Platform owners coordinate ecosystems mainly via incentives, control and boundary resources; the arrival of algorithmic governance and generative AI is now reshaping how participation, quality and innovation are governed.

Main Finding

The paper synthesizes platform governance literature via a systematic review (85 core papers coded from an initial 644 hits) and proposes an integrative framework that groups governance mechanisms into three core categories—Incentives, Control, and Boundary Resources. It argues that effective orchestration requires portfolios of these interrelated mechanisms and calls for research on how emerging technologies (algorithmic governance, generative AI, agentic AI) reshape governance practice and platform economics.

Key Points

• Three core governance categories:
  • Incentives — monetary (subsidies, pricing, revenue-sharing, access vs. usage fees, freemium, subscription, transaction fees) and non-monetary (visibility, promotion, exclusivity, marketing privileges).
  • Control — formal controls (input screening, registration, compliance checks), and other control modalities used to ensure quality and align participant behavior (paper emphasizes the need to balance openness with quality/risks).
  • Boundary Resources — technical and informational artifacts (APIs, SDKs, documentation, sandboxes, developer tools) that lower entry costs, enable integration, and foster third-party innovation.
• The authors derive 13 concrete governance mechanisms (from axial/selective coding) though organized into the three high-level sets above.
• Governance must be configured dynamically across platform lifecycle stages (e.g., onboarding vs. post-critical-mass management) and aligned with platform technology.
• Emerging technologies pose new governance challenges and opportunities:
  • Algorithmic governance (automated matching, algorithmic management) changes how rules are enforced and how transactions are coordinated.
  • Generative AI spawns new GenAI platform ecosystems around foundational models, altering complementor roles and business models.
  • Agentic/autonomous AI as ecosystem participants raises questions on rights, accountability, and orchestration of autonomous actors.
• Research gaps and agenda: impact of AI-driven governance, how governance portfolios co-evolve with technology and platform capabilities, how participants perceive and respond to governance, and the dynamic capabilities needed for adaptive governance.

Data & Methods

• Method: Systematic literature review following Webster & Watson (2002) and vom Brocke et al. (2015), treated literatures as qualitative data with iterative open/axial/selective coding (Bandara et al., Wolfswinkel et al.).
• Search:
  • Query focused on ("ecosystem" OR "platform") AND ("governance").
  • Targeted leading IS journals (VHB-JOURQUAL B to A+) and major IS conferences (ICIS, ECIS, WI, HICSS, PACIS, AMCIS).
  • Databases used included AISeL and Scopus (per outlet searching).
• Screening & coding:
  • Initial hits: 644
  • After title/abstract screening and exclusion criteria: 190
  • Full-text screening produced 56 directly relevant articles; backward search added 29; selected commentaries added due to foundational influence → 85 final papers coded.
  • Analysis yielded 13 governance mechanisms aggregated into the three categories.
• Supplementary materials: coding matrices and appendices available at the authors’ OSF link (https://doi.org/10.17605/OSF.IO/2M3W4).

Implications for AI Economics

• Market design and pricing:
  • Algorithmic governance and automated pricing/matching change platform market dynamics (speed of price adjustments, personalized pricing, dynamic subsidization), affecting welfare, distribution of surplus, and potential for strategic exploitation by platform owners.
  • Generative AI platforms introduce new complement markets (models, prompt services, fine-tunes). Economic models must account for multi-sidedness where the “product” is a model, data, compute, or downstream outputs.
• Competition and market power:
  • AI-enabled boundary resources (proprietary APIs, SDKs, model access) create strong lock-in and winner-takes-most dynamics; the governance choices about openness vs. control materially affect entry barriers and competition.
  • Revenue-sharing and access restrictions influence complementor incentives; AI-driven discoverability and promotion algorithms can amplify or attenuate winner-take-all effects.
• Transaction costs, information asymmetry, and externalities:
  • Algorithmic enforcement reduces some transaction costs (screening, matching) but may increase information asymmetries (opaque ranking/recommendation decisions) and negative externalities (bias propagation, misuse of generative outputs).
  • Agentic AI participants add complexity to property-rights frameworks and liability models—who captures value from autonomous agents, and how are rights/prices assigned when agents transact?
• Welfare, fairness, and regulation:
  • Automated governance raises fairness concerns (discrimination, opaque enforcement), needing econometric and normative models to evaluate distributional outcomes and regulatory interventions.
  • Platform governance choices (e.g., revenue split, promotion algorithms) should be examined for their distributional impacts across users, complementors, and the platform owner.
• Research directions for AI economics:
  • Extend theoretical models of platform competition to include algorithmic governance rules (automated matching, promotion) and AI-created complements.
  • Empirical studies measuring how AI-driven governance changes pricing, entry, survival of complementors, and consumer surplus—using field data, natural experiments, or platform collaborations.
  • Laboratory and field experiments on perception and behavioral responses to algorithmic governance (trust, participation, pricing acceptability).
  • Models of dynamic capability investments: how much should platforms invest in governance automation vs. human oversight, and what are the returns under varying competitive environments?
  • Study of agentic actors: formalize ownership, incentive, and liability structures when autonomous agents participate economically on platforms.
• Practical policy implications:
  • Regulators and platform policymakers should consider how governance automation shifts competitive balance and worker/user protections; transparency, auditability, and contestability of algorithmic governance merit central attention.

If you want, I can (a) extract the 13 specific mechanisms the authors list and map each to economic questions, or (b) draft a short research proposal (3–4 research questions + methods) focused specifically on how generative AI changes platform revenue-sharing and market structure. Which would you prefer?