Comment by Patrick Hedfeld

This formulation does not attribute consciousness, intentionality, or moral personhood to AI systems. Nor does it collapse distinctions between humans and machines. [...] Preserving this distinction is essential because it allows the paper to acknowledge the structuring force of AI while retaining the primacy of human and institutional responsibility. In this respect, the socio-technical concept of AI offers a middle path between two inadequate extremes: the view of AI as a neutral instrument with no independent structuring effects, and the view of AI as an autonomous quasi-person capable of bearing responsibility in its own right.