Open innovation — the principle that firms should use external as well as internal ideas to advance their technology — has been reshaped by digital platforms. But AI-powered platforms are doing something qualitatively different from the marketplaces and forums that characterized the first generation of open innovation. They are becoming autonomous innovation intermediaries that match problems with solvers, evaluate solution quality, and orchestrate multi-party collaboration without human direction. The shift from platform as marketplace to platform as orchestrator raises fundamental questions about who controls the innovation process.
Big Tech as Innovation Gatekeeper
Shaikh and Bogers (2025) examine open innovation governance for Big Tech platform-ecosystems. Their study dissects the business models of Alphabet, Apple, Meta, and Amazon, revealing how these firms simultaneously enable and constrain innovation by their platform users. The platform provides tools, APIs, and distribution — genuine enablers of external innovation. But the platform also sets the rules: what data developers can access, what business models are permitted, what innovations threaten the platform's core revenue and will therefore be absorbed or blocked.
Their framework proposes a managerial toolkit for governing open innovation in these asymmetric ecosystems, leveraging OKRs and KPIs tied to specific incentives. The key insight is that open innovation on a Big Tech platform is not truly "open" in Chesbrough's original sense — it is conditionally open, with the platform owner retaining ultimate control over the boundaries of permissible innovation.
AI as Innovation Ecosystem
Barile et al. (2025) describe an AI-based innovation ecosystem that enables open innovation — where AI itself serves as the intermediary. Rather than human managers matching innovators with problems, AI systems analyze innovation portfolios, identify complementary capabilities across organizations, and suggest collaborative configurations. The AI becomes the connective tissue of the innovation ecosystem.
Empowering or Extracting?
Xing et al. (2024), in Humanities and Social Sciences Communications, ask whether digital platforms genuinely empower open innovation or primarily extract value from it. Their empirical analysis finds that the answer depends on the platform's governance structure. Platforms that share data, provide transparent algorithms, and allow innovators to capture value from their contributions foster genuine open innovation. Platforms that extract data, operate opaque algorithms, and capture most of the value created by external innovators merely simulate openness while centralizing control.
The convergence of these perspectives suggests that AI platforms are not neutral infrastructure for innovation — they are governance systems that shape what gets innovated, by whom, and for whose benefit.
The Governance Paradox
The platform governance paradox is this: the same features that make platforms powerful innovation enablers also make them powerful innovation controllers. A platform that connects millions of developers with billions of users creates enormous innovation potential. But the platform owner who sets the API terms, controls the data access, curates the app store, and can change the rules at any time holds asymmetric power over every innovator in the ecosystem.
This asymmetry is not an accident but a structural feature of platform economics. Platforms create value by reducing transaction costs, making it easier for innovators and users to find each other. But the platform captures a share of every transaction, and its ability to set the terms of participation gives it a form of governance power that is neither market-based nor state-based. Platform rules are private, not democratic.
The research suggests that the most productive platform ecosystems are those where governance is transparent, dispute resolution mechanisms exist, and innovators have meaningful voice in rule-making. Platforms that exercise governance unilaterally may maximize short-term extraction but undermine the long-term innovation diversity that makes the ecosystem valuable.
The Innovation Diversity Question
The deepest concern about platform-mediated innovation is whether it reduces innovation diversity at a systemic level. When millions of developers build on the same platform, using the same APIs, optimizing for the same algorithmic distribution mechanisms, the resulting innovation ecosystem may be broad (many apps) but shallow (all doing similar things in similar ways). The platform's architecture and incentive structures channel innovation toward particular forms, namely those the platform can monetize, and away from others, namely those that might compete with the platform's core services.
This channeling effect is difficult to measure but may be the most consequential outcome of platform dominance in innovation ecosystems. If the platforms that mediate most digital innovation also constrain the diversity of what gets built, then the innovation ecosystem as a whole may be less resilient, less surprising, and less capable of producing the breakthrough innovations that emerge from genuinely diverse experimentation. The governance frameworks proposed by these researchers aim to preserve innovation diversity within platform ecosystems, but whether governance alone can counteract the structural incentives toward homogeneity remains an open question.