The proliferation of artificial intelligence (AI) systems in creative domains has generated substantial debate regarding the nature of authorship, creative agency, and the role of automation in human creative processes. This paper examines the theoretical and empirical landscape of human–AI co-creation, focusing on the tension between automation efficiency and creative complexity. Drawing on computational creativity9 research, human-computer interaction studies, and creative systems design literature, this work presents an automation versus creative-complexity framework to evaluate whether AI systems enhance or suppress human creative agency. Through analysis of mixed-initiative paradigms, enactive models of creativity, and empirical studies of co-creative systems, this paper argues that the distribution of creative control between human and machine significantly determines the quality and originality of collaborative outputs. The findings suggest that while automation can streamline executional tasks, preserving human interpretive authority and exploratory capacity is essential for maintaining creative complexity. This framework offers scholars and practitioners a lens for evaluating AI's role in co-creative environments and designing systems that augment rather than replace human creativity.