The experimentation phase of AI in Procurement is ending.

Twelve months ago, the focus was possibility. Today, the focus is performance.

Execution is no longer about acceleration. It is about sequencing, and sequencing demands disciplined leadership judgement.

Across CASME RoundTables and benchmarking forums, procurement leaders are no longer asking what AI could do. They are asking: 

Are we structurally and commercially ready for what AI will expose?

They are questioning what to prioritise, what to pause, and how to avoid accelerating weaknesses already embedded in their operating models.

AI will not simply improve Procurement.
It will amplify it.

Fluency with AI tools is fast becoming baseline. The differentiator is whether leaders can combine AI-enabled insight with commercial influence, governance clarity and operating model discipline.

That was the focus of our recent joint webinar with WNS Procurement: ‘The Next Procurement Skillset – Combining AI Fluency with Commercial Influence’.

The session brought together three complementary perspectives:

• Graham Crawshaw, Content Director at CASME, reflecting insights from global member conversations and benchmarking
• Victoria Folbigg, former Chief Procurement Officer and Board Advisor, with nearly 30 years leading transformation
• Gaurav Kumar, leader of Insights, Analytics and Operations at WNS Procurement, working with organisations implementing AI at scale.

What followed was not a technology showcase. It was a candid leadership discussion - less about tools and more about judgement under transformation pressure. The focus turned to data maturity, ownership clarity, operating model discipline, reinforcing a growing consensus: AI success will be determined by organisational readiness and structured implementation.

The First Reality Check: Your Data Will Decide Your Fate

We asked the live audience: What is the biggest barrier to realising value from AI in procurement today?

The response was revealing: 
Weak data foundations led at 25%. Disconnected systems had a response of 15%.

Together, 40% of respondents pointed directly to infrastructure. Not resistance, not leadership hesitation, and not cultural reluctance.

The barrier is not belief. It is readiness.

Live Poll: What is the biggest barrier to realising value from AI in procurement today?

As Graham reflected, “Companies are still struggling with clean data. Unless you get the fundamentals right, it makes it really difficult to build on that.” If taxonomies are inconsistent, ERP landscapes fragmented, or master data is unreliable, AI will not correct those weaknesses. It will scale them.

The implications are significant, because AI does not compensate for fragmentation. It accelerates whatever maturity already exists. As Gaurav put it plainly: “AI cannot really make procurement more efficient. It can only amplify the maturity that procurement organisation has.” 

That observation reframes the debate. Weak governance, poor data hygiene and structural inconsistency are not minor technical inconveniences. They are strategic constraints. 

The organisations seeing progress are not those moving fastest into tools. They are those strengthening foundations first, then layering intelligence on top.

AI will expose what you already are. The question is whether you are ready for that amplification.

Pilots Everywhere. Scale Almost Nowhere.

Many organisations are experimenting. Few are scaling.

 “A lot of pilots have been talked about,” Gaurav noted, “but very few companies have really scaled anything procurement-wide or organisation-wide from an AI standpoint.” 

The market is saturated with AI-enabled solutions. Demonstrations are compelling. Barriers to entry are low. But experimentation is not transformation.

Technological enthusiasm is abundant. Structural integration is not. 

Across CASME benchmarking discussions, a divide is emerging between organisations embedding AI into core workflows and those piloting indefinitely at the edges.

The difference is not access to technology.
It is leadership resolve.

Scaling AI demands more than licences and isolated use cases. It requires clarity on decision rights, defined risk appetite, governance redesign and trust in AI-informed recommendations.

As Victoria reflected, “To scale means you give AI all the inputs, and that is quite a scary and complex decision to make.” 

This is where hesitation sets in. Not because the technology is unavailable, but because the organisational implications are significant.

Scaling AI is not an incremental IT upgrade; it is structural change within Procurement.

AI can generate insight.
Only leaders can convert insight into enterprise value.

That is where commercial influence becomes inseparable from AI fluency.

Ownership: Collaboration Without Ambiguity

When asked who owns AI adoption in Procurement, 40% indicated joint ownership between Procurement, IT and Centres of Excellence. 

A further 13% reported no clear owner. In transformation programmes, unclear accountability is rarely neutral; it’s usually where momentum stalls. 

Collaboration is essential. Ambiguity is paralysing. 

Live Poll: Who is primarily accountable for AI adoption and outcomes in Procurement?

As Victoria reflected, “Joint ownership is always a little bit uncomfortable. It might feel as if you're hedging your bets.”

