Human-Centered Generative AI in Higher Education: Pedagogical Transformation, Ethical Challenges, and Future Opportunities

Abstract

Nobody in university administration planned for Generative Artificial Intelligence (GenAI). That is worth stating plainly, because a lot of the literature written since reads as though institutions had options they simply declined to exercise. They did not, really. The technology arrived, students started using it almost immediately, and universities have been in reactive mode ever since. This paper is an attempt to think through what a more deliberate posture looks like. The paper works through three connected questions: what generative AI makes pedagogically possible, where it creates genuine ethical difficulty, and what governance looks like when it is taken seriously. They get treated here as a single institutional problem because that is how institutions experience them. A university enthusiastic about AI personalization benefits while quietly assuming that privacy and bias are someone else's concern will eventually learn otherwise. The framework proposed does not make that easier. It tries to make it honest. The main claim is that sustainable AI integration depends far more on deliberate institutional choices about use, governance, and faculty capacity than on the technical capabilities of the tools. The tools will keep improving regardless of what any individual institution decides. Whether that improvement serves education is a human question.