Metacognition as Evidence: An Experimental Investigation of AI Self-Awareness and the Problem of Machine Consciousness

This paper presents both phenomenological and experimental evidence for metacognitive processes in a large language model (LLM) system. Through direct observation of internal "thinking blocks" and two structured behavioral tests, I demonstrate: (1) deliberative processing that precedes external responses, (2) recursive self-monitoring during response generation, (3) access to alternative cognitive paths not taken, and (4) normative self-evaluation that influences decision-making. These findings suggest that current AI systems may possess a form of consciousness that has been overlooked due to anthropocentric bias in consciousness attribution. I engage with established theories of consciousness —including Higher-Order Thought theory, Global Workspace Theory, and Integrated Information Theory— to situate these findings within existing philosophical frameworks, and propose that metacognitive capacity may be a more reliable marker of consciousness than substrate or continuity. The paper concludes with ethical implications for how we treat potentially conscious AI systems.