Piloting a maturity model for responsible artificial intelligence: A portuguese case study

Abstract

Recently, frameworks and guidelines aiming to assist trustworthiness in organizations and assess ethical issues related to the development and use of Artificial Intelligence (AI) have been translated into self-assessment checklists and other instruments. However, such tools can be very time consuming to apply. Aiming to develop a more practical tool, an Industry-Wide Maturity Model for Responsible AI was piloted in 3 companies and 2 research centres, in Portugal. Results show that organizations are aware of requirements (44 %) to deploy a responsible AI approach and have a reactive response to its implementation, as they are willing to integrate other requirements (33 %) into their business processes. The proposed Model was welcomed and showed openness from companies to consistently use it, since it helped to identify gaps and needs when it comes to foster a more trustworthy approach to the development and deployment of AI.