Large Language Models (LLMs) are increasingly deployed to curate and rank human-created content, yet the nature and structure of their biases in these tasks remains poorly understood: which biases are robust across providers and platforms, and which can be mitigated through prompt design. We present a controlled simulation study mapping content selection biases across three major LLM providers (OpenAI, Anthropic, Google) on real social media datasets from Twitter/X, Bluesky, and Reddit, using six prompting strategies (\textit{general}, \textit{popular}, \textit{engaging}, \textit{informative}, \textit{controversial}, \textit{neutral}). Through 540,000 simulated top-10 selections from pools of 100 posts across 54 experimental conditions, we find that biases differ substantially in how structural and how prompt-sensitive they are. Polarization is amplified across all configurations, toxicity handling shows a strong inversion between engagement- and information-focused prompts, and sentiment biases are predominantly negative. Provider comparisons reveal distinct trade-offs: GPT-4o Mini shows the most consistent behavior across prompts; Claude and Gemini exhibit high adaptivity in toxicity handling; Gemini shows the strongest negative sentiment preference. On Twitter/X, where author demographics can be inferred from profile bios, political leaning bias is the clearest demographic signal: left-leaning authors are systematically over-represented despite right-leaning authors forming the pool plurality in the dataset, and this pattern largely persists across prompts.

翻译：大型语言模型（LLMs）正日益被用于策划和排序人类生成的内容，但在此类任务中，其偏差的性质与结构仍未被充分理解：哪些偏差在不同提供商和平台间具有稳健性，哪些可通过提示设计加以缓解。我们通过控制模拟研究，基于Twitter/X、Bluesky和Reddit的真实社交媒体数据集，对三家主要LLM提供商（OpenAI、Anthropic、Google）的内容选择偏差进行了映射，采用了六种提示策略（通用、流行、吸引、信息性、争议性、中立）。通过54种实验条件下从100条帖子池中进行的540,000次模拟前10选择，我们发现偏差在结构性和对提示的敏感性方面存在显著差异。极化在所有配置下均被放大，毒性处理在吸引导向与信息导向提示之间呈现强烈反转，情感偏差则主要为负面。提供商对比揭示了不同的权衡：GPT-4o Mini在跨提示行为上最为一致；Claude和Gemini在毒性处理上表现出高适应性；Gemini在负面情感偏好上最为显著。在Twitter/X上，当作者人口统计学信息可从个人资料中推断时，政治立场偏差是最清晰的人口统计信号：尽管数据集中右倾作者构成大多数，左倾作者却被系统性过度代表，且这一模式在提示中基本持续存在。