Procurement’s AI Adoption Gap in 2026: What Closes It

BCG‘s 2026 supply chain research shows that only 35% of procurement teams have adopted AI. That is the lowest adoption rate among the 13 business functions it studied. Supply chain and sales lead at 44%. Finance follows at 40%, while HR stands at 37%. The biggest hurdle is that most AI initiatives across business functions are still stuck at the co-pilot stage, showing only marginal returns. Procurement’s AI adoption gap sits widest of them all.

The Adoption Gap Is Real

BCG’s May 2026 Executive Perspective, drawing on a 1,250-company global study, ranks supply chain and sales as AI’s most-adopted functions and procurement as its least-adopted. Even the AI leaders get stuck. Most are stuck on narrow use cases, and only about 30% of companies say planning has actually delivered value they can measure. A real redesign means rewiring workflows end to end, not just patching pieces here and there. That takes cross-functional trade-offs only a CEO can resolve. Most companies haven’t gotten there yet. It’s one piece of a broader shift toward agentic AI in supply chain management that’s reshaping the enterprise.

Why AI Redesign Stalls

Redesign stalls because the skills to run it barely exist yet. Gartner found demand for AI-skilled supply chain roles has grown 387% since 2023, outpacing the entire labor market. BCG found the reason. 92% of the companies don’t have the people to handle unstructured data. On top of that, 88% report silos getting in the way of teams working together on AI, and 71% are short on AI talent. It’s not really a headcount problem. What’s actually stalling redesign is messy data and handoffs that keep breaking.

The harder barriers sit inside the organization. The Hackett Group’s 2026 research found that 59% of procurement teams lack the AI expertise to manage current deployments. What’s missing is capability and structure, not willingness to change. The fastest movers redesign around the gap instead of hiring out of it.

One global consumer goods company, in a case BCG documented, didn’t bother with copilots. It jumped straight to letting AI agents handle replenishment recommendations. In-stock rates rose 2% to 4%, fill rates 4% to 10%, and administrative costs fell 40% to 60%, without new headcount. ProcureAbility looked into this for its 2026 CPO-CIO Report. Turns out 60% of organizations are already putting AI or data people right inside procurement teams, mostly redeployed rather than newly hired. So the fix isn’t more hires. It’s building the work around the gap.

BCG’s Fix: A Build-and-Buy AI Platform

BCG’s answer isn’t “hire more AI specialists.” It’s a hybrid build-and-buy approach: buy the foundational planning layer, build the custom orchestration on top. BCG’s own reasoning for the “buy” half is blunt. Vendors are building agents faster than any single company could on its own. They’re also picking up lessons across industries that no one company could learn alone. That lines up directly with the gap Gartner just measured. You don’t have to out-hire a 387% spike in demand. Not if someone’s already built the foundation for you. Most procurement teams are only starting to treat build-and-buy platform strategy as a deliberate choice instead of a default.

BCG adds one more condition: AI decisions have to be transparent, auditable, and explainable. Plans need to show their data sources, their assumptions, and their trade-off logic, not just a recommendation, but the reasoning behind it. That’s what builds trust across commercial, operations, and finance. Not the model’s confidence, its paper trail.

In BCG’s framing, supply chain teams don’t disappear from the loop. They verify and approve the agents’ plans, then spend the time they get back on strategic work only people can do. That’s augmentation, not replacement. Gartner’s Marco Sandrone warned in February 2026 against treating agentic AI as a “blunt instrument” for headcount reduction, and BCG’s model doesn’t try to. The specialist talent gap becomes the platform provider’s problem to solve, not every individual procurement team’s. That’s the same tension running through the wider debate over AI and procurement headcount that Gartner’s February survey surfaced.

What Changed in 2026

This conversation is accelerating, not settling. BCG’s research came out in May 2026, its summary followed on June 2, and Gartner’s 387% figure landed two weeks later, on June 15. That’s three converging data points in six weeks. Six months ago, most procurement commentary was still asking whether AI adoption would happen at all. Now the data says adoption is happening, just unevenly, and procurement’s AI adoption gap is the widest one measured.

The Bottom Line for Procurement in 2026

Procurement doesn’t close this AI adoption gap by outcompeting everyone else for scarce AI talent. It closes it with a platform built the way BCG describes: engineered as the foundation, verified by the team. Transparent enough that nobody has to take its word for it.