Political Persuasion and Endorsement in Large Language Models

2026-06-04

AI Summary

This study addresses the susceptibility of large language models (LLMs) to persuasive language in politically sensitive contexts, highlighting their limited reliability in simulating human political cognition. It presents the first systematic investigation into how political persona prompts influence LLMs’ endorsement of persuasive content. The authors construct a dataset incorporating real-world media persuasion techniques and evaluate model responses using a five-point Likert scale, combined with political persona prompt engineering and comparative analysis across six regionally diverse LLMs. Results show that, without prompts, models generally reject persuasive messages; however, when prompted with left- or right-wing personas, their endorsement becomes significantly polarized. This polarization varies by persuasion strategy and issue domain, revealing substantial bias risks and context dependency in LLMs’ simulation of political judgment.

Abstract

Large Language Models (LLMs) are increasingly employed as proxies for human behavior in computational social science. However, their tendency to internalize biases from training data raises concerns about their reliability in politically sensitive domains, specifically in regard to their susceptibility to persuasive language. In this work, we examine whether LLMs endorse persuasion-infused messages and whether partisan persona prompting modulates such endorsement. We evaluate six LLMs from different geographic regions on content annotated with persuasion techniques drawn from real-world media sources, measuring the likelihood of endorsement using a five-point Likert scale. The models are prompted as either a neutral social media user or as a user with left- or right-leaning political views. Results show that without political conditioning, LLMs generally do not endorse messages containing persuasion techniques, though model-level differences emerge, and that partisan persona prompting increases polarization of endorsement, particularly for persuasion-infused content. Endorsement further varies by persuasion technique and topic. These findings raise concerns about agentic LLM deployments in politically sensitive environments and complicate their use as reliable simulators of human political cognition.

Problem

• Political Persuasion
• Large Language Models
• Endorsement
• Partisan Bias
• Computational Social Science

Innovation

• political persuasion
• partisan persona prompting
• large language models
• endorsement polarization
• computational social science