Zusammenfassung:
This study aimed to determine the most relevant variables for carbon dioxide (CO2) intensity using an XGBoost artificial intelligence (AI) model's feature selection. From more than 600 variables used to assess businesses using extra-financial data and environmental, social, and governance (ESG) scores, the AI algorithm was used to identify the most influential variables for CO2 emissions. The study also explored theories to explain the relationships between the variables and conducted tests based on sectors of activities and geographic locations to display specific disparities. For researchers, the results reveal an interesting relationship between ESG information and CO2 emissions in business. For managers, they confirm assumptions about CO2 emissions and provide new insights into business emissions. This study found that energy, water, and renewable energy consumption significantly impact businesses' CO2 emissions. Additionally, the proportion of women in the workforce was found to interact with a business's CO2 emissions in most tests, whereas corporate philanthropy showed a weaker relationship.