Recursive nested innovation outcomes in AI-enabled learning: the role of capability renewal

This study examines how artificial intelligence (AI)-enabled learning generates heterogeneous innovation outcomes within organizations and how firms manage tensions among these outcomes through higher-order dynamic capabilities. Using a qualitative multi-case approach, this research investigates digitally intensive SMEs that have adopted AI in innovation activities. The findings show that AI adoption does not produce uniform innovation effects; instead, it generates three coexisting, interacting innovation outcomes within a recursive, nested system: efficiency-seeking innovation outcomes, pseudo-agility innovation outcomes, and transformative innovation outcomes. Efficiency-seeking innovation outcomes emerge when AI is primarily utilized for automation, standardization, and process optimization. Pseudo-agility innovation outcomes arise when AI enhances responsiveness and accelerates experimentation without sufficient integration of organizational learning processes. In contrast, transformative innovation outcomes develop when AI is embedded in more exploratory, strategically oriented organizational learning processes. The findings further reveal that the relationships among these innovation outcomes are recursive in nature. Strategic transformation remains influenced by deeply embedded efficiency logic, while transformative change simultaneously pressures operational efficiency logic. To manage these tensions, organizations need to develop higher-order capabilities, including optimization capability, adaptive coordination capability, and strategic renewal capability. This study contributes to the literature on AI-enabled learning, organizational learning, and dynamic capabilities by demonstrating that AI-driven innovation outcomes are not linear but emerge through recursive interactions among organizational logics, learning processes, and ongoing capability renewal.