ERP, WMS, TMS, OMS, and planning systems remain essential. But AI is introducing a new layer in supply chain technology: systems that evaluate conditions continuously, incorporate context, weigh tradeoffs, and support or initiate action.

From Systems of Record to Systems of Decision

Supply chain technology has evolved in layers.

The first layer was built around transaction integrity. Orders had to be captured. Inventory had to be recorded. Shipments had to be tendered. Labor had to be scheduled. Invoices had to be matched. Financial and operational records had to reconcile.

This was the era of systems of record.

ERP, warehouse management, transportation management, order management, procurement, and related enterprise systems gave supply chains a durable transactional backbone. They remain essential. No AI architecture can replace the need for accurate orders, inventory positions, receipts, shipments, invoices, and master data.

The second layer extended this foundation into planning. Demand planning, supply planning, inventory optimization, network design, transportation planning, and scenario modeling helped companies move beyond recording what happened toward preparing for what might happen.

Those capabilities also remain essential.

But a third layer is now emerging.

AI is introducing systems of decision.

This new layer does not replace systems of record or systems of planning. It operates across them. It evaluates changing conditions, incorporates context, weighs tradeoffs, and supports or initiates action. It is less concerned with storing transactions than with improving decisions that affect cost, service, inventory, capacity, and execution.

For a deeper look at how AI is moving from architecture to operational execution, download the full ARC Advisory Group white paper: AI in the Supply Chain: From Architecture to Execution.

Systems of Record Still Matter

There is a temptation in AI discussions to talk as if legacy systems are obsolete. That is wrong.

Systems of record remain the foundation of supply chain execution. A warehouse cannot operate on probabilistic inventory. A transportation team cannot tender loads against uncertain shipment records. A finance organization cannot settle invoices against ambiguous transactions. A customer service team cannot make reliable commitments if order status is not accurate.

The core enterprise systems preserve operational truth.

But they were not designed to resolve every decision problem. They are very good at capturing and executing structured transactions. They are less effective at deciding what should happen when conditions change across multiple functions at once.

A supplier misses a commitment. A vessel is delayed. A key SKU is running below safety stock. A customer places an unexpected order. A transportation lane tightens. A facility loses capacity.

The record may show the event.

The decision is something else.

Planning Helps, But the Plan Keeps Changing

Planning systems were designed to help companies make better forward-looking decisions. They improved forecasting, inventory policy, capacity planning, allocation, network modeling, and supply-demand balancing.

But planning has historically been periodic. Monthly. Weekly. Sometimes daily. Even when planning systems use sophisticated optimization, the plan often becomes stale as execution begins.

That is not a failure of planning. It is a function of the operating environment.

Demand shifts faster than planning cycles. Carrier capacity changes faster than procurement processes. Supplier reliability changes faster than static lead-time assumptions. Disruptions can invalidate a plan before it is fully executed.

The supply chain does not need planning less. It needs planning to become more connected to execution.

This is where systems of decision become important.

What a System of Decision Does

A system of decision does not merely report what happened. It helps determine what should happen next.

It may consume data from ERP, TMS, WMS, OMS, planning systems, supplier portals, visibility platforms, risk feeds, and customer systems. It may use machine learning, optimization, business rules, retrieval-augmented generation, graph reasoning, or agentic workflows. But its purpose is not technology for its own sake.

Its purpose is to improve decisions.

A system of decision may support questions such as:

Which late shipments create real customer or production risk?

Which supplier disruption requires action versus monitoring?

Which orders should receive constrained inventory?

Which loads should be expedited, consolidated, delayed, or rerouted?

Which alternate suppliers are operationally feasible, not merely theoretically available?

Which customer commitments should be revised?

Which exception should be escalated to a planner, and which can be resolved automatically?

These are not simple reporting questions. They require context, judgment, constraints, and execution linkage.

The Decision Layer Cuts Across Functions

The reason systems of decision matter is that many important supply chain decisions are cross-functional.

A transportation delay is not only a transportation issue. It may affect inventory, customer service, warehouse scheduling, production sequencing, procurement, and finance.

A supplier disruption is not only a procurement issue. It may affect manufacturing, fulfillment, substitution rules, customer commitments, working capital, and risk exposure.

A demand spike is not only a planning issue. It may affect allocation, replenishment, labor, freight capacity, production capacity, and customer prioritization.

Traditional systems tend to see the problem through functional lenses. A decision system must evaluate the broader operating consequence.

