Research Information System
OPEN PRAXIS, vol.17, no.2, pp.206-210, 2025 (ESCI)
This paper critically examines the integration of generative and agentic AI systems within the educational landscape, arguing that they are not merely advanced tools but active participants in a complex reconfiguration of student agency. Moving beyond a simplistic human-vs-machine dichotomy, this paper posits that as AI systems gain the ability to generate novel academic content and execute research tasks autonomously (machine agency), they paradoxically erect an "algorithmic panopticon" that subtly constrains and reshapes human agency if adopted blindly. This paper will explore how student intent and intellectual exploration are increasingly mediated, guided, and superseded by AI-driven frameworks. This leads to a critical risk of "modal collapse"-an intellectual homogenization where student outputs converge on a narrow set of algorithmically-preferred styles and conclusions. This paper argues that these technologies function as powerful accelerants of a "post-truth" condition within education, a state where the lines between fact and fiction are deliberately blurred and trust in established knowledge is eroded. The paper further analyzes the implications of this shift for the core educational tenets of autonomy, critical thinking, and academic responsibility.