刻板印象还是个性化？用户身份偏见影响聊天机器人推荐
Stereotype or Personalization? User Identity Biases Chatbot Recommendations
AI总结：
研究发现，LLMs 在生成推荐时会因用户身份特征产生种族刻板印象偏见，且未能透明披露该影响，此现象在多个主流模型和种族群体中普遍存在。
摘要：
尽管用户通常希望获得个性化推荐，但在实际操作中很难区分偏见与个性化：我们发现，无论用户是通过明确指示有意透露身份，还是通过隐含线索无意中暴露身份，模型都会生成带有种族刻板印象的推荐。我们证明，当人们使用大型语言模型（LLMs）生成推荐时，LLMs 产生的回复既反映了用户想要什么，也反映了用户是谁。我们认为，聊天机器人应当透明地表明推荐何时受到用户透露的身份特征的影响，但观察到它们目前未能做到这一点。我们的实验表明，尽管用户透露的身份显著影响模型推荐（p < 0.001），但模型回复在响应用户查询时掩盖了这一事实。这种偏见和缺乏透明度的现象在多个流行的消费级 LLMs 和四个美国种族群体中持续存在。
英文摘要：
While personalized recommendations are often desired by users, it can be difficult in practice to distinguish cases of bias from cases of personalization: we find that models generate racially stereotypical recommendations regardless of whether the user revealed their identity intentionally through explicit indications or unintentionally through implicit cues. We demonstrate that when people use large language models (LLMs) to generate recommendations, the LLMs produce responses that reflect both what the user wants and who the user is. We argue that chatbots ought to transparently indicate when recommendations are influenced by a user's revealed identity characteristics, but observe that they currently fail to do so. Our experiments show that even though a user's revealed identity significantly influences model recommendations (p < 0.001), model responses obfuscate this fact in response to user queries. This bias and lack of transparency occurs consistently across multiple popular consumer LLMs and for four American racial groups.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