This is one reason AI has strategic relevance. AI can help connect signals across systems, identify relationships, evaluate tradeoffs, and surface recommended actions faster than manual coordination can typically support.

The goal is not to remove human judgment. The goal is to reduce decision latency.

Decision Latency Is the Real Constraint

Most large supply chains already have more data than they can use effectively.

They have orders, shipments, inventory positions, forecasts, carrier events, supplier records, risk alerts, customer commitments, and exception reports. The problem is not always lack of visibility. Increasingly, the problem is the time required to convert visibility into coordinated action.

A shipment delay is detected. Transportation sees the issue. Inventory planning checks exposure. Procurement considers alternatives. Customer service updates expectations. Finance evaluates cost. Operations weighs feasibility.

Each function may respond rationally from its own position. But the response is often sequential, fragmented, and slow.

That is decision latency.

AI’s value is not simply faster analysis. Its higher value is reducing the time between signal, judgment, and execution.

A system of decision is useful only if it shortens that gap.

Not Every AI System Belongs in the Decision Layer

As AI moves closer to execution, the stakes change.

A chatbot that summarizes policy documents is one thing. A system that changes a transportation route, reallocates inventory, recommends a supplier switch, or revises a customer commitment is something else.

The closer AI operates to financial or physical consequence, the greater the requirement for determinism, context, governance, and auditability.

A planning recommendation can be reviewed and adjusted. A warehouse movement, routing change, purchase order, supplier substitution, or customer commitment carries immediate consequence. In those environments, probabilistic output must be constrained by rules, thresholds, approval paths, and domain-specific validation.

This is why supply chain AI should not be treated as a single category.

Different decision environments require different levels of autonomy, oversight, explainability, and control. A low-risk recommendation may be suitable for automation. A high-impact decision may require human approval. A regulated or customer-sensitive decision may require audit trails, access controls, and documented rationale.

The suitability of AI depends on domain, consequence, and governance.

What Changes for Technology Buyers

The emergence of systems of decision changes how buyers should evaluate supply chain technology.

The traditional questions remain useful: what function the system supports, what workflows it automates, what integrations it offers, what data it manages, and what reports it produces.

But those questions are no longer sufficient.

Buyers need to ask a second set of questions:

What decisions does the system improve?

Which roles are involved in those decisions?

What data and context are required?

How does the system evaluate tradeoffs?

Does it recommend action, initiate action, or simply report conditions?

What execution systems does it connect to?

What approval thresholds are configurable?

How are outcomes measured?

How are overrides captured?

Can the decision logic be audited?

This shifts evaluation from software functionality to operational impact.

A system that improves a dashboard may be useful. A system that improves a decision that affects service, inventory, capacity, or cost is more valuable.

What Changes for Vendors

This shift also changes the market structure for supply chain software vendors.

Planning vendors, transportation platforms, warehouse systems, visibility providers, procurement platforms, risk intelligence firms, and enterprise software companies are all embedding AI into their offerings. Their starting points differ, but the direction is similar.

They are moving toward decision support, decision automation, or decision orchestration.

This creates overlap between software categories that were once more distinct. A visibility provider may move into exception resolution. A planning vendor may move closer to execution. A TMS vendor may embed real-time decision support. A procurement platform may incorporate supplier risk intelligence and autonomous sourcing recommendations. An ERP vendor may position its AI layer as the enterprise decision fabric.

The market will not be defined only by functional labels. It will increasingly be defined by decision environments: procurement and commercial orchestration, network planning and resilience, logistics and fulfillment execution, exception management, inventory allocation, supplier risk response, customer commitment management, and planning-execution synchronization.

These are not merely software categories. They are operating problems.

Why AI Programs Stall

Many AI programs stall not because the technology is weak, but because the organization is not prepared to absorb it.

Common failure modes include AI insights that are not connected to execution systems, data that is available but not decision-ready, recommendations that are not trusted, unclear decision ownership, governance introduced too late, and workflows that remain manual after the AI output is generated.

In these cases, the enterprise may have AI capability without operational change.

That distinction matters.

The value is not in producing a better recommendation in isolation. The value is in changing the decision process in a way that improves cost, service, resilience, inventory, or speed.

The most successful organizations will not be those that deploy the most AI features. They will be those that redesign decision workflows around AI-supported execution.

Conclusion: The New Layer of Supply Chain Technology

Supply chain technology is not moving away from systems of record. It is building on them.

ERP, WMS, TMS, OMS, procurement, planning, and visibility systems remain essential. They provide the transactional and operational foundation that supply chains require.

