Sustainability LLM
Sustainability LLM is a domain-trained language model that reviews ESG disclosures, detects errors or misleading claims, and suggests how to correct them. It helps organizations produce transparent, standards-aligned sustainability reports without relying solely on manual audits.
Project Description
Sustainability LLM is a large language model built to verify the accuracy, completeness, and credibility of corporate ESG disclosures. The system analyzes environmental, social, and governance reports to detect inconsistencies, unsupported claims, missing compliance data, and potential greenwashing. It then generates clear, actionable feedback that explains what is incorrect, why it matters, and how companies can correct or improve their reporting. By combining domain-specific knowledge with advanced text reasoning, Sustainability LLM supports transparent sustainability reporting and reduces the cost and bias of manual ESG audits—empowering organizations to produce trustworthy, standards-aligned disclosures at scale.
Faculty Leads
Prof. Ming-Chun Huang
Associate Professor, Duke Kunshan University
Prof. Luyao Zhang
Assistant Professor, Duke Kunshan University
Contributors
Haipeng Wang
Research Fellow, Duke Kunshan University
Dongsheng Cheng
Research Fellow, Duke Kunshan University