Your Procurement Team Is Using AI. But Is It Working?

Procurement teams are already using AI.

The real question is whether they are using it well enough to improve business results.

That distinction matters. According to the Work AI Institute, 87% of digital workers now use AI at work, and 75% say it makes them more productive. Yet only 13% say their organization is performing significantly better because of AI. Workers report saving roughly 11 hours per week through AI automation, but much of that value is not reaching the business.

Procurement leaders should pay attention to that gap. AI adoption is easy to see. AI value is harder to capture.

The Problem Is Not Lack of Use

Many procurement teams have moved past the “should we use AI?” stage.

Buyers are using ChatGPT, Copilot, Gemini, Claude, and other tools to draft RFQs, summarize supplier information, review contracts, analyze spend, and prepare negotiation plans.

The AI Readiness in Procurement 2026 Report found that 47% of procurement professionals use AI every working day. It also found that 58% use AI at least four days per week. The report’s conclusion is blunt: the question is no longer whether your team is using AI, but whether your organization has caught up.

AI Can Create Hidden Work

AI does not remove work automatically. In many cases, it moves the work.

The Work AI Institute calls this “botsitting.” It includes feeding AI missing context, checking outputs, debugging mistakes, rerunning prompts, and cleaning up confident but wrong answers. Workers spend an average of 6.4 hours per week botsitting AI.

Procurement has many high-botsitting activities.

A buyer may ask AI to draft an RFQ, then spend an hour fixing unclear requirements. A category manager may ask for supplier risk insights, then spend another hour checking whether the sources are current. A purchasing manager may ask AI to summarize contract terms, then still need to verify payment terms, rebate language, renewal dates, and termination rights.

The tool produced something. If the buyer hasn’t used the tool correctly, they are left with a lot of cleaning up to do.

Rework Erases AI ROI

Workday’s research found that roughly 37% of the time saved through AI is offset by rework. Employees spend significant time correcting, clarifying, or rewriting low-quality AI output. For every 10 hours of efficiency gained, nearly 4 hours are lost fixing the output.

This is not just a technology problem.

Workday points to gaps in skills, role design, and support. Many employees are expected to produce better results with AI, but they have not been trained to use it effectively.

AI may write a supplier email that sounds professional but misses the leverage point. It may summarize spend data but fail to identify the sourcing opportunity. It may draft negotiation questions but ignore switching costs, incumbent dependency, or supplier capacity risk.

The output looks good. But buyers have to finish it.

Unchecked AI Creates Business Risk

AI output can sound confident and still be wrong.

That matters in procurement because AI-assisted work can influence supplier selection, pricing strategy, contract terms, and stakeholder recommendations.

The Work AI Institute found that 69% of AI users admit to shipping slop at work. Slop is AI-generated work that workers have not reviewed, do not fully understand, or could not defend if asked.

That risk is not theoretical.

A buyer who accepts an AI-generated supplier comparison without checking the data may recommend the wrong supplier. A category manager who uses an AI-generated market summary without validating assumptions may build a weak sourcing strategy. A team that relies on an AI-generated contract summary without checking the source language may miss a buyer obligation.

Procurement teams cannot delegate judgment to AI—they need the skills to challenge AI conclusions before those conclusions become business decisions.

Procurement’s Biggest Barrier Is Skills

The AI Readiness in Procurement 2026 Report found that knowledge and skills gaps are the top barrier to AI adoption in procurement, cited by 41% of respondents. That ranked ahead of IT and policy restrictions, budget constraints, and organizational resistance.

This aligns with what many purchasing leaders see in practice.

Procurement professionals need practical AI skills tied to real procurement work. They need to know how to frame the task, provide the right context, guide the output, and review the answer.

How Procurement Teams Can Turn AI Use into AI ROI

AI works better when procurement professionals give it four things.

1. Give AI a Clear Role and Task

First, it needs a clear role and task. “Help with supplier research” is weak. “Act as a manufacturing procurement analyst and compare these suppliers against our qualification criteria” is stronger.

2. Provide Procurement Context

Second, it needs procurement context. Category scope, supplier history, cost drivers, stakeholder priorities, quality issues, contract constraints, and market conditions all change the answer.

3. Define the Output Requirements

Third, it needs output requirements. Buyers should specify the table, fields, scoring method, assumptions, risks, and next-step use.

4. Apply Human Review and Judgment

Fourth, it needs human review. Procurement professionals must validate sources, test assumptions, challenge conclusions, and decide what is safe to use.

These are learnable skills. They are also becoming core procurement skills.

APD’s AI Skills for Procurement Professionals course is designed to build those skills. Participants learn practical prompting, context building, AI limitations, output validation, and procurement use cases such as supplier research, spend analysis, contract review, RFQ development, negotiation preparation, and category planning.

The goal is not to use AI more. The goal is to use AI well enough to improve speed, quality, and business decisions.