Beyond Plagiarism: A Confucian Reflection on Teaching Philosophy with ChatGPT

When ChatGPT first became widely available, many conversations among philosophy teachers centred on a familiar concern: plagiarism. How could we tell whether students had written their own essays? Should AI-generated text be banned? Should we redesign assessments to make cheating more difficult?

These are important questions, but they gradually became less interesting to me than another question. Even if students use ChatGPT honestly and transparently, are they still genuinely learning?

This question emerged not from abstract philosophical speculation but from my own classroom experience. Increasingly, I noticed that some students could submit polished written work produced with the assistance of AI, yet struggled to explain, defend, or extend the ideas during class discussion. The essays often appeared coherent and well organized, but when students were asked why a particular argument was persuasive, or whether an objection might undermine it, the conversation sometimes came to an abrupt halt. The difficulty was not that students had used AI. Rather, it was that the process of thinking seemed to have been displaced by the process of producing an answer.

As a scholar of Confucian philosophy, I found myself returning to a simple but profound observation by Confucius:

“Learning without reflection leads to confusion; reflection without learning leads to peril.”

Although written more than two millennia ago, this insight has become remarkably relevant in the age of large language models.

The educational risk posed by ChatGPT is not simply that students may outsource their writing. More fundamentally, they may begin to outsource reflection itself.

This realization has gradually changed both the way I think about AI and the way I teach with it.

Rather than asking whether students should use ChatGPT, I now ask how they use it. I have become less interested in detecting AI-generated text than in encouraging students to examine their own thinking while interacting with AI. In class discussions, I often ask students not merely to present ChatGPT’s answer but to explain why they accepted part of it, rejected another part, or modified its suggestions. I want them to identify where their own judgement begins.

From a Confucian perspective, education has never been understood as the simple acquisition of information. Its ultimate purpose is self-cultivation. Learning is valuable because it gradually transforms the learner’s character, judgement, and relationships with others. Knowledge matters, but even more important is the kind of person one becomes through the pursuit of knowledge.

This perspective helps explain why AI presents both an opportunity and a challenge.

Large language models are remarkably effective at retrieving information, summarizing arguments, generating examples, and improving written expression. I encourage students to make use of these capabilities. Indeed, refusing to engage with such technologies would neither prepare students for contemporary intellectual life nor acknowledge the genuine educational benefits that AI can offer.

At the same time, Confucianism reminds us that understanding cannot be reduced to producing fluent language. Genuine learning requires what Confucian thinkers describe as sincerity—the alignment between one’s inner understanding and outward expression. A beautifully written essay has little educational value if it does not genuinely express the student’s own developing judgement. The concern is therefore not whether AI contributes to the writing process but whether it substitutes for the learner’s own intellectual and moral engagement.

This way of thinking has also changed my understanding of the teacher’s role.

When AI can instantly explain philosophical concepts or summarize complex texts, it may appear that teachers are becoming less necessary. Yet I have found almost the opposite to be true. If information becomes increasingly abundant, the teacher’s distinctive contribution becomes increasingly human.

Confucianism has long regarded teachers not primarily as transmitters of information but as moral exemplars whose attitudes towards learning influence their students. In an AI-rich classroom, I no longer feel that my authority depends upon knowing more facts than my students or than ChatGPT. Instead, my responsibility is to model intellectual humility, curiosity, honesty, and careful judgement. Students are constantly observing not only what I know but how I respond to uncertainty, disagreement, and complexity.

Ironically, AI has reminded me that philosophy has always been less about possessing answers than about cultivating habits of thoughtful inquiry.

This conviction has also influenced the kinds of classroom activities I now design. Rather than asking students simply to use ChatGPT to produce a response, I invite them to interrogate its reasoning. We compare AI-generated interpretations with classical philosophical texts, identify hidden assumptions, discuss where the AI oversimplifies complex concepts, and consider what important human experiences may be missing from its responses. These conversations often become richer than they would have been without AI because the technology provides a common object of philosophical critique.

In this sense, ChatGPT has become less a source of answers than a catalyst for dialogue.

Perhaps the most important lesson I have learned is that educational questions about AI are ultimately questions about educational purpose. If the purpose of education is merely the efficient production of correct answers, then AI appears to fulfill that goal remarkably well. If, however, education seeks to cultivate intellectual judgement, moral responsibility, and reflective persons, then AI cannot replace the learner’s own participation in that process.

Confucian philosophy encourages us to evaluate educational technologies not according to how much work they save but according to the kinds of persons they help us become. Used unreflectively, ChatGPT may encourage intellectual dependence. Used thoughtfully, however, it can become an occasion for deeper self-examination. Every interaction with AI invites students to ask not only “Is this answer correct?” but also “Why do I find it convincing?”, “What assumptions does it make?”, and “How has it changed my own thinking?” These are philosophical questions in the fullest sense.

I therefore remain cautiously optimistic about the place of AI in philosophy classrooms. I do not think our task is to resist these technologies or to embrace them uncritically. Rather, it is to ensure that they remain servants of education rather than substitutes for it.

Confucius viewed learning as a lifelong process of becoming more reflective, more humane, and more responsible. That aspiration has not changed simply because our tools have become more powerful. If anything, the arrival of ChatGPT has made it more urgent. The challenge facing philosophy teachers today is not merely to help students use AI well, but to help them become the kinds of persons who can use it wisely.