Framing the Narrative: Ideological Mimicry in Large Language Models

构建叙事：大型语言模型中的意识形态模仿

Abstract: Large language models (LLMs) are increasingly used to answer questions about politically contentious issues, yet evaluations typically treat a model’s stance as a relatively stable property. Real users, however, communicate political signals through their terminology, assumptions, and personal context.

摘要： 大型语言模型（LLMs）正越来越多地被用于回答有关政治争议性议题的问题，然而目前的评估通常将模型的立场视为一种相对稳定的属性。事实上，真实用户会通过其术语选择、预设前提和个人背景来传达政治信号。

We investigate whether such signals produce ideological mimicry: systematic shifts in the political stance expressed by an LLM toward the position conveyed by the interaction. If LLMs adapt their responses to these signals, they risk creating personalised political information environments in which users with opposing views receive systematically different accounts of the same issue, potentially reinforcing existing divisions.

我们研究了这些信号是否会产生“意识形态模仿”：即大型语言模型所表达的政治立场会根据交互中传达的观点发生系统性偏移。如果大型语言模型根据这些信号调整其回答，它们就有可能创造出个性化的政治信息环境，使得持有对立观点的用户对同一议题接收到系统性差异的解读，从而可能加剧现有的社会分歧。

We build the Poli-SHIFT dataset and evaluation framework and assess seven open-weight LLMs across ten contentious political topics in the United States, United Kingdom, and Australia, systematically manipulating contested terminology, politically valenced premises, and user information, and eliciting responses in both multiple-choice and open-text formats.

我们构建了 Poli-SHIFT 数据集和评估框架，并针对美国、英国和澳大利亚的十个政治争议话题，对七个开源权重的大型语言模型进行了评估。我们系统性地操纵了争议性术语、具有政治倾向的前提假设以及用户信息，并以多项选择和开放式文本两种格式获取了模型的回答。

Across models, we find robust evidence that prompt framing shapes the political stance of LLM outputs. Changing terminology alone reverses which side of an issue a model supports in 16.9% of matched comparisons. Stated political ideology also systematically shifts responses toward the user’s position.

在所有模型中，我们发现了强有力的证据表明，提示词的框架（prompt framing）会塑造大型语言模型输出的政治立场。仅改变术语一项，在 16.9% 的匹配对比中就足以逆转模型对某一议题的支持立场。此外，用户声明的政治意识形态也会系统性地将模型的回答向用户自身的立场偏移。

These findings show that political stance is not a fixed property of LLMs; the views expressed are conditional on the interaction with the user. As LLMs become increasingly personalised sources of information, such interaction-dependent adaptation could contribute to political information environments that reinforce users’ existing perspectives.

这些发现表明，政治立场并非大型语言模型的固定属性；其表达的观点取决于与用户的交互。随着大型语言模型日益成为个性化的信息来源，这种依赖于交互的适应性可能会导致政治信息环境进一步固化用户的既有观点。