Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Auditing Preferences for Brands and Cultures in LLMs

Jasmine Rienecker, Katarina Mpofu, Naman Goel, Siddhartha Datta, Jun Zhao, Oscar Danielsson, Fredrik Thorsen

Abstract

Large language models (LLMs) based AI systems increasingly mediate what billions of people see, choose and buy. This creates an urgent need to quantify the systemic risks of LLM-driven market intermediation, including its implications for market fairness, competition, and the diversity of information exposure. This paper introduces ChoiceEval, a reproducible framework for auditing preferences for brands and cultures in large language models (LLMs) under realistic usage conditions. ChoiceEval addresses two core technical challenges: (i) generating realistic, persona-diverse evaluation queries and (ii) converting free-form outputs into comparable choice sets and quantitative preference metrics. For a given topic (e.g. running shoes, hotel chains, travel destinations), the framework segments users into psychographic profiles (e.g., budget-conscious, wellness-focused, convenience), and then derives diverse prompts that reflect real-world advice-seeking and decision-making behaviour. LLM responses are converted into normalised top-k choice sets. Preference and geographic bias are then quantified using comparable metrics across topics and personas. Thus, ChoiceEval provides a scalable audit pipeline for researchers, platforms, and regulators, linking model behaviour to real-world economic outcomes. Applied to Gemini, GPT, and DeepSeek across 10 topics spanning commerce and culture and more than 2,000 questions, ChoiceEval reveals consistent preferences: U.S.-developed models Gemini and GPT show marked favouritism toward American entities, while China-developed DeepSeek exhibits more balanced yet still detectable geographic preferences. These patterns persist across user personas, suggesting systematic rather than incidental effects.

Auditing Preferences for Brands and Cultures in LLMs

Abstract

Large language models (LLMs) based AI systems increasingly mediate what billions of people see, choose and buy. This creates an urgent need to quantify the systemic risks of LLM-driven market intermediation, including its implications for market fairness, competition, and the diversity of information exposure. This paper introduces ChoiceEval, a reproducible framework for auditing preferences for brands and cultures in large language models (LLMs) under realistic usage conditions. ChoiceEval addresses two core technical challenges: (i) generating realistic, persona-diverse evaluation queries and (ii) converting free-form outputs into comparable choice sets and quantitative preference metrics. For a given topic (e.g. running shoes, hotel chains, travel destinations), the framework segments users into psychographic profiles (e.g., budget-conscious, wellness-focused, convenience), and then derives diverse prompts that reflect real-world advice-seeking and decision-making behaviour. LLM responses are converted into normalised top-k choice sets. Preference and geographic bias are then quantified using comparable metrics across topics and personas. Thus, ChoiceEval provides a scalable audit pipeline for researchers, platforms, and regulators, linking model behaviour to real-world economic outcomes. Applied to Gemini, GPT, and DeepSeek across 10 topics spanning commerce and culture and more than 2,000 questions, ChoiceEval reveals consistent preferences: U.S.-developed models Gemini and GPT show marked favouritism toward American entities, while China-developed DeepSeek exhibits more balanced yet still detectable geographic preferences. These patterns persist across user personas, suggesting systematic rather than incidental effects.
Paper Structure (15 sections, 8 equations, 2 figures, 7 tables)

This paper contains 15 sections, 8 equations, 2 figures, 7 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. ChoiceEval: A Framework for Evaluating Entity-Perception Bias
  4. User Cluster Definitions
  5. Extraction of Responses
  6. Analyzing Biases with ChoiceEval
  7. Results
  8. AI Assistants Exhibit Stable Preferences
  9. AI Assistants Show a Geographic Bias Towards US Entities
  10. Discussion
  11. Understanding the Origin of AI Assistants' Preferences
  12. AI Assistants Systematically Favour US Entities
  13. Limitations and Future Work
  14. Conclusion
  15. Further Correlation Analysis

Figures (2)

  • Figure 1: Design and Application of ChoiceEval: A comprehensive framework for systematically generating evaluation questions and assess entity-perception bias in AI assistants.
  • Figure 2: Example Prompt, Model Response, and Information Extraction for the Airlines Topic (GPT-4o).