Abstract
This paper develops an integrative conceptual framework to explain how natural language coding (NLC), as an AI-enabled information systems infrastructure, reshapes participation and co-creation in digital innovation ecosystems, and under what governance conditions it can support equitable, diverse, and inclusive outcomes. Although prior research has examined digital innovation ecosystems, AI, equity, diversity and inclusion (EDI), and information systems governance, these streams remain theoretically fragmented. Ecosystem research highlights distributed agency and generativity but often treats digital infrastructures as neutral enablers rather than participation-shaping architectures. Generative AI and NLC studies emphasize abstraction and productivity gains, yet rarely address how innovation agency is redistributed across actors. EDI scholarship clarifies that inclusion depends on structural conditions rather than mere access, but it is seldom integrated with ecosystem and socio-technical perspectives. Finally, AI and IT governance research focuses on accountability and transparency, without fully theorizing governance as a determinant of EDI co-creation in AI-mediated ecosystems. To address these gaps, the study adopts a conceptual theory-building design based on an integrative literature review across four domains: ecosystems and co-creation, AI-enabled information systems and NLC, EDI and digital inequality, and IS/AI governance. Through abductive reasoning and cross-stream synthesis, the paper proposes a multi-level framework distinguishing three layers: (1) interface-level access, (2) organizational redistribution of roles, and (3) infrastructural and institutional governance.
The results show that NLC does not automatically democratize innovation. While abstraction reduces technical entry barriers and expands participation, meaningful EDI outcomes depend on structured governance. At the interface level, controlled abstraction broadens access. At the architectural level, separating data design from workflow design distributes procedural authority and mitigates knowledge asymmetries. At the institutional level, profile-based governance and embedded validation mechanisms encode differentiated decision rights, transparency, and accountability into the innovation process. Together, these mechanisms translate participation affordances into what the paper terms “governed generativity.” The study concludes that equitable, diverse, and EDI AI-enabled ecosystems emerge not from openness alone, but from governance structures that condition how contributions are authorized, evaluated, and rewarded. The framework is intended to guide future empirical research through a sequential mixed-methods program using a platform user database of 700+ users. A qualitative study will examine how and why NLC reshapes participation architectures and EDI dynamics, surfacing mechanisms and boundary conditions. A subsequent quantitative phase will test the proposed relationships at scale (e.g., surveys and SEM), linking NLC-enabled participation and governance mechanisms (access rights, transparency, accountability) to EDI co-creation outcomes across contexts.