{{item.videoDuration}}
{{item.title}}
{{item.text}}
{{item.videoDuration}}
{{item.text}}
Key takeaways:
- Supply chain planning is shifting from periodic, calendar-driven processes to continuous, decision-driven capabilities that sense change faster, evaluate trade-offs, and respond as conditions evolve.
- AI is taking planning beyond dashboards and recommendations toward governed agentic execution, where systems can increasingly resolve defined exceptions and act within established decision boundaries.
- Planning is expanding beyond enterprise boundaries as organizations connect more deeply across supplier and partner networks, improving visibility into both emerging disruptions and growth opportunities.
- AI-enabled planning requires more than technology. Operating models, decision rights, and planner roles should evolve so organizations can translate faster insights into better cross-functional decisions.
- Data readiness and governance determine how quickly autonomy can scale, with trusted decision-critical data and clear accountability providing the foundation for AI-enabled execution.
- The path to advanced planning is progressive, moving from stronger visibility and integrated processes toward AI-assisted decisions and governed autonomy that can make supply chains more adaptable and growth-focused.
After weathering multiple disruptions in recent years, companies across industries are competing to build the next generation of supply chains. Besides increasing resilience, the winners will be adaptable, autonomous, and growth-focused, re-architecting toward networks that are more diversified and digitally capable rather than retreating from global trade.1 It starts with planning—the organizational capacity to sense change faster, decide with better information, and execute with less friction.2
The opportunity is clear. Only 23% of supply chain organizations have a formal AI strategy,3 with most just running disconnected experiments and calling it transformation. By 2028, 60% of digital adoption efforts will fall short of expected value, not because the technology fails but because organizations underinvest in the people and processes needed to make it work.4 The opportunity is also central to the intelligent enterprise, where strategy, technology, operations, and governance work as one integrated system, with AI helping turn information into coordinated action.
Success in this environment hinges on understanding how planning maturity translates into outcomes, where AI can create durable value, and what leaders should prioritize—not just to manage risk but to capture the growth possible in this moment.
The three biggest changes in planning as a capability
A shift is underway across three dimensions: how frequently decisions are made, how much decision-making machines can assume, and how far beyond the enterprise the information behind those decisions can extend. These changes build on the shift toward closed-loop planning and ecosystem orchestration that we previously identified as core to the next generation of supply chains.1
From periodic cycles to continuous, closed-loop decision-making
For decades, planning ran on fixed cycles: monthly S&OP reviews, weekly supply runs, quarterly demand refreshes. That logic may no longer hold. Closed-loop planning and execution harness data to continuously plan, execute, monitor, and adapt in real time.1 The bigger change is what that does to decision-making. Decisions that once waited for the next planning cycle now can be made as conditions dictate.
Exceptions become central. Rather than batching deviations into a weekly review, mature environments resolve them as they surface, anchored by a single source of truth for both data and the assumptions behind each decision.5 Disruptions that once took months to play out can now ripple across global networks in days.
Agentic planning: Autonomy in narrow, governed tasks
“Autonomous” has been applied to supply chains loosely for years, and it now raises a more precise operational question. Which planning decisions should machines be allowed to make? Agentic AI is moving from pilot into production, and Gartner projects SCM software with agentic AI capabilities will grow from under $2 billion in 2025 to $53 billion by 2030, with 60% of enterprises adopting agentic features by then, up from 5% today.6 The investment trajectory may be clear but the governance infrastructure to support it is not.
Successful early deployments share one trait: They start narrow. An agent can detect a shipment delay, check alternative carrier availability, reroute the order, notify the customer, and escalate to a human only when the exception falls outside its approved scope. This is a different category of operational infrastructure, one that compresses the time between detecting a problem and resolving it for a single, well-bounded decision.
The risk is mistaking agentic planning for progress. Speed without the right decision architecture just means faster mistakes. Without strong data foundations, clear decision rights, and operating models redesigned for machine-led execution, agentic workflows can accelerate activity without improving decision quality.7 Done well, the payoff is both avoided risk and more confident growth decisions (which SKUs to expand, which capacity bets to make, etc.) made with better information and less lag.
Organizations should treat agentic AI as a near-term discipline, picking two or three exception-handling workflows that are technically feasible and commercially safe, building governance around them, and scaling from there.
Planning beyond the enterprise: Extending decision-making across the network
The third shift changes the boundary of planning itself through ecosystem orchestration by establishing mutually beneficial partnerships with external parties built on continuous data sharing and consensus decision-making.1 For most of the past two decades, planning systems were built around what an organization could see and control, things like its own inventory positions, production capacity, and Tier 1 supplier commitments.
That perimeter is no longer sufficient. The risks that interrupt operations increasingly originate from where planning systems don’t reach. A 2026 report by Supply Chain 24/7 shows that visibility drops off quickly after Tier 1: Only 12% of organizations say they have visibility into more than half of their Tier 2 suppliers, and Tier 3 is still largely out of reach.8
The compliance environment is now reinforcing what the operational case already demanded. Labor prevention acts and sustainability regulations require visibility well beyond Tier 1 suppliers. But extended visibility is only half the equation. Ecosystem orchestration also surfaces growth signals earlier like shifting demand, emerging capacity, and new partner capabilities that a Tier-1-only view would miss entirely.
