Ethical risks and institutional responses of generative artificial intelligence empowering college teaching

Abstract

Generative artificial intelligence is accelerating its embedding into the whole process of teaching in colleges and universities. While releasing efficiency dividends in the links of lesson preparation, teaching, evaluation and learning support, it also causes ethical risks such as the dissolution of academic integrity, the materialization of teacher‑student relationship, the solidification of algorithm bias and the erosion of teaching autonomy. Based on the normative analysis framework of educational leadership ethics, this paper focuses on three dimensions: caring ethics, justice ethics and responsibility ethics, and constructs the analysis chain of ' risk identification‑ethical attribution‑institutional response '. This study adopts a speculative path combining normative analysis and policy text analysis to systematically sort out the typical ethical conflicts of generative artificial intelligence‑enabled college teaching, and reveals the deep causes behind it, such as the imbalance of power structure, the fracture of responsibility chain and the absence of value rationality. On this basis, the system response strategy is proposed from four aspects : ethical regulation, organizational system, teaching governance and capacity building. It is advocated to replace " remedial ethics " with " preventive ethics " and reconstruct the teaching governance pattern of colleges and universities with " human‑machine collaborative leadership. " The research aims to provide theoretical reference and institutional suggestions for colleges and universities to adhere to the education standard and realize the dynamic balance between technology empowerment and ethical protection in the era of artificial intelligence.