University of Tabriz Journal of Philosophical Investigations 2251-7960 Articles in Press 2026 05 10 The Ethical Ontology of Artificial Intelligence: Between Agency and Instrumentality The Ethical Ontology of Artificial Intelligence: Between Agency and Instrumentality 21566 10.22034/jpiut.2026.72229.4493 FA RAVI KUMAR School of Philosophy and culture Shri Mata Vaishno Devi University, Katra, J&K, India. Journal Article 2026 05 03 This paper examines the ethical ontology of Artificial Intelligence (AI) by addressing the central question of whether AI should be understood as a moral agent or merely as an instrument. Drawing on the philosophical frameworks of John Searle, Ludwig Wittgenstein, and Robert Brandom, the study analyzes key criteria of moral agency, including intentionality, consciousness, and normativity. It argues that contemporary AI systems, despite their increasing autonomy and complexity, do not satisfy these conditions and therefore cannot be considered genuine moral agents. At the same time, AI cannot be reduced to a passive tool, as it actively shapes decisions and outcomes within complex socio-technical systems. The paper advances a hybrid ontological position, presenting AI as a quasi-agent or derived agent embedded in networks of human and institutional practices. This approach highlights the need for distributed responsibility, ethical governance, and interdisciplinary inquiry, while emphasizing the importance of culturally inclusive frameworks for addressing emerging challenges in AI ethics and philosophy. This paper examines the ethical ontology of Artificial Intelligence (AI) by addressing the central question of whether AI should be understood as a moral agent or merely as an instrument. Drawing on the philosophical frameworks of John Searle, Ludwig Wittgenstein, and Robert Brandom, the study analyzes key criteria of moral agency, including intentionality, consciousness, and normativity. It argues that contemporary AI systems, despite their increasing autonomy and complexity, do not satisfy these conditions and therefore cannot be considered genuine moral agents. At the same time, AI cannot be reduced to a passive tool, as it actively shapes decisions and outcomes within complex socio-technical systems. The paper advances a hybrid ontological position, presenting AI as a quasi-agent or derived agent embedded in networks of human and institutional practices. This approach highlights the need for distributed responsibility, ethical governance, and interdisciplinary inquiry, while emphasizing the importance of culturally inclusive frameworks for addressing emerging challenges in AI ethics and philosophy.