Generative AI platforms as institutional catalysts of digital entrepreneurship: Enablement, dependence & power dynamics

This study theorizes how recent generative AI (Gen AI) platforms operate as institutional catalysts of platform-dependent entrepreneurship (PDE). Integrating institutional theory, the external enablement framework, and innovation platform theory, we propose an integrative framework for explaining the emergence of PDE under reliance on non-substitutable, platform-governed capabilities. Using an abductive mixed-method case study of OpenAI’s ecosystem (2020–2025), we trace how governance, boundary resources, and institutional signals shape entrepreneurial feasibility, scaling, and vulnerability. Our analysis identifies four catalytic institutional mechanisms—Infrastructure Provision, Capability Scaffolding, Market Legitimization, and Ecosystem Orchestration—that enable venture creation while simultaneously generating dependence. Temporal analysis reveals an enablement–dependence paradox: platforms accelerate entry by democratizing frontier capabilities, yet accumulate dependencies that expose ventures to governance shocks. Empirically, we show how enthusiasm gave way to crisis during OpenAI’s GPT-5 release, illustrating governance overreach, trust erosion cascades, and choice removal as control. We conclude with theoretical and practical implications. • This study explores how generative AI platforms as institutional catalysts • Institutional catalysts enable platform-dependent entrepreneurship (PDE) • The paper draws on institutions, external enablement and platform innovation • An embedded case study approach is adopted to analyze 20 PDE • PDE emerges through resource access, conservation, and compression mechanisms