Artificial intelligence (AI) is increasingly applied in Supply Chain Management to improve efficiency, decision-making, and operational performance. While these capabilities are often associated with sustainability benefits, the evaluation of sustainability outcomes remains fragmented and inconsistent across the literature. Existing studies rely on heterogeneous indicators, proxy measures, and process-specific assessments, which limit comparability and obscure trade-offs between environmental, operational, economic, and social objectives. As a result, the sustainability impact of AI-driven initiatives is often inferred rather than explicitly assessed. Based on a structured literature review and conceptual framework development approach, this study argues that sustainability in AI-enabled supply chains is mis-evaluated due to the absence of structured, trade-off-aware, and multi-level evaluation approaches. In response, it proposes an integrative framework that links AI technologies, supply chain processes, and sustainability outcomes through a standardised measurement layer. The framework incorporates trade-off evaluation, multi-level interactions, and cross-cutting enablers, enabling consistent and transparent assessment of sustainability performance. By shifting the focus from technological capability to evaluative consistency, the framework provides a basis for understanding how AI-driven decisions translate into measurable sustainability outcomes and how trade-offs emerge across competing objectives. The study contributes by addressing the fragmentation and limited comparability of existing evaluation approaches in AI-enabled supply chains. In addition, it offers a coherent structure supporting research, managerial decision-making, and policy-oriented sustainability evaluation.
Daios et al. (2026) studied this question.