Retailers tap into AI to cut down on return fraud

ST PETERSBURG, Fla. — Return fraud is becoming a bigger problem for online retailers, and now companies are being forced to look at how they can crack down on the expensive issue.

According to the National Retail Federation, retailers are expected to hit nearly $850 billion in returns after this holiday season wraps up. According to David Sobie, the head of UPS-owned logistics company Happy Returns, nearly 1 in 10 items returned is fraudulent in some way.

Sobie says 83% of companies he and his team polled this year, consider retail fraud a very serious issue. Retail fraud can range from returning an empty box in hopes of a refund, to more complex schemes like returning a similar looking item of lesser value and hoping the employee processing the return doesn’t notice.

Now, retailers are starting to use AI as a way to crack down on this problem. Sobie is doing the same with a handful of his retail partners using his new Return Vision solution that utilizes AI.

Return Vision works by scoring a return based on different characteristics, and if it’s deemed a higher risk, an employee will open the return and process the item using AI. The technology catches things the employee could miss.

He used the example of returning a less-expensive cotton sweater in place of an expensive one.

“AI catches subtle differences,” he said. “It could be a different neckline, different stitching, or where the buttons are placed. Some things that someone doing their job and doing their job well may not catch.”

Getting caught processing a fake return could get a customer banned from using that online retailer or even result in legal action.