Reconfiguring Human Subjectivity in the Age of Generative Artificial Intelligence: A Philosophical Reflection and Evidence-Informed Framework

Abstract

Generative artificial intelligence (GenAI) has moved from a specialized tool to an everyday cognitive partner. By producing fluent text, images, and plans at scale, GenAI now intervenes directly in how people perceive, decide, create, and relate to others. These interventions redraw long-standing boundaries of human subjectivity: the assumed primacy of human cognition, the autonomy of value judgment, and the exclusive right to build and inhabit meaningful worlds. Such shifts are ambivalent. Human–AI co-creation can expand the horizon of expression, accelerate sense-making, and enable new forms of collective intelligence; yet deep algorithmic mediation may also weaken agency through over-delegation, outsource normative reasoning, and intensify alienation. From an interdisciplinary perspective that combines phenomenology, philosophy of technology, and Marxian theories of subjectivity, this paper develops a structured framework for analyzing how GenAI reconfigures subjectivity across cognition, practice, and intersubjective life. Building on this framework, we propose a ‘subjectivity-in-the-loop’ path that couples technical affordances with value constraints and reflexive governance. A teaching implementation case is presented to operationalize the framework in higher education, and an evidence-informed evaluation protocol is provided with simulation-based analyses to demonstrate the measurement pipeline. Our central claim is that GenAI is not simply an external antagonist to subjectivity; it is a historically specific condition under which subjectivity can be reconstituted. Safeguarding human subjectivity therefore requires rational scrutiny of technological mediation, explicit value regulation, and institutional designs that keep responsibility and meaning-making anchored in human communities.