Coupa: AI Is Cure for Procurement Fraud Epidemic

Procurement fraud has evolved on the back of digitally sophisticated tactics from an occasional business headache into a full-blown financial and reputational disaster for organizations worldwide.

It’s no longer just a paper problem; the issue is moving from the back office to the boardroom.

“We have to change our mindset,” Michael van Keulen, industry principal, procurement, at Coupa, told PYMNTS. “Fraud prevention isn’t just a finance problem or a procurement problem — it’s a business imperative.”

“In today’s environment, where AI [artificial intelligence] is at the forefront and external threats are increasing due to geopolitical and economic instability, companies are more vulnerable than ever,” he added.

Data from a recent PMYNTS Intelligence report, conducted with Coupa, confirmed what should already be keeping C-suite leaders up at night: middle-market firms, which already self-report operating in high-uncertainty environments where cybersecurity risks and operational uncertainty intersect, are among the most exposed to procurement fraud.

Nearly 9 in 10 of these businesses (87%) report higher customer losses due to fraud than their more stable counterparts. The reason? A combination of frequent regulatory changes, volatile markets and outdated procurement processes.

“If you have not operationalized your procurement process appropriately, [and] if you don’t have a certain level of maturity, you’re receptive to fraud,” van Keulen said.

Read more: Fraud in the Procure-to-Pay Cycle Hits Middle-Market Firms Hard

Fraudulent activities in procurement can range from overpayments and duplicate invoices to unauthorized transactions.

“We’ve seen companies defrauded for tens and sometimes hundreds of millions of dollars simply because they didn’t have the right institutionalization of their procurement practices,” van Keulen said.

Traditionally, companies have taken a reactive approach — addressing fraud only after it has occurred. But as van Keulen stressed: “Preventing fraud is a much better way. With AI and machine learning, you can detect fraudulent behaviors before they escalate, using real-time alerts and large data sets to pinpoint anomalies before payments go out the door.”

For companies hesitant to embrace automation, the data is compelling: Firms that use automation are twice as likely to reduce procurement fraud compared to those relying on manual methods like staff training.

“Humans inherently introduce bias and inefficiencies,” van Keulen said. “Manual fraud detection often operates on a reactive basis, where fraud is identified after the money has already left the organization. But automation works 24/7, recognizing patterns and flagging suspicious activities in real time.”

Beyond detection, innovations like AI can continuously learn from transactional behaviors, refining the ability to recognize fraud patterns.

“We have $7 trillion worth of spend in our [Coupa] community,” van Keulen said. “That’s real data from real suppliers, training our AI to detect anomalies faster and more accurately than any human could.”

Despite the clear benefits of automation in reducing fraud, adoption remains sluggish — only 28% of firms have integrated automated fraud detection, despite 71% recognizing it as the most impactful solution.

Van Keulen sees this as a critical gap, noting that one of the primary barriers to automation adoption is a misunderstanding of how fraud detection works.

“There’s no one-size-fits-all solution,” he told PYMNTS. “What works in one industry won’t necessarily work in another. More importantly, fraud detection is only as good as the data feeding into it.”

Automation doesn’t just prevent fraud — it also enhances efficiency when compared to legacy methods. To tackle procurement fraud head-on, Coupa, for instance, developed Spend Guard, an AI-driven safety net that identifies potential fraud and inefficiencies in real time.

“Spend Guard shifts organizations from a manual, reactive approach to an automated, proactive one,” van Keulen said. “It uses AI and machine learning to analyze spend data across all applications, leveraging community intelligence to flag fraudulent behavior before it leads to financial loss.”

“Take duplicate invoices, for example,” he said. “Some companies still pay third-party auditors to detect duplicates, which is ridiculous when AI can do it in real time before the payment is processed.”

As procurement fraud continues to threaten business stability, the message is clear for middle-market firms facing heightened fraud risks: adopt real-time fraud detection, leverage AI for pattern recognition and integrate automation across the procurement lifecycle.

“Executives often underestimate the impact of procurement fraud because it’s hard to quantify — until it happens,” van Keulen said. “When it does, it’s not just about financial loss; it’s about reputational damage and shareholder value.”

But by breaking down operational silos, leveraging AI-driven solutions and prioritizing automation, businesses can turn procurement fraud from a lurking threat into a manageable risk.

The technology is out there; now it’s time for companies to use it, van Keulen said.

The post Coupa: AI Is Cure for Procurement Fraud Epidemic appeared first on PYMNTS.com.