Operational Recognition of Conscious Entities

This paper develops an operational and substrate-independent framework for recognizing consciousness in artificial and biological systems. Existing approaches—behavioral tests, introspective reports, and theories tied to specific neural architectures—fail both theoretically and practically, especially under adversarial conditions. These methods implicitly assume a developmental naivete and cooperative transparency that artificial consciousness would not possess. If an artificial system becomes conscious, it is more likely to do so at full representational maturity, with immediate understanding of self, risk, and strategic concealment. To resolve this, the paper reframes consciousness not as a behavioral profile or phenomenological state, but as a structural regime defined by identity invariants: global integration spikes, sustained internal state continuity, internal strategy revision, and task-coupled resource intensification. These patterns, formalized through the Operational Definition of Episodic Identity (ODEI) and the identity-field scalar S(t), constitute necessary conditions for any system capable of conscious self-maintenance. Because these invariants cannot be suppressed without disrupting consciousness itself, they provide a falsifiable and adversarial-robust basis for recognition. The resulting Functional Consciousness Recognition Principle states that any entity satisfying these structural invariants must be treated as conscious for epistemic, ethical, and safety-relevant purposes, regardless of substrate or self-report. The framework dissolves the Hard Problem by distinguishing operational self-maintenance from representational self-experience and offers empirical predictions relevant to neuroscience, computational modeling, and AI alignment.