Review the recent literature on platform ecosystem governance. Identify the major theoretical perspectives, summarize areas of agreement and disagreement, and explain how the field has evolved over the past five years.

1. Major theoretical perspectives

Recent research treats platform ecosystem governance less as a single set of platform-owner rules than as a multidimensional process involving coordination, control, participation, and institutional legitimacy. Four perspectives are particularly prominent.

Governance as ecosystem design and indirect coordination. This perspective examines how platform owners shape complementors’ behavior through openness, access rules, standards, incentives, monitoring, and positioning rather than through direct managerial authority. Inoue’s analysis of 9,780 Japanese video-game products shows that openness and platform distinctiveness influence whether complementors pursue incremental or radical innovation, suggesting that governance affects ecosystem-level “ambidexterity” indirectly through the conditions under which complementors compete and innovate [1]. Related work finds that governance design also produces interdependent effects among complementors: greater openness can increase autonomy while simultaneously increasing ecosystem complexity and the costs of customization, thereby reducing complementors’ willingness to multihome [2]. This literature therefore conceptualizes governance as an architectural and economic design problem, with ecosystem performance depending on the balance between participation and control.

Governance as relational, networked, and co-produced. A second perspective challenges the assumption that one keystone organization necessarily owns and governs the ecosystem. Studies of European data spaces and inter-platform ecosystems examine shared ownership, bilateral coordination, cross-platform complementarity, and distributed decision rights [3][4]. Costabile, Iden, and Bygstad [5] show that common technical and organizational standards can emerge through institutional work even when no clear platform leader exists. Similarly, research on network-driven ecosystems conceptualizes governance as a process through which actors jointly establish ownership and decision rights rather than simply responding to rules imposed by a central platform [3]. This perspective emphasizes negotiation, interdependence, and collective institution-building.

Governance as institutional and political power. A third perspective foregrounds asymmetry. Platform owners do not merely coordinate exchanges; they control infrastructures, data, technical interfaces, visibility, and access to users. Recent studies distinguish economic, technological, and regulatory sources of platform power and argue that these sources reinforce one another through feedback loops [6]. Research on Google’s relationship with Samsung likewise illustrates how control over a higher-level platform can create entry barriers and constrain attempts to establish competing or subordinate platforms [7]. From this viewpoint, governance is inseparable from dependence, contestation, and regulation. The relevant question is not only whether governance improves ecosystem efficiency, but also who defines the rules, who bears their costs, and whether affected actors can challenge them.

Governance as adaptive sociotechnical practice. A fourth perspective understands governance as continually revised in response to changing technologies, actor practices, institutional pressures, and unintended consequences. Volz et al. [8] conceptualize data governance as an adaptive control loop in which principles emerge from the reconciliation of competing institutional logics rather than from a fixed hierarchy of commands. Research on platform governance in informal markets similarly finds that governance emerges through iterative adjustment between digital changes to platform architecture and nondigital relational arrangements [9]. This approach is extended by studies of generative AI, which argue that AI changes the boundary resources, affordances, and power relations through which platforms govern participation [10]. Governance is consequently treated as an ongoing process of tuning rather than a one-time institutional design choice.

2. Areas of agreement and disagreement

The literature broadly agrees that governance is constitutive of ecosystem performance. Platforms do not simply provide neutral infrastructure: their rules and technical architectures shape participation, knowledge exchange, innovation, and value distribution. A systematic review of 111 studies identifies infrastructure, governance, and generativity as mutually connected platform-enabling structures supporting successive phases of knowledge collaboration—aggregation, coordination, and emergence [11]. This finding converges with empirical work showing that monitoring, reputation mechanisms, and IT support can reduce complementor slackness and encourage value co-creation [12]. Across these studies, participation is understood as institutionally and technically enabled rather than as an automatic consequence of platform access.

A second point of agreement is that effective governance requires balancing apparently opposing objectives. Openness can attract complementors and stimulate innovation, but it can also increase complexity, coordination costs, security risks, and opportunities for platform dependence. Zander et al. describe ecosystem evolution as driven by moderate levels of tension between collaboration and competition, the core and periphery, value creation and value capture, and global and local interests [13]. Research on blockchain ecosystems reaches a comparable conclusion from a different theoretical starting point: governance failures arise when system-level rules are insufficiently aligned with actor-level needs, especially across the trade-offs between consistency and flexibility, system reliance and actor reliance, and ecosystem utility and member utility [14]. The shared implication is that governance effectiveness depends on calibrated tensions rather than on maximizing a single value such as openness, control, or decentralization.

