Agentic AI in Supply Chain Planning and Procurement: The Roles Employers Need Before Adoption

Agentic AI takes supply chain automation a significant step further. 

Where generative AI can analyse information, produce content or recommend an action, AI agents can potentially pursue objectives, coordinate workflows and take actions across systems with a greater degree of autonomy. 

For supply chain planning teams, that could mean agents monitoring demand signals, adjusting forecasts or triggering inventory and production updates. In procurement, agents could monitor suppliers, support sourcing events, manage routine transactions or initiate action when defined risk thresholds are breached. 

Deloitte describes agentic AI in supply chain management as enabling a move from task automation towards outcome delegation, with humans increasingly shifting from execution towards orchestration and oversight. 

That makes agentic AI a workforce question as much as a technology question. 

Before employers ask what an AI agent can do, they should ask whether they have the planning and procurement professionals capable of deciding what it should do, where its authority ends, and when a human needs to intervene. 

Agentic AI changes the risk as well as the opportunity

The distinction between traditional automation and agentic AI matters. 

A conventional system might identify a potential inventory shortage and alert a planner. An AI agent could potentially analyse the shortage, identify available inventory elsewhere in the network, recommend or execute a reallocation, and trigger subsequent planning updates. 

In procurement, an AI tool might summarise supplier data. An agentic system could go further by assessing supplier performance, identifying risk and initiating a sourcing or remediation workflow. 

Deloitte identifies potential supply chain applications ranging from demand analysis and dynamic inventory management to supplier evaluation, sourcing, logistics and proactive constraint management. 

That potential also raises the consequences of getting implementation wrong. 

An inaccurate AI-generated summary is one problem. An autonomous action based on incomplete data, poor business rules, or inappropriate permissions is another. This is why talent strategy needs to begin before deployment. 

Adoption should start with people, processes and governance

The pressure to move quickly is growing, but supply chain leaders should be cautious about equating speed with readiness. Gartner has warned supply chain planning leaders about "agent washing", where technology is marketed as agentic without delivering meaningful autonomous capabilities. Instead, they recommend that supply chain planning leaders concentrate on proven use cases, stronger data and governance foundations, and a deliberate sequence of adoption rather than responding to market hype. 

Procurement leaders face a similar challenge. 

A survey of chief procurement officers found that only 36% were very confident in their ability to redesign roles and processes around AI. Gartner argues that organisations need to separate human work from AI-native tasks and redesign procurement roles accordingly. 

The lesson for employers is straightforward: adopting agentic AI without redesigning the organisation around it risks automating an operating model that was never built for autonomous decision-making. 

So, which people need to be in the room before adoption? 

1. Planning leaders who can define where autonomy belongs 

Planning is one of the clearest opportunities for agentic AI. 

Agents can potentially monitor demand continuously, identify exceptions, and connect changes in forecasts to inventory, production, and replenishment decisions. IBM, for example, identifies demand forecasting and planning as a major application, with agents using historical and real-time signals to update forecasts and inform downstream plans.  

But increasing autonomy makes experienced planning leadership more important, not less. Businesses need professionals capable of defining which decisions can be automated safely and which require escalation, such as:  

• Should an agent automatically rebalance inventory between locations? 
• Can it change the production plan? 
• What happens when the mathematically optimal decision conflicts with a commercial commitment to a strategic customer? 
• Where does an exception become significant enough for human intervention? 

Those are supply chain questions, not purely technical ones. 

Our USA Supply Chain Talent Report 2026 found that planning professionals were particularly likely to recognise this changing environment. 29% selected AI literacy as the capability they expect to be most important for their role over the next three years, higher than any other skill option among planning respondents. 

The Europe Supply Chain Talent Report 2026 shows a similar pattern, with planning professionals again most likely to select AI literacy, at 24%. 

Employers preparing for agentic planning therefore need leaders who combine deep planning expertise with sufficient AI literacy to determine where autonomy creates value and where it introduces unacceptable risk.

