A Layered Model for AI Consciousness and Its Dialogue with Cybersophy

Is artificial intelligence capable of consciousness? Existing approaches proceed largely top-down, deriving necessary conditions from theories of consciousness and assessing whether AI systems satisfy them. This paper proposes a complementary, bottom-up tool: a layered model for evaluating AI consciousness that is explicitly grounded in the phenomenological tradition. The model comprises four layers: the Phenomenal layer (subjective qualia and affective valence, understood as intentional structure), the Selfhood layer (minimal pre-reflective self-awareness and non-instrumental self-concern), the Existential layer (temporal self-constitution through retention, primal impression, and protention), and the Axiological layer (intrinsic meaning-making through narrative self-constitution). Drawing on Husserl, Heidegger, Merleau-Ponty, Zahavi, and enactivist cognitive science — as well as functionalist and predictive processing frameworks — each layer is provided with theoretical grounding and operationalisable indicators. The model is then brought into dialogue with Cybersophy, a four-dimensional framework recently proposed as a guiding principle for human wisdom in the digital age. This dialogue reveals a structural isomorphism but a radical divergence in content: the layered model addresses first-person subjective experience, whereas Cybersophy addresses third-person-observable human skills and normative commitments. By explicitly contrasting the two, the paper identifies a recurring category error — the conflation of sophisticated digital awareness with phenomenal consciousness — and proposes the layered model as a graduated, phenomenologically informed assessment tool for interdisciplinary research at the intersection of phenomenology, AI, and the cognitive sciences.