AI in Education Supporting Student Agency

The expanding use of generative artificial intelligence in education has raised concerns that AI may displace students' intellectual work and weaken their agency. This essay examines the conditions under which AI can instead support student agency. It brings together psychological accounts of agency, Biesta's concepts of qualification and subjectification, Arendt's accounts of thinking, judgment, plurality, and action, and research on cognitive offloading, interactivity/interpassivity, knowledge creation, and collaboration. The central distinction is between delegating limited operations to AI and delegating intellectual responsibility for the assignment. AI can support learning by providing access to information, diverse tools, feedback, and assistance adapted to students' developing expertise. That can free students cognitive resources for knowledge creation, value-based judgment, and collective inquiry. These benefits depend on pedagogy that preserves students' responsibility for formulating questions, evaluating outputs, justifying claims, and acting on their conclusions. The essay argues that the educational aim should be AI-augmented epistemic agency, in which students use AI to expand what they can investigate and create while retaining understanding, authorship, judgment, and responsibility.