Publication details

The Principle of Non-Discrimination in the Infosphere: A New Ethics



Year of publication 2019
MU Faculty or unit

Faculty of Education

Description The High-Level Expert Group on Artifical Intelligence (AI), which works at the European Commission, presented seven points that must be met in the European space by companies developing the artificial intelligence. Among the seven points that can be described as "a new laws of robotics" is the principle of diversity, non-discrimination, and fairness. It is not allowed to create artificial intelligence that could discriminate anybody. In Western culture, the ethical framework of reflection on non / discrimination is on a scale between two poles – deontology and consequence-based ethics. However, AI research shows that such a scale is traditionally formulated from the anthropocentric point of view, and thus does not adequately cover the transformations that our world is going through. Above all, it is unable to reflect: • phenomena of human cooperation with AI; • AI autonomous behavior; • responsibility in hybrid cognitive systems. Even the above-mentioned "the laws of robotics" have this anthropocentric ethical character. Our contribution is based on Lucian Floridi's information ethics, which emphasizes the need to include all entities that are capable of processing information and thus contributing to the formation of the infosphere. We propose principles of non-discrimination that can be extended from people to all inforgs (i.e. people as well as biotic and technical entities in the infosphere) and show what practical impacts such a change in ethical perspective can bring.