In recent years, important guidelines and principles have been published for the design, development, and use of applications that incorporate Artificial Intelligence (AI). These support technology creators in the development of ethically correct software. Such guidelines, however, face a challenge like that of the Declaration of Human Rights. They attempt to be globally valid, while the interpretation of what both human-centered design and AI can and should do varies based on the different regions and cultures across the world. In considering what a successful and ethical computation, action, and presentation of results by an AI-assisted algorithm might look like, general agreement can be reached in some cases. However, there are differences that need to be recognized, such as the consideration of cases where something could go wrong with an automated process. We advocate for the need of recognizing these differences in the research and design process. In not doing so, a design or its research might result in being unethical.

Our example is rather taking the perspective of research as the basis for human-centered design and human-centered AI design. We are, too, convinced, about the positive values of universal guidelines and principles yet an over-generalization might leave behind certain groups of people, regions or cultures and their interests, needs and contexts.