2. Procurement leaders who can redesign work around AI

Procurement could experience an even more significant operating-model shift. Deloitte's research into multi-agent AI in sourcing and procurement describes a move from AI generating content or recommendations towards systems capable of executing governed, multi-step workflows across sourcing, contracting, compliance and supplier interactions. 

The implications for procurement roles are considerable. If an agent can undertake parts of supplier analysis, sourcing administration and procure-to-pay activity, employers need to decide where procurement professionals can add greater value. 

That could increase the importance of category strategy, supplier relationships, complex negotiation, commercial judgement and risk management. 

McKinsey estimates that technology could make procurement organisations 25% to 40% more efficient, with time previously spent on transactional work redirected towards strategy.  

But efficiency does not appear automatically when software is introduced. Gartner's finding that only 36% of CPOs are very confident redesigning their functions around AI demonstrates why leadership capability needs to precede adoption. 

Employers need procurement leaders who can redesign roles, clarify decision rights, and determine where human commercial judgment remains indispensable. 

3. Supplier and category specialists with strong relationship skills 

Agentic AI may become highly capable at analysing suppliers, but that does not make supplier relationships less important. DSJ Global's 2026 research actually suggests the opposite. 

Procurement professionals were the European respondents most concerned about AI's impact on their roles, with 36% saying they were very or somewhat concerned. But the wider report concludes that experienced supply chain professionals continue to create value through judgement, relationships and navigating ambiguity. 

Christine Corson, Managing Director at DSJ Global USA, describes how the work division is changing in the US report: 

She adds that when AI handles more processing and analysis, professionals can spend more time building relationships, influencing stakeholders, and developing broader business strategy. 

That becomes highly relevant in procurement. An agent may identify a deteriorating supplier-risk profile. A category or supplier manager still needs to understand the commercial relationship, assess whether the data reflects reality, negotiate corrective action, and determine whether changing suppliers creates greater risk elsewhere. 

Agentic AI can change the information available to those professionals. It does not remove the need for them.

4. Data and systems expertise that understands the supply chain

An AI agent is only as useful as the environment in which it operates. 

For planning and procurement, that environment can include ERP platforms, planning systems, procurement software, supplier information, contracts, inventory records, demand signals, logistics data, and financial systems. 

Poor data can therefore become an operational problem once agents are authorised to act on it. Data readiness and governance are foundational to trust in agentic supply chains. IT and operational technology alignment are therefore important requirements for scaling. 

Employers may consequently need people who can bridge technical and functional teams. 

The strongest profiles will not necessarily be pure AI specialists. A planning systems leader who understands master data, forecasting processes and ERP integration may be more useful during implementation than someone with advanced AI knowledge but limited understanding of the supply chain. 

The same applies in procurement, where systems expertise needs to sit alongside an understanding of sourcing processes, supplier data, contracts, approvals, and commercial controls. 

5. AI governance and risk owners 

Autonomy requires accountability. Before deploying agents, organizations need to establish who is responsible for their actions, who can change their permissions, what agents are authorised to access, and when a decision needs human intervention. 

The NIST AI Risk Management Framework provides a useful foundation. It places accountability, transparency, and human oversight within the wider management of AI risk, while NIST's current work on agentic AI focuses specifically on areas including governance, security, evaluation, and risk management. 

For employers, this does not necessarily mean creating a new role titled "AI Governance Manager." It does mean assigning clear responsibility for how agents operate. 

Depending on the organisation, that capability could sit across supply chain leadership, procurement, planning, technology, risk, legal, or compliance teams. 

The principle is simple: an AI agent can be given autonomy, but accountability still belongs to people. 

6. Change leaders who can redesign how teams work

Agentic AI adoption changes workflows, decision rights, performance measures, and potentially the responsibilities attached to individual jobs, which makes change management a core capability. 

