Due to increasing ESG corporate reporting and the lack of alignment on reported metrics, traditional ESG rating systems are nowadays unable to process and distill all the information reported by companies leading to an increase in the use of, often, low-quality metrics to influence investors’ perceptions, i.e. greenwashing. This project aims to quantify the discrepancy between the predicted and official rankings as a measure of greenwashing. Additionally, it aims to develop detection models, based on AI techniques, to automatically detect greenwashing. We will use recent advances in text and vision transformers to develop models capable of accurately quantifying sustainability-related messaging depending on the category a corporation belongs to. This will improve the explainability and interpretability of existing models and allow mapping of various textual and visual sources of information onto sustainability axes described in SDGs, such as 14 and 15 dealing with biodiversity. We will produce leads that facilitate the rating of companies based on a ‘greenwashing score’.