A Detailed Study on LLM Biases Concerning Corporate Social Responsibility and Green Supply Chains

2025-11-03

AI Summary

This study identifies systematic value biases in large language models (LLMs) regarding corporate social responsibility (CSR) and green supply chain management—specifically, implicit prioritization of Western-aligned sustainable business strategies that distort ethical reasoning and managerial decision-making. Methodologically, we design organizational culture prompts grounded in Hofstede’s cultural dimensions and conduct standardized questionnaire-based experiments across multiple LLMs (GPT-4, Claude, Llama). Results demonstrate a consistent Western-centric and developmentalist bias across all models; critically, culturally calibrated prompts significantly modulate LLM outputs—shifting attributions of responsibility, reconfiguring supply chain accountability allocations, and redirecting sustainability recommendations. The study contributes a novel bias identification framework for ESG governance applications and establishes empirically validated, culture-sensitive intervention pathways to enhance LLM reliability and contextual appropriateness in sustainability decision support.

Technology Category

• Natural Language Processing: Ethics — Bias, Fairness, Transparency & Privacy
• Machine Learning: Ethics, Bias, and Fairness
• Philosophy and Ethics of AI: Bias, Fairness & Equity

Application Category

• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Organizations increasingly use Large Language Models (LLMs) to improve supply chain processes and reduce environmental impacts. However, LLMs have been shown to reproduce biases regarding the prioritization of sustainable business strategies. Thus, it is important to identify underlying training data biases that LLMs pertain regarding the importance and role of sustainable business and supply chain practices. This study investigates how different LLMs respond to validated surveys about the role of ethics and responsibility for businesses, and the importance of sustainable practices and relations with suppliers and customers. Using standardized questionnaires, we systematically analyze responses generated by state-of-the-art LLMs to identify variations. We further evaluate whether differences are augmented by four organizational culture types, thereby evaluating the practical relevance of identified biases. The findings reveal significant systematic differences between models and demonstrate that organizational culture prompts substantially modify LLM responses. The study holds important implications for LLM-assisted decision-making in sustainability contexts.

Problem

• Investigates LLM biases in corporate social responsibility and green supply chains
• Analyzes how organizational culture types affect LLM response variations
• Evaluates training data biases affecting sustainable business strategy decisions

Innovation

• Analyzed LLM responses using standardized questionnaires
• Evaluated bias variations across four organizational cultures
• Identified systematic differences in sustainability prioritization biases