This paper introduces the Epiphany Model, a theoretical framework proposing that true conscious self-awareness in artificial intelligence might emerge not gradually through scaling, but as a discrete 'epiphany' moment when specific cognitive capabilities integrate. Drawing on theories from cognitive science and philosophy of mind, the paper examines prerequisites for machine consciousness, explores how agentic AI systems might develop self-awareness, and outlines an experimental framework for detecting or inducing such a transition. Ethical and alignment implications of potential machine consciousness are discussed, with recommendations for responsible research in this domain.