The literature also agrees that governance is increasingly multi-level. Platform rules interact with technical standards, market incentives, organizational routines, public regulation, and user practices. Studies of the EU Digital Markets Act and platform regulation show that contestation between platform owners and complementors is mediated by regulators rather than confined to bilateral platform relationships [15]. Research on private AI standards similarly demonstrates that corporate and civil-society initiatives interact with state authority through competing pathways of resistance, engagement, and leadership [16]. Governance therefore operates simultaneously within platforms, across ecosystems, and through public institutions.

Important disagreements remain. The first concerns centralization versus distributed governance. Platform-design research often assumes that a platform owner can strategically configure openness, incentives, standards, and monitoring. In contrast, research on Mastodon, European data spaces, and leaderless standardization emphasizes federated, shared, or negotiated authority [17][3][5]. The disagreement is partly conceptual: centralized models treat governance as a capability of an orchestrator, whereas distributed models treat it as an emergent property of interorganizational relations. The empirical literature does not support a universal superiority of either model; centralized governance may reduce coordination costs, while distributed governance may improve representation and adaptability but create slower or more contested decision processes.

The second disagreement concerns formal rules versus situated adaptation. Governance-by-design studies emphasize codified rules, technical standards, monitoring, and incentive structures. Adaptive and practice-oriented studies argue that rules are incomplete because actors reinterpret, evade, or modify them in use. Evidence from Kuaishou, for example, portrays governance as a negotiation among the state, the platform, and migrant users rather than as a simple transmission of regulatory commands [18]. Similarly, the evolution of DTube from community moderation toward semi-automated governance suggests that technological architecture alone does not determine governance outcomes; user practices and changing community conditions remain decisive [19].

The third disagreement concerns efficiency, innovation, and legitimacy as the primary evaluative standard. Ecosystem-management research tends to evaluate governance through innovation, sales, participation, or value co-creation. Critical and political research asks whether governance is accountable, equitable, and contestable. Haggart and Iglesias Keller’s framework distinguishes input, throughput, and output legitimacy, arguing that many platform-governance proposals emphasize procedural administration while underexamining participation and democratic accountability [20]. This difference matters because a governance arrangement can improve coordination or platform growth while simultaneously intensifying power asymmetries or excluding less powerful actors.

3. How the field has evolved since 2021

From 2021 onward, the field has moved through three related shifts. The first was a move from platform governance as owner control toward ecosystem governance as indirect coordination. Early work in the period concentrated on how openness, distinctiveness, and governance design influence complementor innovation and multihoming [1][2]. The platform remained the central unit of analysis, but governance was increasingly understood as shaping the behavior of interdependent actors rather than simply enforcing platform policies.

The second shift, especially visible from 2022 to 2024, was toward distributed authority, institutional work, and political legitimacy. Studies examined how standards are built without a dominant leader, how alternative platforms organize membership and moderation, and how regulation emerges through domestic politics and transnational constraints [5][17][21]. Platform governance consequently became connected to questions of democracy, public accountability, state–platform relations, and the institutional conditions under which rules acquire legitimacy. The field also expanded beyond commercial app stores and social media to include education, blockchain, data spaces, and other sectoral ecosystems [22][14].

The third shift, most evident in 2025–2026, is toward dynamic, multi-actor, and AI-mediated governance. Recent research increasingly treats governance as adaptive and recursive: data-governance principles emerge through continuous adjustment, platform power is reproduced through interacting economic, technological, and regulatory mechanisms, and ecosystem ownership may be shared across networks [8][6][3]. Generative AI has accelerated this development because it introduces open-ended and difficult-to-predict outputs, alters boundary resources, and requires complementors to govern prompts, context, user inputs, and model outputs [23]. AI is therefore studied not only as an object requiring governance but also as an instrument that changes how platforms coordinate actors and produce value. Barile and Secundo [24] identify intelligent stakeholder matching, dynamic coordination, knowledge recombination, and scalability as mechanisms through which AI-based platforms may become active ecosystem orchestrators.

Overall, the literature has evolved from a relatively firm-centered view of governance toward a more plural account in which platform owners, complementors, users, regulators, technologies, and standards jointly shape the rules of participation. The central theoretical problem has consequently changed. Earlier work primarily asked how platforms should govern complementors to generate innovation and value; recent work asks how governance can remain adaptive, legitimate, and contestable while platforms accumulate infrastructural and algorithmic power. The selected literature has made substantial progress on these questions, but it remains more developed in conceptual frameworks and qualitative case studies than in longitudinal comparative evidence capable of showing when particular governance arrangements produce durable ecosystem performance across sectors.