But AI is creating a new layer above and across these systems.

That layer is focused on decisions.

It connects signals, context, reasoning, governance, and execution. It helps organizations move from knowing what happened to deciding what should happen next. It reduces decision latency. It supports coordination across functions. It creates the possibility of more adaptive, resilient, and responsive supply chains.

The next competitive advantage in supply chain technology will not come from better dashboards alone.

It will come from better decisions, connected to execution.

That is the shift from systems of record to systems of decision.

The post From Systems of Record to Systems of Decision: How AI Is Changing Supply Chain Technology appeared first on Logistics Viewpoints.

Global supply chains do not run only on ships, ports, warehouses, and trucks. They also run on data. Undersea cables are becoming part of the same infrastructure risk conversation as canals, straits, pipelines, power grids, cloud platforms, and payment networks.

Undersea Cables Are Supply Chain Infrastructure

For most of modern logistics history, the word “chokepoint” meant a physical place.

The Strait of Hormuz. The Suez Canal. The Panama Canal. The Strait of Malacca. A congested port. A rail corridor. A border crossing. A bridge.

That definition is now too narrow.

Global trade also depends on digital chokepoints. These are less visible than ports and canals, but they are increasingly central to the movement of goods, money, documents, instructions, and commitments. Beneath the ocean floor, submarine fiber-optic cables carry the data layer of the global economy. They support financial transactions, cloud computing, customs documentation, logistics visibility, port systems, carrier communications, manufacturing coordination, and the routine exchange of commercial information that allows supply chains to function.

The recent discussion by Iranian-linked media about fees, permits, and potential control over undersea internet cables passing through the Strait of Hormuz is a useful reminder of this shift. The Strait of Hormuz has long been understood as an energy and maritime chokepoint. The newer concern is that the same geography may also become a digital pressure point.

That does not mean a disruption is imminent. It does mean supply chain leaders need to broaden how they think about infrastructure.

The supply chain is no longer only physical. It is physical, financial, digital, and computational at the same time.

The Digital Layer of Trade

Modern supply chains require continuous information flows.

A container move depends on booking data, customs filings, bills of lading, port community systems, carrier status updates, bank payments, purchase orders, warehouse instructions, customer notifications, and inventory commitments. A disruption in physical movement is obvious. A disruption in digital movement can be less visible at first but can rapidly affect execution.

If transportation management systems cannot receive status updates, visibility degrades. If customs platforms slow down, cargo can be delayed. If payment networks are disrupted, commercial settlement becomes uncertain. If cloud services or data routes become unstable, companies may lose access to systems that manage planning, fulfillment, sourcing, and customer communication.

This is why undersea cables should be understood as supply chain infrastructure.

They are not peripheral telecommunications assets. They are part of the operating environment for global logistics.

Hormuz as a Digital Chokepoint

The Strait of Hormuz is already central to global energy flows. Its role in oil and gas markets is well understood. What is receiving more attention now is the overlap between energy routes, maritime routes, and data routes.

The operating significance is not whether a particular proposal becomes formal policy. The significance is that undersea cables are being discussed in the same strategic vocabulary historically applied to oil tankers, naval transit, and regional trade.

That is the change.

Digital infrastructure is now part of geopolitical bargaining.

A country does not need to stop container vessels to create supply chain pressure. It can threaten energy flows, interfere with port systems, disrupt payment channels, target cloud infrastructure, or place legal and operational pressure on communications networks. The practical effect can be similar: greater uncertainty, higher risk premiums, slower execution, and reduced confidence in the reliability of trade lanes.

This matters because supply chains increasingly depend on near-real-time information. Visibility platforms, transportation management systems, supplier portals, customs systems, warehouse systems, and customer service applications all assume that the data layer will remain available.

That assumption deserves more scrutiny.

Why This Matters to Supply Chain Executives

Most supply chain risk programs are still built around familiar categories: supplier failure, port congestion, natural disasters, labor disruption, geopolitical conflict, cyberattack, inventory shortages, and transportation capacity.

Those categories remain valid. But they do not fully capture the infrastructure dependencies now embedded in supply chain operations.

The modern supply chain depends on several connected infrastructure layers:

Physical infrastructure: ports, roads, rail, warehouses, airports, canals, ships, and trucks

Energy infrastructure: fuel, electricity, LNG, refining, and grid stability

Digital communications infrastructure: undersea cables, terrestrial fiber, satellite backup, and telecom networks

Computational infrastructure: cloud platforms, data centers, AI systems, and enterprise applications

Financial infrastructure: payments, trade finance, insurance, credit, and settlement systems

A shock in one layer can cascade into others.

