Towards a Business Case for AI Ethics

Abstract

The increasing integration of artificial intelligence (AI) into software engineering (SE) highlights the need to prioritize ethical considerations within management practices. This study explores the effective identification, representation, and integration of ethical requirements guided by the principles of IEEE Std 7000–2021. Collaborating with 12 Finnish SE executives on an AI project in autonomous marine transport, we employed an ethical framework to generate 253 ethical user stories (EUS), prioritizing 177 across seven key requirements: traceability, communication, data quality, access to data, privacy and data, system security, and accessibility. We incorporate these requirements into a canvas model, the ethical requirements canvas. The canvas model serves as a practical business case tool in management practices. It not only facilitates the inclusion of ethical considerations but also highlights their business value, aiding management in understanding and discussing their significance in AI-enhanced environments.