News & Resources


5 Ways to Stop AP Leakage at Scale With AI

BY: Chih-Hsuan Wu | 05/27/21

The accounts payable (AP) team has several responsibilities to juggle. Not only are they the gatekeepers of an organization’s precious funds when companies are facing a cash crunch from the COVID-19 crisis, but they are also the first line of defense against fraud. Now, their roles are more important than ever.

Each day, invoices flood an AP team’s inbox. With each invoice, they need to determine if the invoice is a duplicate, if each line item is within policy, and if the pricing is correct, among a plethora of other standard checks. To complete this task, AP teams cross-reference previous invoices, purchase orders (PO), and contracts.

In addition to the standard checks, new checks are being put in place because of the current economic climate. Organizations have issued mandates to tighten budgets, making significant cuts for certain categories of purchases. The AP team also needs to verify if the purchase is still allowed or if the purchase would exceed budget caps.

However, the time it takes to process each invoice isn’t measured in seconds or minutes. So, without a large enough team to read through each request, how can an AP team run efficiently enough to pay their vendors on time without compromising the level of analysis needed for financial control?

Teams are forced to randomly select a small set of invoices to manually review, while letting most of the lower-value invoices pass through—a compromise they need to make due to limited resources.

Some AP teams may be lucky enough to have the aid of some automation tools (be sure to read the other feature articles in this special section), or shared services departments, but without incorporating the proper context from contracts or other key documents, the checks are not as thorough as they could be. The team is left with the same problem they started with: they never feel the coverage is enough, and it’s impossible to ever get to 100% coverage.

The only thing AP teams can do is hope that there aren’t too many invoices coming in the door, so their workload isn’t overwhelming. And once an invoice is paid, they hope that it isn’t later flagged by an internal audit. Either way, the payments have already been made, marking the start of the inefficient and ineffective AP recovery process.

This audit dilemma almost seems perfect for artificial intelligence (AI) to tackle at scale. Not only does AI have the capability to interpret large datasets and process decisions quickly, but it also can identify anomalies (such as sudden spikes in pricing) from past trends.

That is the reason AI has become the central component of the latest generation of AP auditing solutions. With AI-powered AP audit solutions, AP teams can audit 100% of spending in their organizations without employing extra resources. Errors, waste, fraud, and noncompliance spending can all be identified prior to payment.

1. Verify Every Transaction Against Policy

Many organizations have a spending limit policy based on the requestor level. Considerations may include the amount a requestor can request during a set period, or the type of allowable purchase based on the requestor’s job function. These are common examples of policies auditors need to manually check. With the help of AI-powered AP audit solutions, every transaction can be automatically audited against all spend policies in the same manner as a human auditor. Moreover, these solutions can approve low risk items, so auditors have the time and energy to focus on studying the identified problems without spending significant time discovering the issues that need to be fixed.

2. Check Duplicate Invoices

By examining 100% of the invoices, expense reports, and corporate card transactions, AI-powered AP audit solutions can quickly flag suspicious transactions. The most advanced solutions are also designed to extract contextual information directly from documents by reviewing manually entered information in comparison with information on vendor, purchase date, and total receipt amount. If any criteria match previously submitted invoices, the solutions then mark the invoice as high-risk, whether the duplicate is from the same vendor or listed under a different name.

3. Validate Payment Terms, Discounts

Sometimes, inadvertently, there are payment terms and discount discrepancies between what was contractually negotiated and what is invoiced. Unless every invoice gets evaluated side by side against the contract, this leakage is usually overlooked. With the contractual analysis AI-powered AP audit solutions can perform, every invoice’s payment term and promised discounts can be verified against its contract. In addition, the best-in-class solutions in the market also can interpret complex pricing tables, discount structures, and labor rates to eliminate the need for refund requests from vendors.

4. Stop Paying the Wrong Vendor

In addition to the abundance of ways an outsider can attempt to commit fraud, payments can inadvertently be made to the wrong vendor due to simple manual entry errors. AI-powered AP audit solutions can find spending leakage before payments are made. By learning the payment types and payment patterns, AI can detect if there’s a spike in the invoice amount, an anomaly in items purchased, or a change in bank details, without any manual comparisons.

5. Reduce Risk Exposure

The AP team can be the last line of defense against risk. When a Foreign Corrupt Practices Act (FCPA) violation happens, the costs add up to an enormous sum. The cost of an FCPA penalty averaged $88 million in 2019, according to Stanford Law School, and $25 million is the starting cost of a civil FCPA violation for book- and record-keeping for corporations. What if it was simply a mistake? If an investigation is triggered, the average monthly cost of an FCPA investigation is $2.2 million. In addition to the internal audit and potential legal costs, immeasurable damages might come to the business’s reputation.

So, how can an auditor remember all of the names of the heads of state and the businesses they are associated with to be in compliance with the FCPA or identifying politically exposed persons (PEPs)? AI-powered AP audit solutions can assist the AP department in avoiding those costly mistakes. Using the technology to aggregate and analyze spending from individuals and vendors and compare it with external databases and sources provides the coverage organizations need to ensure adherence to proper governing laws. The leading AI-powered AP audit solutions also can identify expired agreements and missing provisions (i.e., anti-bribery corruption provisions for high-risk vendors) as an extra layer of protection for the organization.

In short, the risk of AP leakage is not worth the risk. If investment in AI-powered AP audit solutions isn’t something that you’ve considered, now is definitely the time to reevaluate given the uncertain economic landscape with COVID-19. Quickly and effectively stopping money walking out the door is one of the keys to weathering the storm ahead.

Chih-Hsuan Wu is Product Marketing Manager at AppZen who has a background in systems and strategy consulting.