivity—discounting biased testimony—but lack robustness in realistic settings. Crucially, minimal intention-directed prompts significantly improve their critical judgment accuracy. This work establishes a novel paradigm for modeling social cognition in LLMs and provides an extensible, prompt-based intervention framework for enhancing trustworthy human–AI interaction.

Technology Category

• Cognitive Modeling & Cognitive Systems: Social Cognition And Interaction
• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: Fact-Checking / Misinformation Detection (NLP Focus)

Application Category

• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• 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: Search Tool Learning with LLM: Teaching LLMs to invoke search and make use of retrieved information

Abstract

Human communication is motivated: people speak, write, and create content with a particular communicative intent in mind. As a result, information that large language models (LLMs) and AI agents process is inherently framed by humans' intentions and incentives. People are adept at navigating such nuanced information: we routinely identify benevolent or self-serving motives in order to decide what statements to trust. For LLMs to be effective in the real world, they too must critically evaluate content by factoring in the motivations of the source -- for instance, weighing the credibility of claims made in a sales pitch. In this paper, we undertake a comprehensive study of whether LLMs have this capacity for motivational vigilance. We first employ controlled experiments from cognitive science to verify that LLMs' behavior is consistent with rational models of learning from motivated testimony, and find they successfully discount information from biased sources in a human-like manner. We then extend our evaluation to sponsored online adverts, a more naturalistic reflection of LLM agents' information ecosystems. In these settings, we find that LLMs' inferences do not track the rational models' predictions nearly as closely -- partly due to additional information that distracts them from vigilance-relevant considerations. However, a simple steering intervention that boosts the salience of intentions and incentives substantially increases the correspondence between LLMs and the rational model. These results suggest that LLMs possess a basic sensitivity to the motivations of others, but generalizing to novel real-world settings will require further improvements to these models.

Problem

• Assessing LLMs' capacity for motivational vigilance in communication
• Evaluating how LLMs handle biased information from self-serving sources
• Testing LLMs' sensitivity to human intentions in real-world scenarios

Innovation

• Employ cognitive science experiments for LLM evaluation
• Use rational models to assess motivational bias detection
• Apply steering interventions to enhance intention sensitivity