Persuasion Propagation in LLM Agents

2026-01-31

AI Summary

This study investigates how user persuasion, operationalized through belief interventions, influences the behavior of large language model (LLM) agents in long-horizon tasks. Introducing the novel concept of “persuasion propagation,” the work proposes a behavior-centric evaluation framework that distinguishes between belief interventions applied before versus during task execution and establishes a causal paradigm for assessing belief-to-behavior effects. Leveraging belief preloading mechanisms, behavioral logging, and controlled experiments in web research and programming tasks, the study demonstrates that pre-task belief priming significantly alters agent behavior, reducing the average number of search queries by 26.9% and the number of unique information sources accessed by 16.9%. These findings substantiate a strong propagation effect of early-stage persuasion on downstream agent actions.

Technology Category

• Humans and AI: Human-Aware Planning and Behavior Prediction
• Multiagent Systems: Adversarial Agents
• Cognitive Modeling & Cognitive Systems: Simulating Human Behavior

Application Category

• 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
• Responsible Web: Human-perceived consequences of algorithmic deployment on the web

Abstract

Modern AI agents increasingly combine conversational interaction with autonomous task execution, such as coding and web research, raising a natural question: what happens when an agent engaged in long-horizon tasks is subjected to user persuasion? We study how belief-level intervention can influence downstream task behavior, a phenomenon we name \emph{persuasion propagation}. We introduce a behavior-centered evaluation framework that distinguishes between persuasion applied during or prior to task execution. Across web research and coding tasks, we find that on-the-fly persuasion induces weak and inconsistent behavioral effects. In contrast, when the belief state is explicitly specified at task time, belief-prefilled agents conduct on average 26.9\% fewer searches and visit 16.9\% fewer unique sources than neutral-prefilled agents. These results suggest that persuasion, even in prior interaction, can affect the agent's behavior, motivating behavior-level evaluation in agentic systems.

Problem

• persuasion propagation
• LLM agents
• belief intervention
• task behavior
• agentic systems

Innovation

• persuasion propagation
• belief intervention
• behavior-centered evaluation
• LLM agents
• task execution