Controlled autonomy will define AI-enabled disruption management

An Institute for Supply Management roundup examines how agentic AI, real-time analytics and warehouse automation could reshape supply-chain disruption management. It cites Gartner’s prediction that 60% of supply-chain disruptions could be resolved without human intervention by 2031.

The more operationally relevant message is one of controlled adoption. Gartner recommends beginning with low-risk decisions, strengthening data quality and governance, assessing the impact of autonomy on supply-chain roles and maintaining contingency procedures for rapid human intervention when automated decisions fail.

The article also places the forecast in context. Prologis research indicates that automation is present in roughly 30% of logistics space, while fully automated warehouses remain uncommon at an estimated 3% to 5%. It also notes that tariffs and geopolitical volatility are pushing retailers toward regional, diversified and multihub sourcing models.

In his commentary, Manas Verma argues that the headline should not be interpreted as a declaration of full automation. He positions AI as a process-improvement layer that handles repetitive, lower-risk decisions while experienced operators remain responsible for supplier relationships, geopolitical risk and contingency execution.

Operational perspective: autonomous disruption management requires more than a model detecting an event. Organizations must define which decisions may be automated, what data supports them, who owns the constraints and guardrails, and when a person must intervene. Every autonomous action should be logged, reversible where possible and measured against service, cost and resilience outcomes.

Early deployment should focus on bounded decisions with reliable data and limited failure consequences. More consequential decisions—such as changing suppliers, reallocating scarce inventory or modifying production plans—need approval rules, escalation paths and reconciled execution through ERP, APS, TMS or control-tower systems.