The subproject Machines as moral, political and legal actors“ is led by Luise K. Müller (TU Dresden, associated member of the SOCAI centre). The project asks how the effects of aggregated — individually efficient and rational — algorithmic decision procedures, and the resulting costs and benefits can be distributed fairly. The current focus on the moral competence of AI in the machine ethics literature seems to be of limited use, as it is neither necessary nor sufficient in these cases for machines to make moral decisions in a strong sense. What is rather needed is rule-based coordination of AI in favour of a fair distribution of the costs and benefits that result from aggregated decision-making. The core question thus is: How do we design an effective structure of distribution, that ensures respect for the freedom and equality of moral subjects in the context of algorithmic decision procedures? The focus on aggregated algorithmic decision-making, which are in and of themselves not unfair, but may produce patterns of structural injustice, arises from the perspective of political theory and philosophy. As one possible way of responding to this problem, the project seeks to analyse whether it makes sense to understand AI as bearer of duties.
Similarly complex questions exist in law. These questions range from the consideration about a potential constitutional right to ways in which private law would address contract formation by machines. Legal dogmatics would reach its limits once machines themselves become actors of social practices. Particularly in the area of legal responsibility, the key issues remain unclear.