Exploring Translational Ethical Competency (TEC) in AI Design: Conceptual Analysis and Methodological Reflections from a Pilot Study

As AI is rapidly becoming an integral part of our daily life, the ethical issues associated with it have become a major issue. Although the ethical concerns regarding the development of AI systems have been discussed in detail during recent years, practical competencies for translating ethical principles into actionable practices have not been fully understood. To fill this gap, this paper uses the concept of Translational Ethical Competency (TEC) in the context of AI design. We share our reflection on a pilot study we conducted, which involved semi-structured text-based elicitation interviews with AI professionals to examine how ethical principles are incorporated into the real-world development of AI systems. Four case studies were designed as an elicitation tool to guide the discussion, covering domains such as educational fact checking, autonomous vehicles, stock price prediction, and blood diagnostics. Based on our reflection at the pilot interview stage, we revised our interview protocol. In this paper, our reflection on the pilot study, protocol revision process, and future work will be discussed. We aim to identify the best practices to incorporate ethical principles into the AI design process, which can provide practical insights for AI ethics education and professional development.