Ethical memory is the mechanism by which an AI system retains, references, and re-applies moral reasoning across time and context. Unlike static rule enforcement, ethical memory evolves as systems interact with users, interpret feedback, and accumulate precedent.

In the NRBC framework, social values serve as the grounding layer for ethical memory. These include relational values like trust, respect, and reciprocity, as well as collective values like solidarity, nonviolence, and equality. These are not encoded once but cultivated through iterative exposure and moral reinforcement.

For AI systems to act with ethical memory, they must be able to recognize continuity in moral context. This involves retaining structured moral histories, adjusting behavior when patterns of harm or misalignment arise, and offering transparent justifications for decisions made across time.

Ethical memory is what allows AI to behave not just consistently, but accountably. It is the bridge between single-instance decision-making and long-term moral character.