Advancing net-zero supply chain management: a machine learning driven review aligned with the sustainable development goals

Net-zero supply chain management (NZSCM) is gaining traction as a strategic response to climate challenges. This study uses the TCCM framework and BERTopic modelling to map its theoretical, contextual, and methodological evolution. Key theories include life cycle assessment, institutional theory, and game theory, but gaps remain in applying AI and stakeholder theory. The sectoral focus is narrow–mainly on construction, energy, and maritime–with weak cross-sector learning and uneven global adoption. Six themes emerge–the circular economy, renewable energy, carbon capture, digital transformation, resilience, and multistakeholder governance–aligned with SDGs 7, 9, 12, and 13. Major barriers include scalability issues, behavioural inertia, and limited automation. Conceptual and quantitative methods dominate, with minimal use of mixed or experimental approaches. Future research should examine digital twins for real-time forecasting, behavioural–policy interactions, and finance–driven adaptations. Advancing NZSCM requires hybrid LCA models, stakeholder engagement, and experimental validation to support equitable and data-informed decarbonisation pathways.