Recent research has highlighted how large language models (LLMs)—the same type of technology that powers ChatGPT—can be trained to identify ESG-related statements in corporate reports. The study, released on arXiv in February 2025, demonstrated how AI can classify disclosures into environmental, social, and governance categories, accelerating ESG analysis and improving consistency across sectors. This could revolutionize the way ESG data is interpreted by both investors and regulators.

However, the same study warns of challenges: AI outputs are only as reliable as their training data. If trained on biased or narrow datasets, models may reinforce inaccuracies or overlook key context. This is why GenESG’s dual focus on technical competence and ethical reflection is so vital. The project encourages learners not only to use AI tools—but to question their limitations and apply them with integrity.

Reference: Meunier, A., Sahin, O., & Rezaei, J. (2025). Leveraging LLMs to classify ESG-relevant content in corporate disclosures.