Data Extractivism and Strategic Value Appropriation: Rethinking Firm Advantage in AI-Centric SME Ecosystems

Abstract

Artificial intelligence–enabled platforms are transforming the foundations of competitive advantage in digital market ecosystems. Small and medium-sized enterprises (SMEs) generate substantial transactional and behavioral data through platform participation, yet control over data aggregation and model-training architectures typically resides with platform sponsors. This structural decoupling challenges the core assumption of the resource-based view that ownership and control of valuable resources ensure rent appropriation. Integrating resource-based theory, value appropriation logic, data-enabled learning research, and platform governance scholarship, this article develops a conceptual framework explaining how data extractivism operates as an architecture-mediated mechanism of value capture. The model argues that competitive advantage in AI-centric ecosystems increasingly derives from control over aggregation infrastructures rather than localized data generation. Cross-SME data pooling produces compounding learning rents that disproportionately accrue to actors controlling centralized architectures, especially under conditions of high switching costs, limited data portability, and governance opacity. By reframing advantage as architecture-dependent, the study extends strategic management theory and clarifies how SME performance becomes ecosystem-conditioned in AI-driven markets.