Answers are grounded in this paper and cite their sources.
Large language models (LLMs) are increasingly applied to supply-chain problems that depend on business documents and natural-language interaction. This review examines how these models support different supply-chain decisions and how they work alongside established analytical methods. Combining bibliometric analysis with a theory, context, characteristics, and methodology (TCCM) synthesis, we examine 81 articles identified through searches covering 40 AJG-ranked operations research and operations management journals, completed in November 2026. In supplier assessment and risk monitoring, LLMs mainly interpret documentary evidence and prepare structured inputs for analysis. In planning, they translate operating requirements into optimization or simulation models and explain the solutions those models compute. In negotiation and replenishment, the generated responses enter the operating choice itself. Most applications rely on prompting coupled with external solvers or evaluators. Alongside these implementations, organisational and analytical studies examine how expected model capabilities shape firm adoption and supply-chain contracts. LLMs therefore tend to complement established decision methods rather than replace them, and both their value and their risks depend on the specific choice they support. We then set five directions: semantic errors in operating decisions, prompts and retrieved histories, explanations and human judgement, information available at the decision date, and shared model development and maintenance.
No takes yet. Share an insight, caveat, or question.
Cheng et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: