The rapid integration of artificial intelligence (AI) into educational contexts has initiated a profound reconfiguration of pedagogical theory, instructional practice, and the epistemic foundations of learning. Generative AI systems, particularly large language models such as ChatGPT, are no longer peripheral technological tools but have become active participants in knowledge production, assessment, and instructional mediation. This article develops a comprehensive, theoretically grounded analysis of AI-integrated pedagogy by synthesizing perspectives from motivation theory, ethical educational design, teacher professional knowledge, and learning ecology frameworks. Drawing strictly on contemporary and foundational literature, the study conceptualizes generative AI as a pedagogical actor that reshapes intrinsic motivation, learner autonomy, and instructional authority. A qualitative, integrative methodology is employed to examine how AI alters pedagogical relationships, assessment practices, and inclusivity across educational levels. The findings suggest that AI integration is not pedagogically neutral; rather, it redistributes epistemic agency, redefines teacher expertise, and challenges conventional notions of academic integrity and learner authenticity. Through deep theoretical elaboration, the article argues that effective AI pedagogy requires a shift from instrumental adoption toward ethically informed, motivation-sensitive, and epistemically transparent learning ecologies. The discussion highlights tensions between automation and human judgment, equity and access, and innovation and regulation. The article concludes by proposing a conceptual framework for responsible AI pedagogy that aligns intrinsic motivation, ethical governance, and professional teacher knowledge, offering implications for curriculum design, teacher education, and future research in AI-enhanced education.