Impact pathway: beyond automation and augmentation – a relational ontology of AI in operations and supply chain management

Abstract

Purpose As companies deploy increasingly agentic forms of AI to coordinate and manage workflows, AI can no longer be understood as a discrete technological tool added to existing operations and supply chains. We propose a relational ontology as an alternative theoretical foundation, conceptualising human–AI coordination as an emerging configuration that continuously reconstitutes roles, processes and performance criteria in OSCM practice. Design/methodology/approach We analyze recent practitioner reports on AI in OSCM (2023–2026) to assess how far the discrete-tool paradigm dominates discourse and to identify where its limitations become visible. This grounds our proposed impact pathways. Findings We show that practitioner discourse is systematically founded on the assumption that AI is a discrete tool. Using well-known AI use cases, we demonstrate that this paradigm cannot account for observed shifts in roles, processes and performance criteria. Research limitations/implications Adopting a relational lens, we develop three impact pathways that reframe common empirical puzzles – unexpected role drift, unstable process boundaries and contested performance attribution – as objects of study rather than implementation failures. Originality/value We challenge a foundational and largely unexamined assumption in OSCM research – that AI is a discrete tool deployed onto existing operations – and offer a relational alternative. This shifts the unit of analysis from the technology to the evolving human–AI configuration, helping managers anticipate how AI deployment will reshape jobs, workflows and performance metrics.