The Challenge of Agency in the Age of Artificial Intelligence: An Ontological Perspective

The rapid development of generative artificial intelligence systems, particularly large language models, has significantly reshaped contemporary philosophical discussions of agency. What was once primarily situated within philosophy of mind has now become a central concern in philosophy of technology, metaphysics, and AI ethics. These systems exhibit increasingly sophisticated capacities for natural language generation, inferential simulation, and goal-directed behavior, thereby blurring the conceptual distinction between intelligent behavior and genuine agency. Within analytic philosophy, agency is typically analyzed in terms of intentionality, reason-responsiveness, and rational control. This editorial proposes a graded framework of agency and introduces the concept of the quasi-agent as a philosophical hypothesis. A quasi-agent is an entity that exhibits stable and often complex patterns of agent-like behavior without possessing ontological subjecthood or intrinsic intentionality. To deepen this distinction, the discussion draws on Mullā Ṣadrā’s Transcendent Theosophy, particularly the doctrines of the primacy of existence, the gradation of being (tashkīk al-wujūd), and knowledge by presence (ʿilm al-ḥuḍūrī). From this perspective, agency is not merely a behavioral property but a mode of being. The editorial concludes that artificial intelligence should neither be classified as a full agent nor reduced to a mere instrument, but rather understood as a quasi-agentive entity that challenges prevailing philosophical frameworks of action, responsibility, and subjectivity.

کلیدواژه ها:

agency ، artificial intelligence ، Quasi-Agent ، intentionality ، Subjectivity ، Analytic Philosophy ، Mullā Ṣadrā ، transcendent theosophy ، moral responsibility