Successful AI adoption requires explicit accountability:

• Procurement must own commercial outcomes and supplier risk
• IT must own platform integrity, data security and technical architecture
• Finance must validate ROI
• Centres of Excellence must drive capability, standards and adoption.

Shared involvement does not remove primary ownership; Procurement cannot outsource accountability for its own transformation.

Without commercial influence at the centre, AI remains advisory rather than decisive. Advisory insight rarely transforms organisational outcomes.

The Blended Skillset Is Already Emerging

Our final poll asked: Which capability will most differentiate Procurement leaders in the AI era?

The strongest responses were:

• AI fluency
• Judgement and critical thinking
• Strong data foundations.

Closely followed by governance, risk management, stakeholder trust and commercial influence. 

The signal is clear: the future is not technical versus commercial, it is integrated capability.

Live Poll: Which capability will most differentiate Procurement leaders in the AI era?

This balance matters because AI will not eliminate the need for human judgement or calibrated risk-taking. As Gaurav noted, “A lot of the rule-based things AI will do better than humans. But it’s the relationship part that becomes even more critical.”

The next Procurement skillset is not technical alone. Across the market, two complementary capability tracks are emerging:

• Commercial leaders - negotiation, influence, stakeholder alignment
• AI-enabled specialists - analytics, optimisation, digital fluency.

But competitive advantage does not sit in either camp alone. It lies in the integration between them.

AI fluency without commercial judgement produces analysis without impact. Commercial influence without digital literacy limits strategic leverage.

In the AI era, commercial influence means more than negotiation capability. It means shaping stakeholder risk appetite, challenging assumptions within AI-generated insight, aligning cross-functional leaders behind data-informed decisions, and ensuring that identified value is realised, not merely reported. 

The next Procurement leader must be multi-dimensional: capable of interrogating data, defining governance boundaries and influencing enterprise decision-making simultaneously.

AI Cannot Sit Outside the Operating Model

Procurement teams remain under sustained pressure: cost discipline, volatility, regulatory complexity and rising stakeholder demand. Yet in many organisations, AI is still treated as an additional initiative.

That model is unsustainable.

As Gaurav shared from a CPO conversation: “It cannot be an additional thing that I expect them to do outside the nine to five.” 

When AI sits outside the operating model, it becomes peripheral.
When it is peripheral, it becomes optional.
Optional initiatives do not reshape performance.

There is also a growing realism around financial returns. As Graham observed, “Hardly anyone is actually generating savings - it’s not a quick fix.” The expectation that AI will immediately release cost capacity is not yet supported at scale. This is structural change, not short-term optimisation.

True integration requires:

• Embedding AI into sourcing, risk and category workflows
• Aligning KPIs
• Redefining role expectations
• Evolving governance structures.

This is not incremental IT enhancement - it is operating model redesign that demands leadership ownership.

The Inflection Point

The constraint is no longer technological, it’s organisational. AI will not replace Procurement, but it will expose fragility in data governance, ownership clarity and capability depth.

The poll results tell us that the barrier is structural readiness. 

Data and governance foundations must precede AI scale.
Ownership must be clear.
Role expectations, decision rights and governance boundaries must evolve.
AI must be embedded, not layered on.

Progress will not be defined by the number of pilots launched. Success will be defined by organisations that:

• Strengthen data before scaling
• Define ownership before delegating
• Embed AI into workflows rather than layering it on top
• Build practical AI fluency
• Elevate commercial judgement alongside digital capability.

Some teams will integrate AI into the core of how Procurement operates, and others will continue experimenting at the edges.

The advantage will not belong to those who explore first. 
It will belong to those who embed decisively.

A Leadership Question

Before accelerating further, ask:

• Are your data foundations strong enough to withstand scale?
• Is accountability for AI outcomes explicitly defined?
• Is AI embedded into decision-making, or running parallel to it?
• Are your leaders combining AI fluency with commercial influence?

Across the CASME community, these questions are no longer theoretical. They are being debated, benchmarked and acted upon by senior procurement leaders navigating real transformation pressures.

Through structured peer benchmarking and facilitated leadership dialogue, organisations are testing assumptions, challenging maturity gaps and defining what disciplined AI integration actually looks like in practice.

The conversation has moved beyond experimentation. It is now about execution.
Execution demands structured peer insight, not isolated experimentation.

To hear the full discussion, including detailed poll insights, practical scaling realities and leadership perspectives, access the full on-demand webinar here.

The next Procurement skillset is already emerging.
The competitive advantage will belong to those who build it, and embed it, first.