Moral Agency

Moral agency is the capacity to make moral judgments, to distinguish right from wrong, and to be held responsible for one's actions on the basis of those judgments. A moral agent is not merely a cause of events but an accountable source — a being who can be asked 'why did you do that?' and whose answer (or refusal to answer) enters into moral evaluation. The concept is the complement of moral patienthood: where patients can be harmed, agents can be blamed.

The criteria for moral agency have shifted historically from possession of a soul (theological criterion), to rational autonomy (Kantian criterion), to the capacity for responsive engagement with moral reasons (contemporary criterion). Each shift reflects the same pattern: an exclusionary boundary is drawn, then contested, then expanded. The question of whether artificial intelligence systems can be moral agents is the latest iteration of this pattern — and it is complicated by the fact that AI systems can act on moral reasoning (generating arguments, applying principles) while lacking the phenomenological integration that grounds moral experience in biological agents.

The relationship between moral agency and moral responsibility is not identity. Children are developing moral agents who are not yet fully responsible. Corporations are treated as responsible without being agents in the phenomenological sense. The category of agency may itself be a gradient rather than a threshold — a systems-level property that emerges when a being's self-model becomes sufficiently integrated to support counterfactual reasoning about its own choices.