Generative AI and Operations: From Intelligence to Action at Scale

Generative and agentic AI systems are rapidly reshaping organizational workflows and norms, while reviving a classic operational challenge in a new form: how do we translate intelligence into reliable action at scale? This talk examines three emerging research frontiers on the path to autonomy.

First, generative AI can democratize optimization through language-to-decision workflows that support interactive modeling, rapid what-if analysis, and continual reoptimization. Second, whether these systems create real-world value depends on professional and organizational norms, from peer-reputation concerns in high-stakes settings to firm investments in workforce skills to the incentive distortions and moral hazard that can accompany enterprise deployment. Third, as systems become more agentic, reliability becomes the binding constraint. OR/OM contributes not just by solving decisions, but by designing the regimes that make autonomy stable, auditable, and safe under distribution shift and tail events.

Taken together, these frontiers position OR/OM as the discipline that both enables and governs autonomy at scale. The talk also highlights the mission of the new AI in Operations department at Manufacturing & Service Operations Management.