Researchers have developed a new artificial intelligence (AI) system to detect accounting fraud within individual companies and across supply chains and industries.
Machine learning technique FraudGCN analyzes trends in financial data and corporate relationships to identify and predict fraudulent activity. It uses graph theory and machine learning to examine the network of relationships between companies, their auditors, and industry peers.
“It’s a never-ending mathematical arms race between the authorities and the fraudsters,” Chen Xu Wanglead author of the paper and associate professor in the School of Software Engineering and the Smart Grid and Network Security Key Laboratory at Xi’an Jiaotong Universitysaid in a press release.
The development comes as financial markets grapple with the impact of accounting fraud. report reveals that 62% of financial institutions with assets over $5 billion are reporting an increase in financial crimes, revealing growing vulnerabilities in the U.S. banking sector. As fraud methods evolve, including the potential use of AI by criminals, there is growing interest in developing more effective detection methods.
Current methods of detecting financial fraud
Traditional fraud detection methods often rely on audits, which can be labor-intensive and fail to differentiate between a company’s real results and manipulated figures. These hurdles mean that many companies can remain unchecked for long periods of time.
Paul WnekFounder, CEO and Principal Solutions Architect at DevelopAPtold PYMMTS that there are several common types of fraud in businesses: “Invoice fraud, such as fictitious invoices for goods or services that were never delivered or legitimate invoices altered to divert funds. Vendor fraud, such as setting up fake vendors to receive payments for nonexistent goods or services or bribes to reward vendors or employees who approve contracts or invoices. Payment fraud, which can occur when fraudsters access payment systems or manipulate approval processes to authorize fraudulent payments.”
These patterns can be difficult to detect using conventional methods.
“What is needed is an efficient and accurate algorithm to automatically identify accounting fraud and leave behind the days of random audits,” said Mengqin Wanganother researcher involved in the FraudGCN project, according to the statement.
FraudGCN attempts to solve this problem by building multi-relational graphs representing the connections between companies. This allows the system to analyze trends across corporate networks.
When tested on data from publicly traded Chinese companies, the researchers found that FraudGCN outperformed current approaches by a margin of 3.15% to 3.86%.
However, the practical implications of these improvements for fraud detection are not yet clear.
The role of AI in fraud detection and perpetration
As AI’s role in fraud detection expands, experts note its potential to both detect and assist with fraud. Joe StephensonDirector of Digital Intelligence at Interteldiscusses the dual nature of AI in this context.
“In the insurance industry, we are salespeople, and as such, we often overlook the potential implications of emerging technologies like artificial intelligence on claims,” Stephenson told PYMNTS. “While AI is very useful for underwriting, we also see criminals leveraging ChatGPT and AI to enable fraudulent activity, whether through the development of synthetic identifiers or metadata.”
This introduces new challenges, as Stephenson explains: “Metadata is not traditional and the use of social media makes it easy for anyone to exaggerate allegations or organize criminal groups.”
However, AI can also be used to analyze large volumes of data.
“Advanced algorithms can analyze social media activity, identifying patterns and anomalies that might go unnoticed by human investigators,” Stephenson said.
THE “Financial Fraud Prevention Manual“The PYMNTS study examines how financial institutions can leverage advanced technologies like behavioral analytics and machine learning to combat digital-age fraud tactics, including AI-powered schemes, malicious bots and synthetic identities that evade traditional security measures.
Automation of fraud prevention
Alongside AI detection tools, the industry is also implementing automation of financial processes as a preventative measure.
“Automated accounting systems designed with the best security measures offer built-in fraud detection capabilities, such as anomaly detection and invoice matching algorithms,” Wnek said. “The best platforms are one-stop shops for all accounts payable tasks, resulting in fewer systems that data must flow between.”
This shift towards automation is an ongoing trend in the financial sector.
“While traditionally resistant to change, accounting teams have begun to recognize the value of automation to improve the efficiency and accuracy of processes such as accounts payable (AP) and accounts receivable (AR),” Wnek said. “AP automation completely eliminates this line item, reducing costs by 40 to 95 percent.”
Additionally, by reducing manual intervention in financial processes, automation can reduce the risks of certain types of fraud.
However, adopting these technologies comes with its challenges.
“The two biggest barriers to adopting accounts payable and accounts receivable digitization are cost and complexity,” Wnek said. “But these barriers are easily addressed when companies consider the return on investment compared to outsourcing customer support and financial processing.”
