The rise of generative artificial intelligence (GenAI) is driving a paradigm shift in human-AI interaction, creating new opportunities and challenges for Human-Centered AI (HCAI). This chapter establishes human-AI co-creation as a new interaction paradigm that moves beyond traditional input-output models to embrace turn-taking, improvisation, and social collaboration. This shift introduces complex interaction dynamics, such as emergent communication strategies and coordination, which current design and evaluation frameworks are not equipped to handle. To address this gap, this chapter proposes the cognitive science theory of enaction as a theoretical foundation for co-creative systems. Enaction frames social cognition as a participatory sense-making process where meaning emerges dynamically through interaction, offering a lens for designing co-creative partnerships that align with HCAI principles of user agency and collaboration. Grounded in this framework, the chapter provides a survey of the co-creative landscape, including its core dimensions, interaction types across different domains, evaluation methods, and concrete design recommendations. By identifying enaction as a guiding theory, this work provides a new conceptual toolkit for building the next generation of human-centered, collaborative AI systems.