Hegemonikon and AI

Abstract

The central ethical problem raised by artificial intelligence is not whether AI systems can "reason" in a functional sense, but whether their use preserves a centre of judgment that can be held responsible. Beginning with large language models, it distinguishes linguistic fluency from scientific validity, ethical commitment, and accountability. Enter the Stoic concept of the hegemonikon—the capacity for assent that directs the mind—as a framework for understanding intellectual creation and decision-making. This framework is extended by analogy to institutions, clarifying when AI supports human judgment and when it replaces decision-making authority. The argument is applied to various fields and leads to five principles for the ethical use of AI: disclosure, attribution, critical engagement with AI outputs, meaningful human oversight, and augmentation rather than replacement. Responsible AI governance requires institutions to preserve decision-making structures that are identifiable, open to challenge and revision, and subject to accountability.