Can Generative AI Help Stakeholders Identify Ethical Firms? A Study of CSR Reports

Stakeholders are increasingly interested in identifying ethical firms. One way they can do this is through the language leaders use to describe their firm, which is a reflection of the firm’s ethical culture and/or climate. Artificial Intelligence (AI), such as a Large Language Models (LLMs), offers a way to analyze large textual data sets and measure the degree to which firm leaders use moral language in their stakeholder communication. Can LLMs help stakeholders gain insight into a firm’s ethicality? Focusing on the Moral Foundations Theory’s (MFT) Care moral intuition, we empirically test if our customized GPT-4 model can recognize moral language in corporate social responsibility (CSR) reports. Comparing GPT results with human coders, we find mixed evidence on the ability of LLM to measure moral language in CSR reports.

In The Proceedings of 85th Annual Meeting of the Academy of Management, Copenhagen, Denmark, 2025