alignments

Alignments are a framework for describing an agent's or character's moral and thematic outlook along two axes: law versus chaos and good versus evil. The most widely known scheme uses nine alignments, ranging from Lawful Good to Chaotic Evil, with variations such as Neutral Good, True Neutral, and Lawful Neutral.

In role-playing games and fantasy fiction, alignments are used to guide behavior, influence decisions, and shape interactions with other characters. They are typically descriptors rather than rigid rules; players may improvise or allow characters to deviate from their stated alignment. The system has been influential but also criticized for oversimplifying moral psychology and constraining nuanced storytelling.

In ethical and philosophical discussions, alignment is considered as the degree of concordance between beliefs, values, and actions, or between normative theories and actual judgments. It is used to analyze consistency, responsibility, and the implications of moral theories in real-world contexts.

In artificial intelligence, alignment denotes the effort to ensure that an AI's goals and behavior align with human values, safety requirements, and societal norms. Research topics include value alignment, corrigibility, robustness to distributional shifts, and interpretability. The field addresses theoretical foundations as well as practical safeguards to prevent harmful or unintended outcomes.

Outside gaming and AI, alignment can refer to harmonizing plans with standards in engineering, aligning narratives and character arcs in writing, or aligning organizational policies with ethical principles. Across domains, alignment emphasizes the fit between an agent’s motivations or rules and the broader values or objectives it is meant to serve.