SAS Institute expert explains how data can expose procurement fraud

Companies can use data already held in their business systems to spot suspicious purchasing patterns, according to Natasa Vlajic, a procurement integrity expert at SAS Institute.

At the SAS Innovate on Tour conference in Stockholm, Vlajic described how a system can flag purchases above a set threshold. Someone seeking to avoid scrutiny might try to learn that limit and repeatedly stay just below it. A single transaction may look normal, she said, but links among suppliers, contracts, invoices and payments can reveal patterns.

Other warning signs include duplicate invoices, purchases at prices that differ from contract terms, and a new supplier with ties to a previously blacklisted party. Similar supplier names may merit closer review, although name similarity alone does not prove fraud. Vlajic said analysis tools can also help identify indirect connections that manual checks may miss.

For telecom operators, she pointed to network construction and maintenance, where work involving several layers of subcontractors can be difficult to standardize. Comparing maintenance invoices across regions may reveal recurring differences from contracts or signals of possible coordination among suppliers.

Vlajic said companies can combine business-system records with external data. Once an investigator identifies a suspicious pattern, an AI assistant can search for it across the organization’s data. Some customers also use agentic AI to compare purchases with contract prices and terms on an ongoing basis. She said this could automate parts of the review of organizations with thousands of contracts, while responsibility for the work varies across internal audit, procurement, finance and compliance teams.