Managing employee expenses is a mundane, time-consuming, and expensive job that not only chews up significant resources but is also incredibly difficult to do well. The swift reimbursement of out-of-pocket expenses to employees is an essential task if you want to keep those staff happy — but is riddled with areas of unnecessary spend, waste, and even fraud that require almost microscopic analysis to find.
One solution is to field an army of highly skilled expense auditors — but even IF your organization could afford to scale up this elite team of finance professionals, they would still only be able to identify a small proportion of the overall wastage.
To do this job properly, you need a team of operatives with infinite scale, full access to all of the enterprise’s information systems, and who never get tired.
To do this, you need artificial intelligence.
Common Areas of Unnecessary Spend, Waste, and Fraud
But before we discuss how to solve the problem, let’s first explore some of the most common areas where expense policies come under abuse.
Bites and Booze
Expense policies almost universally define expectations around food and drink very clearly. But that doesn’t stop employees from pushing those clearly expressed boundaries. Whether trying to sneak alcoholic drinks through as meal items or simply submitting breakfast, lunch, and dinner receipts that total more than the agreed daily stipend, careful processing of bites and booze is necessary.
No-one will submit more than one copy of the same receipt, right? Wrong. Whether unwittingly or deliberately, duplicate expenses can cost enterprises heavily if not identified. Examples of where to look for duplicates include:
a. In different monthly reports for a single employee
b. In reports for multiple employees
Preferred supplier lists in the procurement function are commonplace, but this concept is increasingly important to employee spend, especially for travel expenses. Being able to validate certain expense types against specific vendor lists quickly is now a core capability.
Artificial Intelligence automation dramatically enhances the finance team’s ability to quickly and accurately process expenses and identify spurious claims. The scale of AI offers organizations the possibility of achieving 100% coverage — applying every policy against every expense line and report. But this level of automation alone won’t find the detailed exceptions that exist deep within expense reports.
Details, Details, Details
Straightforward expense abuse, such as the variations shown above, are easy to spot for humans and AI alike. But there are many more scenarios where policy violation can occur, each increasingly nuanced and hard to identify. Below we detail a series of genuine scenarios that organizations are currently using AI to intercept:
Work from Home
- Claims for hotel stays, car rental, or meals while an employee is working from home.
- Expenses for both Home Internet and Co-working space within the same month.
- Hotel, car rental expenses and travel meals in the employee’s home/work location.
- Items on expense reports that are also processed via Accounts Payable.
- Duplicate expenses where one submission is via a receipt and one as a corporate credit card entry.
Bites and Booze
- Number of guests on a restaurant invoice matches the number on the expense claim
- Claims for a per diem expense BUT where also listed as an attendee in another person’s meal expense on the same day.
- Gifts/recognition items distributed to one recipient from multiple expense users.
- Flights booked less than 7 days before departure
- Limousine or private car service booked outside of allowed hours
- Plane tickets purchased for individuals with the employee’s same last name but different first name.
- Air travel for multiple dates with no hotel reservation.
- Rounded-up value purchases on credit or store cards
- Claims that break custom thresholds for hotel, airfare, and alcohol expenses based on GL/project code, reflecting expense policies that vary by client contracts.
- Expenditures outside of work hours (i.e. weekends, evenings, PTO, holidays)
- Validity of vendors – especially relevant in geographies such as India where all vendors should have registered trading ID (GSTIN)
Each of these requires the auditor to look beyond the simple expense line, to include external data, and to even look across other expense reports if necessary. In short, simple policy automation doesn’t cut it here – AI needs to do more.
Expense-Specific AI Automation
Automation that delves deeper into expense auditing and explores the areas of waste and fraud requires an additional set of capabilities. These extended requirements are needed to expand the scale of checks performed on each expense line, enrich the available collection of data to cross-check against, and continually learn from historical examples of expense abuse.
The AI platform needs to offer the following extended capabilities:
All expenses across all dimensions
AI systems have access to vast amounts of processing power, so provide the ability to explore expense comparisons across multiple dimensions. For example, when looking for duplicate expenses, an AI can search across
- All expenses in a single report from a single employee
- All expenses across time from a single employee
- All expenses across time from across all employees
- Across all offices
- Across all countries
Integrated data from other sources
Humans auditors apply common-sense and external data to expense processing all the time. AI systems add this context and flavor to automation by integrating data sources that enrich their view of the expense, including:
- Internal systems such as HR records and travel portals
- External data sources such as internet reviews, mapping systems, and even social media.
Learning from previous activity
A human trait is to learn from previous examples, and advanced AI platforms also excel at this. This allows an AI to “remember” that an employee has been known to add extra miles to their mileage claims in the past — this then becomes a prominent area to check for in that particular employee’s reports and a known exception to check for with all other employees.
Each organization has unique, and often complex, custom policies. The application of these policies across all of the dimensions described above is something that an AI can achieve far more cost- and time-effectively than human auditors.
Improving the efficiency of their department is high on the priority list for all finance leaders. Technology is frequently seen as the tool to provide the desired improvements — with artificial intelligence increasingly touted as the final piece in the jigsaw to deliver true automation in the finance function. Yet think finance automation, and you usually think about invoice processing or procurement, not expense auditing. But that view is changing.
The move to remote working has increased the number of employees submitting monthly expenses, and expense fraud is on the rise. Finance teams still operating manual processes for expense management are drowning in reports and exceptions — but those organizations who are applying AI automation in this field are experiencing significant benefits. From 100% coverage of policies to automated identification of spend wastage and fraud detection — AI automation is transforming expense auditing from an expensive cost center into a leading example of how technology can positively impact every part of the finance function.