A maritime conflict may raise fuel prices and delay cargo. It may also affect cable security, cloud access, payment confidence, insurance pricing, and carrier risk calculations. A cyberattack may begin in software but interrupt physical operations. A data center disruption may affect inventory planning, customer service, and freight execution.

Supply chain resilience therefore cannot be limited to inventory buffers and alternate suppliers. It must include digital continuity.

Visibility Platforms Depend on Invisible Infrastructure

There is irony in the current technology environment. Supply chain visibility platforms are sold on the promise of knowing where everything is. But the platforms themselves depend on infrastructure that is mostly invisible to users.

Container tracking, predictive ETAs, supplier portals, warehouse dashboards, and transportation control towers all depend on the movement of data. That data often crosses national boundaries, cloud regions, telecom networks, and undersea routes before appearing as a dot on a screen.

When those communications pathways are stable, they disappear into the background. When they are threatened, the enterprise discovers that visibility is not simply a software capability. It is an infrastructure dependency.

This becomes more important as supply chains become more AI-enabled. AI systems need real-time signals, external context, transaction histories, exception data, and access to enterprise systems. The more supply chain decision-making depends on continuous data access, the more exposed it becomes to communications infrastructure risk.

AI does not reduce infrastructure dependency. In many cases, it increases it.

A supply chain that uses AI for demand sensing, dynamic routing, supplier risk monitoring, customs documentation, and customer service automation may be more responsive than a traditional supply chain. But it may also become more dependent on data availability, system interoperability, cloud access, and secure communications.

That does not argue against AI. It argues for a more complete resilience model.

The New Infrastructure Questions

For years, companies asked whether their suppliers were dual-sourced, whether their ports had alternatives, whether their carriers had capacity, and whether their inventory policies were resilient.

Those questions still matter.

But new questions are emerging:

What digital infrastructure supports our most critical supply chain workflows?

Which cloud, telecom, cable, and data exchange dependencies are embedded in our operations?

Do key logistics, planning, and visibility systems have regional redundancy?

Which workflows fail if real-time data is degraded?

Can we operate in a limited-connectivity mode?

Are escalation procedures defined for digital infrastructure disruption?

Do supplier portals, customer portals, and carrier integrations remain usable under degraded conditions?

These are not traditional supply chain questions. But they are becoming operationally relevant.

The executive issue is not whether a supply chain manager should become a telecom engineer. The issue is whether the organization understands the dependencies that support its ability to plan, execute, communicate, and recover.

Digital Chokepoints Behave Differently

Digital chokepoints are not identical to physical chokepoints.

A blocked canal is visible. A damaged bridge has a location. A closed port has a queue. A data route may degrade in more complex ways. Traffic may reroute. Latency may increase. Systems may remain partially available. Some applications may function while others fail. The business impact may depend on architecture, redundancy, vendor configuration, cloud region, access rights, cybersecurity posture, and contractual service levels.

This makes digital infrastructure risk harder to see and harder to assign.

It can sit between IT, supply chain, risk management, procurement, legal, and finance. Everyone may own part of it. No one may own the full operating consequence.

That is the governance gap.

A modern supply chain resilience program should identify which digital services are mission-critical, who owns their continuity, how disruptions are escalated, and which manual or alternate processes can sustain operations when systems degrade.

Resilience Under Degradation

The answer is not to build a fully redundant version of every system. That is unrealistic.

The better approach is to tier workflows by operational criticality.

Some workflows can tolerate delay. Some cannot. A weekly analytics report can wait. A customs filing, shipment release, carrier tender, customer commitment, or production signal may not.

Supply chain leaders should work with IT and enterprise risk teams to classify critical workflows, map system dependencies, and define continuity requirements. This includes not only core enterprise applications, but also third-party logistics platforms, visibility providers, supplier portals, carrier networks, payment systems, and external data sources.

The practical goal is resilience under degradation, not perfect immunity.

Can the enterprise still prioritize shipments? Can it still communicate with carriers? Can it still release orders? Can it still issue customer updates? Can it still make inventory allocation decisions? Can it still comply with regulatory requirements?

If not, the organization has a digital infrastructure exposure.

Conclusion: The Supply Chain Runs on Data

The supply chain has always depended on infrastructure. What has changed is the definition of infrastructure.

