Advances in AI have led to the increasing deployment of AI-based care robots in institutional care settings. While ethical debates have addressed responsibility in relation to such systems, insufficient attention has been paid to their causal responsibility under conditions of inherent causal indeterminacy. This article argues that AI-based care robots can be causally responsible for outcomes to which they contribute, even though they lack moral agency. Drawing on three assumptions (graduality, evaluation, and scope) the analysis clarifies when robotic actions qualify as necessary part causes within responsibility practices. At the same time, it shows that care robots exhibit forms of inherent causal indeterminacy arising from (1) distributed design histories, (2) opaque runtime processes, and (3) embodied contingencies. This indeterminacy does not negate causal responsibility but constrains how determinately it can be specified. Distinguishing causal from moral responsibility is therefore essential: while causal responsibility concerns whether an action counts as a necessary part cause, moral responsibility requires additional normative conditions such as control, will or knowledge. Recognizing robots as causally relevant actors refines institutional accountability practices without attributing moral agency to machines.
Mario Kropf (Wed,) studied this question.