Bringing more signals into the planning environment creates value only if the internal structures are designed to act on them, which means rethinking decision rights, oversight structures, and the role of the planner.
Considerations: Operating model, data, and governance
While AI can expand planning capabilities, organizational readiness determines whether that value is realized. Trusted data, clear decision rights, and connected workflows matter more than model sophistication, and until organizations build those foundations, AI will improve visibility faster than decision quality.9
Operating model redesign and the planner's evolving role
In many cases, planning remains organized around functional silos, and those silos now span more than planning alone. Demand, supply, procurement, and finance often operate on different clocks and through different systems, which makes a planning recommendation hard to validate against margin impact or execution feasibility. Decisions move stepwise across those silos, held together by meetings and escalation paths. AI can change that. It helps surface and evaluate trade-offs across the network in real time. Instead of being constrained by process flow, planning is governed by decision flow.
Operating structures built around functional ownership can’t absorb decision-making that cuts across those boundaries. Organizations see more, faster, but still struggle to act. While nearly all organizations expect to move toward more horizontal, networked models, far fewer operate that way today.10
Planning organizations should move from owning processes to owning decisions, explicitly defining which decisions are automated, which are augmented, and which remain human-led. Some decisions should be fast because speed creates value there. Others need to stay slow because that’s where judgment matters more.
That redesign changes the planner’s job as much as the org chart. Planners become stewards of decision quality rather than process coordinators. Legacy planners were integrators, pulling data and building plans manually. Now the system integrates and the planner interprets, focusing on what the system can’t do, including evaluating trade-offs, prioritizing the exceptions that matter, and bringing commercial context into decisions. Most organizations aren’t there yet, but left unresolved, this creates a new failure mode. Planning gets faster but weaker, and decisions are made quickly with less understanding behind them.
Data readiness should determine the pace of autonomy
AI readiness in planning is often a data question before it is a model question. Decision intelligence and agentic planning fall apart conflicting hierarchies, fragmented ownership, and inconsistent master data long before algorithmic sophistication becomes the constraint. Decision-ready data is more important than larger data lakes. Companies that strengthen their foundations as they increase AI use see nearly twice the improvement in AI-driven performance compared with those scaling on weaker foundations.11
87% of operations leaders say poor data quality has hampered their ability to achieve value for digital initiatives, and only 30% report significant improvement over the past two to three years.
Source: PwC’s 2026 Digital Trends in Operations SurveyLeading organizations improve targeted, decision-critical data iteratively rather than waiting for perfect enterprise-wide data quality. Planning leaders should ask whether the organization has:
- Consistent product, customer, and location hierarchies across functions
- Current and governed assumptions for lead times, service levels, and constraints
- Cost data to make financial tradeoff decisions
- Clear ownership for the data used in high-impact planning decisions
- The ability to move data between systems without manual rework or re-entry
Governance and responsible AI
Effective governance requires clear decision rights, defined escalation paths, and accountability measured differently for system-driven versus human-led decisions. Where this isn’t defined, behavior fills the gap: Planners override outputs, spreadsheets come back, and the operating model fails.
Trust follows the same logic. If planners don’t understand how decisions are made, they won’t rely on them. If leaders can’t explain those decisions, they won’t defend them. Scaling AI requires clear accountability, transparency, and human oversight where stakes are high, and adoption depends on the ability to interrogate and challenge outputs, not just receive them.12 Without that, trust may erode and value can disappear.
The same standard applies to privacy. As planning moves from recommending to executing, who sees sensitive supplier, allocation, and cost data—and with what oversight—becomes an operational question. Only about one-third of operations leaders are comfortable assigning AI agents to execute full end-to-end processes,10 typically a sign that escalation logic hasn’t been built into the architecture.
At minimum, organizations should demonstrate:
- Testing of agent behavior before release
- Monitoring of decision boundaries and data access
- Defined escalation paths for high-risk decisions
- Clear ownership across business, technology, and risk teams
How you can get there
Many organizations may be able to define AI-enabled planning, but the challenge is building a practical path to reach it. This often involves a broader maturity curve—from spreadsheet-driven planning to integrated visibility to AI-assisted decision support to governed agentic execution and, ultimately, adaptable, autonomous, and growth-focused supply chains.1
In addition to a more advanced planning function, this is planning as part of an intelligent enterprise, connecting data, technology, people, and governance so decisions can translate into coordinated action. Consider this 90-day operating cadence for moving through the earliest stages of that maturity curve.
Organizations that scale AI successfully build “AI fitness” across the enterprise, and the most AI-fit companies achieve a 7.2x performance boost over peers.11
This roadmap helps compress the gap between intent and action, with planning becoming a driver of enterprise performance and competitive advantage.
Adriana Arroba Hurtado, Eric Miao, Jimmy He, Joelle Azouri, Mackenzie Snow, Richard Ward, Sarah Kopik, and Steve Puricelli also contributed to this perspective.
AI is redefining supply chain planning
From periodic plans to continuous decisions
{{filterContent.facetedTitle}}
{{contentList.loadingText}}
{{contentList.loadingText}}
Follow us
