Algorithmic Ethics: Formalization and Verification of Autonomous Vehicle Obligations

We develop a formal framework for automatic reasoning about the obligations of autonomous cyber-physical systems, including their social and ethical obligations. Obligations, permissions prohibitions are distinct from system's mission, necessary part specifying advanced, adaptive AI-equipped systems. They need dedicated deontic logic to formalize them. Most existing logics lack corresponding algorithms system models that permit verification. demonstrate how particular logic, Dominance Act Utilitarianism (DAU), is suitable starting point formalizing systems like self-driving cars. its usefulness by subset Responsibility-Sensitive Safety (RSS) in DAU; RSS an industrial proposal cars should not behave traffic. show certain logical consequences undesirable, indicating further refine proposal. also can change over time, which long-term autonomy. then model-checking algorithm DAU formulas on weighted transition illustrate it car controller literature.