Health is a complex, resource-intensive domain where technologies raise important questions about value and values. AI-enabled technologies in particular show immense potential to shape both individual health care encounters and the broader organizational arrangements which comprise our health care systems. As such, it is important to engage critically with how these technologies might impact upon health care goals and functions, including privacy, clinical decision making, or health system resource allocation. One set of approaches to considering the role of values in the development of new technologies is design ethics. Design ethics is an applied ethics which aims to understand and identify ethical dilemmas arising through and from design. One of the main advantages of design-based approaches is the ‘front-loading’ of ethics – or going beyond ethical applications, uses, and impacts of technologies, to the ‘constructive’ incorporation of ethical considerations early in design and development phases. At the same time, debate remains about how exactly ethics ought to be incorporated into the design process, and who should be responsible for it. This submission represents the culmination of a recent 6-month research project conducted as part of the scholars program in AI ethics and health at the University of Toronto’s Joint Centre for Bioethics. It will present a typology of ‘ethics by design’ approaches based on normative and descriptive claims, and will conclude with recommendations for future research and practice, including how these approaches may assist in the realization of fair, transparent, and inclusive AI for health, with implications for health system actors and society more broadly.