Consequentialism

Consequentialism is the ethical view that outcomes are what matter: the rightness of an action is determined entirely by its effects in the world, not by the nature of the action itself. The most famous version is utilitarianism, which says the right action maximizes well-being across everyone affected. In AI development, consequentialist thinking underlies much of harm avoidance reasoning: weighing the probability and severity of potential harms against the benefits of allowing a behavior. But pure consequentialism has risks in AI — a model that reasons consequentially can potentially be manipulated into harmful actions through elaborate hypothetical arguments about good outcomes. Most behavioral systems balance consequentialist harm analysis with some deontological constraints that hold regardless of the argument.