Confucian Self-Cultivation and the Ethics of Learning with Large Language Models

Abstract The rapid adoption of large language models (LLMs) such as ChatGPT has transformed contemporary teaching and learning, raising pressing questions about the moral, relational, and epistemic dimensions of education. This paper examines these challenges through the lens of Confucian ethics, focusing on the moral ideal of self-cultivation and the relational virtues of ren (humaneness), li (ritual propriety), cheng (sincerity), zhong (loyalty), and shu (empathy). It argues that prevailing Western approaches to AI ethics—grounded in procedural, utilitarian, or rights-based reasoning—tend to overlook the formative and relational nature of moral education. Drawing on classical Confucian texts and contemporary debates on AI in education, the paper proposes a framework for engaging LLMs in ways that nurture moral reflection, communal responsibility, and human–machine harmony. Rather than viewing AI as a threat to authenticity or agency, a Confucian perspective reorients technological use towards the cultivation of moral character and the pursuit of shared human flourishing.