In Expense Fraud, Admitting There’s A Problem Is The First Step

For any organization with an attitude of, “It can’t happen to me,” here’s a wakeup call: Corporate fraud is common and on the rise.

Earlier this year, researchers at the Association for Financial Professionals found that instances of check fraud and Business Email Compromise scams increased between 2015 and 2017, with fraud originating both within and outside the enterprise.

One of the easiest areas for internal fraud to occur is in employee expensing. Especially for businesses that rely on manual expense reports, management and auditing processes, an employee that fudges a dollar amount here and there, or a worker who accidentally submits an expense report twice, mean companies can easily fall victim to fraud, both intentional and unintentional.

“Every corporation wants to believe, ‘Our organization is very well run, and all of our employees have the right intent,’” said Anant Kale, founder and CEO of AppZen. “And the reality is that the majority of people absolutely do have the right intent; they do the right thing. But always, in every organization, there are a few who are commuting fraud. There is a lot of unintentional waste that also happens.”

AppZen provides a back-office solution to automate the auditing process of expense reports, using artificial intelligence (AI) technology to identify instances of fraud and error. The company recently announced $13 million in Series A funding, led by Redpoint Ventures and Resolute Ventures.

Investor confidence largely stems from AppZen’s ability to secure large enterprise clients, a feat Kale recently told PYMNTS is only possible by going in and identifying ongoing instances of fraud or misspend in corporates’ expense management processes.

“We go into the data and into the current auditing and compliance process and look at expense report data and show them all of the places where they have either a duplicate reimbursement to an employee or an employee [who] has fallen outside … policy or patterns of intentional misuse by an employee,” he said.

Part of the reason why these scenarios carry on undetected is because the traditional process by which auditors seek out and identify fraud or mistakes is massively burdensome, the CEO said.

“The manual process involves a team of people looking at expense documents — like receipts, credit card statements, travel documents — researching all of these things,” he explained. “It’s a very involved process. It takes a lot of time, and it’s very expensive to do.”

Because of this, only a small fraction of expense reports are ever audited completely, added Kale.

“The majority of expenses get paid out without auditing because of time and cost pressures,” he said.

This leaves the back door open for corporate fraud and other mishaps in expense management, like employee spend that falls out of compliance with company policy — both of which, Kale noted, are incredibly difficult to detect. The ability of artificial intelligence and other technologies like natural language processing and machine learning to sift through troves of expense report data means organizations can stop letting errors, duplicate payouts and ongoing fraud schemes fall through the cracks. Kale said many of these tools are coming in from the consumer world as more innovators wield these technologies’ power for B2B use cases.

“Companies like Google, Facebook, Amazon and Microsoft have all done amazing work in open-sourcing these technologies,” he said. “The challenge is to apply it in a way that solves a specific problem.”

The CEO said AppZen has been able to do so in the use case of identifying fraud in expense reporting, but, he added, the company has plans to expand the implementation of these existing technologies into other areas of the enterprise, like accounts payable — another area that is affected by fraud and expensive errors like duplicate B2B invoice payments.

“Expenses are just another form of invoice being paid to the employee,” he explained of the potential to apply AppZen’s solutions in an accounts payable scenario. “The volume of dollars flowing through accounts payable is far higher than expenses, and there are more controls in place today — everything from PO matching, two-way matching and things like that. But there is still a lot of waste happening in terms of invoices coming up multiple times in different forms, the same item being invoiced twice, fictitious vendors and collusion.”

Technologies like AI can certainly be wielded to identify and remedy these scenarios that are quite costly for the enterprise. But according to Kale, the true power in these technologies goes beyond saving organizations a buck.

“What sets this technology apart is its ability to holistically look at that volume of data and connect the dots,” he said. “You can save money, but the real power is being able to identify patterns of fraud [and] non-compliance, which essentially means there is a bigger problem.”