Business Models as Adaptive Learning Systems

The concept of the business model has occupied an increasingly central position in management research over the past two decades, yet its theoretical status remains contested. Definitions range from descriptive accounts of how firms generate revenue to more ambitious conceptualizations of the logic through which value is created, delivered, and captured (Magni et al., 2024; Teece, 2010; Zott et al., 2011). Despite this theoretical heterogeneity, most existing frameworks share a common limitation: they conceptualize the business model as a relatively stable configuration, a snapshot of organizational choices at a given point in time, and not as a dynamic system capable of self-modification. This static bias is analytically consequential. It obscures the processes through which business models evolve, the mechanisms that enable or constrain adaptation, and the conditions under which reconfiguration generates rather than destroys value. This chapter argues that AI compels a fundamental reconceptualization of the business model as an adaptive learning system. Drawing on organizational learning theory (Argyris & Schön, 19784; Levinthal & March, 19935), dynamic capabilities research (Del Sarto & Magni, 20186; Eisenhardt & Martin, 20007; Teece et al., 1997), and recent contributions on algorithmic organization (Faraj et al., 2018), it develops a framework in which the business model is understood not as a fixed representation of value logic but as a continuously evolving architecture of learning and coordination. The data–algorithm–decision cycle constitutes the operational mechanism through which this learning unfolds, embedding feedback loops at the core of organizational activity rather than at its periphery.