Ports and ships still matter. So do roads, railroads, warehouses, canals, and aircraft. But the supply chain also runs on fiber-optic cables, cloud platforms, data centers, payment networks, cybersecurity systems, and enterprise software.

Undersea cables are a reminder that the digital economy is not weightless. It has physical routes, landing points, repair constraints, ownership structures, jurisdictional exposure, and geopolitical risk.

For supply chain leaders, the lesson is clear.

Digital infrastructure is now supply chain infrastructure.

The companies that understand this will build more complete resilience programs. The companies that do not may discover, during the next disruption, that their physical network can still move goods, but their digital network cannot support the decisions required to move them wel

The post Why Undersea Internet Cables Matter to Global Supply Chains appeared first on Logistics Viewpoints.

Ocean freight forwarding is an $80+ billion market bogged down by the manual processes related to booking management, documentation services, and the coordination labor that holds it all together.

When working with a freight forwarder, you’re buying three things bundled together:

Carrier relationships — access to capacity, negotiated rates, allocation commitments.
Operational data — knowing which carrier fits a given lane, what documents a particular trade corridor requires, how to handle an exception when a booking gets rejected.
Coordination labor — the booking itself, the documents per container (industry estimates range from 9 to 18 depending on the corridor), the re-keying of data across disconnected systems, the email chains chasing confirmations and clearances.

Shippers have always paid for the bundle because you couldn’t get one piece without the others, but that’s changing.

Where the bundle comes apart

Travel agents used to bundle airline relationships, destination expertise, and the labor of putting trips together into a single fee. Aggregator platforms unbundled the pieces, and the booking layer went first because that’s where the volume was. Ocean freight forwarding is in the same position. More than digitizing booking, though, AI is automating it.

The bulk of the volume and labor cost for freight forwarders is tied up in rate comparisons across dozens of carriers, document preparation and routing by trade lane and commodity classification, booking execution against pre-negotiated contracts, and exception triage on rejected bookings.

But this is all high-volume, rule-governed, multi-system coordination where speed and consistency matter more than creativity. Exactly the type of work that AI agents are well-equipped to handle.

Platforms can now ingest a rate agreement, parse surcharges and FAK provisions into a digital rate profile, compare carriers on cost, transit time, and schedule reliability, and execute a booking based on pre-defined parameters, without a human in the loop.

Automating the entire order lifecycle

Every dollar of margin exposure in ocean freight traces back to a decision made without complete information. That means that every action must be rooted in live network data across shipment flows, carrier performance, and insight from inventory and order systems. A platform with that intelligence can automate and accelerate the full workflow from detecting a supply shortfall, selecting a carrier, booking the container, managing the documents, tracking the shipment, and handling exceptions.

A shipper stitching together a rate tool from one vendor, a booking portal from another, a document system from a third, and a visibility feed from a fourth gets digitization. They get a slightly faster version of the same manual process. The full picture still lives in a person’s head, and the handoffs between systems still require human coordination.

While freight forwarders and other intermediaries are also investing in AI, they’re primarily automating their own coordination labor before someone else absorbs it. But they can’t replicate the data advantage of a platform that sits across the entire supply chain.

A forwarder automating its booking desk draws on its own transaction history. A point solution built specifically for ocean booking draws on booking data. A platform processing millions of supply chain events daily across orders, inventory, carrier performance, and live shipment status, has a different signal base entirely. Carrier selection informed by real-time schedule reliability, live network disruption, and your actual inventory positions is structurally more accurate than carrier selection informed by historical rate tables.

The shrinking intermediary layer

The moats around freight forwarders’ profit margins are eroding, and the lines between legacy endpoint solutions are blurring. High-complexity corridors and specialized commodities still need human expertise, but the bread-and-butter containerized freight that makes up the bulk of forwarder revenue is the volume where automated workflows shine.

Meanwhile, software providers will have a hard time selling dashboards and chatbots to specific teams compared to AI-native platforms offering a single operating system across all supply chain operations, and serving downstream stakeholders.

The question for forwarders is how long they can keep patching automation onto a fragmented architecture with a booking tool here, a document system there, people bridging the handoffs in between. And how much revenue sits in structured, repeatable work that a connected platform absorbs?

For shippers, the choice is whether to invest in a platform that automates the order-to-delivery and exception lifecycle, or keep paying others to hold the pieces together. The second option is a decision to fund the intermediary layer sitting between them and their own data.

The post The Freight Forwarder Moat Is Getting Shallower appeared first on Logistics Viewpoints.