Political Ideology Shifts in Large Language Models

TL;DR

This study examines how large language models' political expressions are influenced by size and persona content, revealing scale-dependent ideological shifts and susceptibility to priming, which have implications for fairness and safety.

Contribution

It provides a systematic analysis of ideological shifts in LLMs due to scale and persona content, using the Political Compass Test across multiple models.

Findings

Larger models show broader, more polarized ideological coverage.

Models are more susceptible to explicit ideological cues as they scale.

Priming with thematic content causes predictable ideological shifts that grow with size.

Abstract

Large language models (LLMs) are increasingly deployed in politically sensitive settings, raising concerns about their potential to encode, amplify, or be steered toward specific ideologies. We investigate how adopting synthetic personas influences ideological expression in LLMs across seven models (7B-70B+ parameters) from multiple families, using the Political Compass Test as a standardized probe. Our analysis reveals four consistent patterns: (i) larger models display broader and more polarized implicit ideological coverage; (ii) susceptibility to explicit ideological cues grows with scale; (iii) models respond more strongly to right-authoritarian than to left-libertarian priming; and (iv) thematic content in persona descriptions induces systematic and predictable ideological shifts, which amplify with size. These findings indicate that both scale and persona content shape LLM political behavior. As such systems enter decision-making, educational, and policy contexts, their latent ideological malleability demands attention to safeguard fairness, transparency, and safety.