Jeremy Clopton, CFE, CPA, ACDA, CIDA
Director, Upstream Academy
A recent Reuters article about the “expense report of the future” included a discussion about the future of fraud risk management and the use of robotics. Though not a new discussion, this article highlighted the most common uses of the technology without using the trendy approach of claiming robots are taking over.
Here are some key takeaways to consider:
Even amazing technology needs an examiner’s expertise. The use of robotics, machine learning, and advanced artificial intelligence is generally focused on increasing effectiveness and efficiency. The discussion around solutions like AppZen highlights each of these benefits. It also discusses what happens with a red flag is found — it’s sent to a human auditor. No matter the complexity of the technology, the research, due diligence and investigation of a possible finding is still most effectively conducted by a human auditor. Rather than focusing on a purely automated solution, think augmented solutions.
No matter what we do, fraud risk and unusual transactions are present. This is highlighted by the closing of the article that discusses the fact that AppZen has yet to run an analysis on a company that comes back clean. This shouldn’t come as a surprise to anyone using technology for fraud prevention and detection. Even if the findings aren’t fraudulent, they are likely exceptions to a policy or some type of unusual circumstance. These are relevant to company management looking to improve policy education and compliance.
Risk management automation improves operational efficiencies as well. Much of the focus around efficiencies is tied to that of the examiner. Using technology will help fraud examiners do more with less and get their investigations done more quickly. Though likely, this isn’t the only benefit. When risk management is automated, the processes being tested have operational efficiency opportunities. Focusing expense reimbursement testing around data, rather than the traditional paper receipt system, increases operational efficiencies for traveling employees as well.
A major consideration for fraud examiners looking to apply next-generation technologies for fraud risk management is understanding the limits of the technology. An examiner needs to know how a system learns, what the parameters are for analysis and the expectation for follow up on the part of the examiner. If an examiner can thoroughly plan the approach to using this technology, not only will fraud risk management become more efficient, it will become more effective. To this author, I’d much rather have a more effective fraud risk program than a faster one. Bring on the robots.