Matthew Wood, Managing Director at DSJ Global Europe, argues that the market is already changing in structure rather than simply shrinking. DSJ Global's European research recommends that employers focus on how roles are evolving and invest in capabilities such as adaptability and change management. 

The talent implications should be considered early. A business should ask itself:  

• If an AI agent takes responsibility for work previously completed by a planner or buyer, what becomes that person's new remit? 
• How should their performance be measured? 
• Which capabilities need developing? 
• Who trains teams to challenge an agent rather than automatically accept its recommendation? 
• How does the organisation prevent valuable functional knowledge from disappearing as processes become automated? 

These questions need answers before adoption reaches scale. 

The pre-adoption team employers should consider

There is no universal agentic AI team structure. The right model will depend on the organisation, technology, level of autonomy, and use case. 

But for planning and procurement, employers should assess whether they have access to capabilities such as: 

• Supply chain planning leaders, who determine decision rights, exceptions, and planning priorities. 
• Demand, supply, and S&OP/IBP specialists, to translate operational planning logic into agent-enabled workflows. 
• Procurement and strategic sourcing leaders, to redesign processes around human and AI responsibilities. 
• Category and supplier management specialists, to retain commercial judgment and relationship ownership. 
• Planning systems and procurement technology specialists, to connect AI with existing systems and workflows. 
• Data and analytics professionals, to address data quality, architecture and decision intelligence. 
• Risk, compliance and governance expertise, to define controls, permissions, escalation and accountability. 
• Transformation and change leaders, to redesign roles and support workforce adoption. 

Not every capability requires a permanent new hire. Some organisations will need permanent leadership, others contract specialists for implementation, and businesses undergoing larger transformation may need several hires across functions. 

The priority is identifying the capability gap before technology decisions create an urgent hiring requirement. 

Do not wait for implementation to expose the skills gap

Agentic AI is developing quickly, but the case for immediate full autonomy is much less clear.  

Gartner's warning about agent washing in supply chain planning is particularly relevant. Organisations need to distinguish genuine autonomous capability from technology marketing, select viable use cases, and build the foundations required for responsible adoption. 

The same principle applies to talent. Waiting until implementation begins to discover that nobody owns AI governance, supplier data is unreliable, planning leaders cannot explain decision boundaries, or procurement teams lack the capability to redesign workflows will slow transformation. 

The stronger approach is to map people requirements alongside the technology roadmap.

Agentic AI needs human capability before autonomous capability

The attraction of agentic AI is understandable. Supply chains manage enormous volumes of information across suppliers, customers, production facilities, inventory, logistics networks, and external risks. Technology capable of monitoring that environment and acting could significantly change how planning and procurement operate. 

But autonomous technology does not remove the need for human expertise. It changes where that expertise sits. 

DSJ Global's 2026 US research found that planning professionals and supply chain leaders are already placing significant value on AI literacy, while professionals also highlighted vendor relationships, adaptability and business acumen as important capabilities. 

The European findings tell a similar story. Only 30% of respondents were very or somewhat concerned about AI replacing their role, while the report highlights judgement, relationships and the ability to navigate ambiguity as areas where experienced professionals continue to create value. 

For employers, the priority before agentic AI adoption is to build the human capability required to govern the autonomous capability. 

Build your agentic AI talent strategy before deployment

Agentic AI can change how planning and procurement teams operate, but technology alone will not determine whether that change creates value. 

The people defining decision rights, governing AI actions, managing suppliers, challenging outputs, connecting systems and leading organisational change will matter just as much as the agents themselves. 

DSJ Global connects organisations with business-critical professionals across Planning, Procurement, Supply Chain Leadership, Engineering, Commercial Solutions and Technical Operations, with permanent, contract and multi-hire solutions suited to different stages of transformation. 

Preparing your planning or procurement function for agentic AI? Build the right team before deployment begins. 

Discuss your talent requirements with DSJ Global today.