Abstract:

Recent developments in agentic artificial intelligence (AI) have intensified long-standing debates about the nature of agency, particularly in contexts where learning, decision-making, and creativity are central (Leonardi 2025; Moruzzi 2022; Patel 2025). This paper examines how agency is understood and cultivated in the contemporary university classroom at a moment when AI tools are increasingly involved in tasks traditionally associated with human initiative, judgment, and authorship. In educational research, agency is commonly defined as the capacity of learners to intentionally shape their learning trajectories by identifying meaningful goals, mobilizing cognitive and motivational resources, and regulating their actions in pursuit of those goals within specific social and institutional contexts (Matusov et al. 2016; Betz and Hackett 1987). Understanding agency in the contemporary classroom, therefore, requires not only protecting human capacities for judgment and creativity but also critically examining how these capacities evolve within increasingly hybrid human–AI environments.

Based on interdisciplinary scholarship in education, philosophy, creativity, and recent AI studies, this paper proposes understanding classroom activity as a space of distributed agency in which students, instructors, and computational systems interact in complex ways. We introduce an assessment framework for balancing agentic AI with preserving and strengthening human agency:
(1) transparency and explainability of AI-mediated decisions,
(2) alignment with pedagogical intents and curricular fairness,
(3) preservation of student voice, choice, creativity, and responsibility, and
(4) accountability mechanisms for instructors and institutions.

This framework conceptualizes “distributed agency” in the learning process as a condition in which human and technological agents contribute different forms of capabilities, while responsibility for meaning-making, judgment, and ethical decision-making remains anchored in human participants. From this perspective, AI systems, particularly emerging AI agents, can function as cognitive and creative partners that expand the range of possible learning practices without displacing the normative foundations of human agency. Sustainable agency in the university classroom emerges from deliberate design and reflective governance.

The paper concludes by suggesting pedagogical approaches to reflective engagement with AI, encouraging students to develop meta-awareness about their choices, tools, and intellectual processes. We also present practical implications for curriculum design and policy development, advocating a participatory approach in which students, educators, and technologists co-create the conditions under which agency is enacted, negotiated, and evaluated.

References:
[1] Betz and Hackett. 1987 “Concept of Agency in Educational and Career Development” Journal of Counselling Psychology 34 (3).
Leonardi. 2025 “Homo Agenticus in the Age of Agentic AI” Information and Organization 35 (3).
[2] Matusov and Kayumova. 2016 “Mapping Concepts of Agency in Educational Contexts” Integrative Psychological and Behavioural Science 50 (3).
[3] Moruzzi 2022 “Creative Agents: Rethinking Agency and Creativity in Human and Artificial Systems” Journal of Aesthetics and Phenomenology 9 (2).
Patel. 2025 “Exploring the Role of Agentic AI in Enhancing Human-AI Collaboration” International Journal Of Engineering Technology Research & Management 09 (08).

Keywords:

Agency, artificial intelligence, higher education, disruptive technologies.