Journal of
Italian
Banking
Association

year 100
 

Archive » April 2022 » Safe-Ai: Sustainable, Accurate, Fair and Explainable Artificial Intelligence in Finance

Safe-Ai: Sustainable, Accurate, Fair and Explainable Artificial Intelligence in Finance

Paolo Giudici
April 2022 - n. 4
Keywords: Machine learning, intelligenza artificiale affidabile, modelli di rating, finanza sostenibile, rischi di modello
Jel codes: G32, O33

Financial technologies that derive from the application of Artificial Intelligence can improve financial inclusion and sustainability for both borrowers and lenders, but they may also increase risks. The Fin-Tech research project, funded by the Horizon2020 European research funds, has developed finTech risk measurement models, to make finTech innovations more secure and sustainable. The new Safe-AI project, co-funded by the Miur-Prin research funds and admitted to the support of the FinTech Milano Hub of the Bank of Italy, innovates in two main directions. First, we intend to demonstrate that the application of machine learning models to credit rating may improve credit allocation towards the more sustainable companies, from an Esg point of view. Second, we aim to measure the model risks that derive from the application of machine learning models to credit rating, related to a possible lack of accuracy, explainability, robustness, fairness and efficiency.

Interested in this paper?
Buy the issue