Implementing Responsible AI: A Life-Cycle Reference Model Based on FATES Principles

This paper presents a conceptual framework for the responsible use of Artificial Intelligence (AI), based on Fairness, Accountability, Transparency, Equity and Safety (FATES) principles. The FATES framework was directly influenced by global initiatives for designing, developing and deploying AI platforms and systems. As AI becomes increasingly embedded in various domains and sectors of industry and society, there is a critical need to monitor its responsible use. This conceptual paper first addresses the challenges of implementing AI for good, and for all. Next, it introduces the FATES principles for ethical AI practices. While discussing good practices for designing, developing and deploying AI responsibly, it synthesises insights from global public and private sector initiatives, differences between developed and developing economies and perspectives of academic scholars, policy makers, technology providers, mass media and, most importantly, the individuals and communities who adopt AI. This paper concludes with a Life-Cycle Reference Model for Responsible AI. We acknowledge that the reference model and its application to the AI implementation life-cycle are contextual. The specifics of the sector and use case determine much of assessing responsible use. Accordingly, we propose a general reference model that may be contextualised.