As supply chains face escalating disruption, leading organizations are moving from AI assistance toward agentic AI operating models capable of acting autonomously. The IBM Institute for Business Value (IBV) study, conducted in partnership with Oracle and Accelalpha, describes how agentic AI models represent a structural shift in supply chain design, enabling systems to adapt dynamically to real-world events with minimal human intervention.
Agentic AI operating models combine ERP data, supply chain applications, ecosystem partner interfaces, and external data sources to drive continuous optimization across procurement, inventory, production, logistics, and service operations.
From Insight to Autonomous Action
Unlike traditional automation, agentic AI operating models are designed to interpret signals and take action in real time. Autonomous agents can reroute shipments, adjust sourcing strategies, negotiate with suppliers, and mitigate risk as conditions change, without waiting for manual approvals.
The study finds that 62% of surveyed supply chain leaders believe embedded AI agents accelerate speed to action, while 76% of chief supply chain officers say process efficiency will improve as agents take on repetitive, impact-based tasks faster than people can. By 2026, 57% of executive respondents expect agentic AI to proactively recommend actions, and 62% expect it to make automation and workflow reinvention more effective.
These models rely on significantly broader data inputs, including weather, geopolitical events, market indexes, and partner data, allowing agents to simulate scenarios and anticipate disruptions before they materialize.
Ecosystem-Scale Resilience
Agentic AI operating models can extend beyond the enterprise. Agent-to-agent interactions across supplier and logistics networks allow organizations to coordinate responses at ecosystem scale. This capability enables more resilient planning, faster recovery from shocks, and shared innovation across partners.
The study highlights dynamic sourcing in procurement as an early area of strong adoption, with agents adjusting supplier selection based on demand signals, pricing shifts, and capacity constraints. Similar approaches apply to inventory optimization, production yield forecasting, and transportation routing.
Visualization and simulation are central to the agentic AI model. AI-enabled virtual representations allow agents to test scenarios, evaluate trade-offs, and generate mitigation plans that can be executed immediately when thresholds are met.
Balancing Autonomy, Governance
While autonomy accelerates performance, executives in the study also cite data accuracy, bias, security, and transparency as major challenges. Of the respondents, 72% point to concerns around data accuracy or bias, and 63% cite data security and privacy as barriers to generative AI in supply chain operations.
As a result, people remain essential to orchestrating agentic AI outcomes. Employees are expected to monitor agent performance, set optimization goals, and adjust autonomy levels as needed. Governance, observability, and accountability are positioned as core requirements, not afterthoughts.
The study emphasizes that organizations achieving top-tier performance are redesigning operating models, not simply deploying tools. Success with agentic AI depends less on technology adoption and more on organizational readiness. Leaders moving the fastest are pairing autonomy with clear accountability, investing early in visibility, testing, and human oversight to ensure AI-driven actions remain aligned with business and compliance expectations.
What This Means for ERP Insiders
Agentic AI points to a new supply chain operating model. Autonomous agents move ERP beyond tracking what has already happened and toward actively adjusting plans and actions as conditions change. This shift alters how planning, sourcing, logistics, and execution are designed, with systems taking on more responsibility for routine decisions.
Stronger connections across partners improve resilience. The IBM research shows that when AI agents can share information and coordinate actions across suppliers, logistics providers, and other partners, supply chain teams respond faster to disruption. This makes integration and data-sharing across ecosystems increasingly important for maintaining performance during uncertainty.
Governance is essential as systems gain more independence. As AI takes on more decision-making, organizations need clear oversight, defined roles, and visibility into how actions are taken. The research suggests long-term value comes from balancing automation with accountability, ensuring autonomy improves outcomes without introducing new operational or compliance risks.
