Leveraging Language Models to Detect Greenwashing

TL;DR

This paper presents a novel approach using a fine-tuned language model to detect greenwashing in corporate sustainability reports, achieving promising accuracy and F1 scores, and introduces a mathematical framework for quantifying greenwashing risk.

Contribution

It introduces a new methodology with a mathematical formulation and a fine-tuned ClimateBERT model for greenwashing detection in sustainability reports.

Findings

Achieved 86.34% accuracy on test data.

F1 score of 0.67 indicates effective detection.

Demonstrated the feasibility of language models for greenwashing risk assessment.

Abstract

In recent years, climate change repercussions have increasingly captured public interest. Consequently, corporations are emphasizing their environmental efforts in sustainability reports to bolster their public image. Yet, the absence of stringent regulations in review of such reports allows potential greenwashing. In this study, we introduce a novel preliminary methodology to train a language model on generated labels for greenwashing risk. Our primary contributions encompass: developing a preliminary mathematical formulation to quantify greenwashing risk, a fine-tuned ClimateBERT model for this problem, and a comparative analysis of results. On a test set comprising of sustainability reports, our best model achieved an average accuracy score of 86.34% and F1 score of 0.67, demonstrating that our proof-of-concept methodology shows a promising direction of exploration for this task.