AI takes on return fraud as holiday returns surge

Data from Happy Returns, a UPS-owned reverse logistics company, shows that nearly one in every 10 retail returns in the United States involves fraud. Retailers now lose an estimated $76.5 billion a year to the problem. 

To slow those losses, Happy Returns, which specializes in boxless in-store returns for online purchases, is testing a new artificial intelligence tool that flags fraudulent returns before refunds go out.

Sign up for my FREE CyberGuy Report
Get my best tech tips, urgent security alerts, and exclusive deals delivered straight to your inbox. Plus, you’ll get instant access to my Ultimate Scam Survival Guide – free when you join my CYBERGUY.COM newsletter

NEW IPHONE SCAM TRICKS OWNERS INTO GIVING PHONES AWAY

How return fraud works in the real world

Return fraud often looks harmless on the surface. A shopper requests a refund for a legitimate item. Instead of sending back the real product, they ship something cheaper, damaged or completely different. Retailers often issue refunds before anyone inspects the item. That speed allows fraud to slip through and drives up costs.

Industry data from Happy Returns and the National Retail Federation shows retailers will handle nearly $850 billion in returned goods in 2025, representing almost 16 percent of total retail sales. According to the same research, an estimated 9 percent of those returns are fraudulent. The report also finds that many shoppers admit to some form of return policy abuse. Importantly, because Happy Returns conducts in-person item verification and uses AI-powered automated flagging plus audit processes to catch mismatches, the rate of confirmed fraud in its network is much lower than the industry-wide estimate. 

Why boxless returns changed the equation

Happy Returns operates nearly 8,000 in-person return drop-off locations inside stores such as Ulta Beauty and Staples, as well as at UPS locations. Shoppers can return eligible items without a box or shipping label, and refunds are often issued quickly after verification. Like any returns channel, fraud attempts can happen, but in-person drop-off, item verification, ongoing flagging and audits help keep confirmed fraud far lower than broader industry averages.

Happy Returns says its boxless, in-person model already blocks many common fraud tactics, including empty boxes, partial returns and fake tracking numbers. “If you never touch the product, you can’t actually know what’s being returned matches what was sold,” the company says. Everlane says that physical handling alone acts as a deterrent. “Just the fact of knowing an individual will physically handle and verify the product at the Return Bar deters fraudsters from even attempting to commit fraud,” said Jim Green, director of logistics and fulfillment at Everlane. 

Still, Happy Returns acknowledges that fraud tactics continue to evolve. Lookalike products and knockoffs can closely resemble the real thing, making subtle differences hard to spot without close inspection.

THE FAKE REFUND SCAM: WHY SCAMMERS LOVE HOLIDAY SHOPPERS

Happy Returns is testing a new artificial intelligence system designed to flag fraudulent retail returns before refunds are issued. (Photo By Helen H. Richardson/The Denver Post via Getty Images)

How Return Vision uses AI to detect fraudulent returns

This holiday season, Happy Returns is piloting its new AI system with select retailers, including Everlane, Revolve and Under Armour, as return volumes spike.

The new AI tool is called Return Vision. It starts working the moment a shopper initiates a return online. The system looks for unusual patterns across return timing, frequency and location. A single return may appear normal on its own. When those signals overlap in suspicious ways, the return is flagged for review before a refund is issued.

At drop-off points, workers can scan item barcodes and see photos of what the item should look like. They can reject obvious mismatches on the spot. Once returns reach Happy Returns hubs in California, Pennsylvania and Mississippi, flagged packages are sent to human auditors. The items are opened and photographed, including images of the front, back and identifying labels.

Those photos are fed back into the AI system, which compares them against official product images and past transaction data. Human teams review the AI assessment and make the final decision. The goal is not automation alone. It is adding multiple layers of review where fraud is harder to hide.

Early results show how effective AI is at catching return fraud

While still in pilot, Happy Returns says Return Vision is showing early results. Less than 1 percent of returns flowing through its network are flagged as high risk. Of those flagged returns, about 10 percent are ultimately confirmed as fraud. The average prevented loss per confirmed case is just over $200. 

Happy Returns says the system focuses on high-confidence cases, allowing most shoppers to move through returns without delay. The company notes that the tool does not address every form of abuse, such as wardrobing, where customers return worn items.

WHY YOUR HOLIDAY SHOPPING DATA NEEDS A CLEANUP NOW

Boxless, in-person returns are helping retailers cut down on common fraud tactics like empty boxes and fake tracking numbers. (Photographer: David Paul Morris/Bloomberg via Getty Images)

Why retailers are turning to AI to stop return fraud

Happy Returns is not alone in turning to AI to stop return fraud. Amazon and FedEx both offer boxless returns and use automated systems to flag risky behavior. The U.S. Postal Service is rolling out similar services. Across retail, 85 percent of surveyed merchants say they use AI or machine learning to combat fraud. Many say the results have been mixed. Happy Returns says combining behavioral signals with physical product verification helps close gaps that data-only systems often miss.
 

Take my quiz: How safe is your online security?

Think your devices and data are truly protected? Take this quick quiz to see where your digital habits stand. From passwords to Wi-Fi settings, you’ll get a personalized breakdown of what you’re doing right and what needs improvement. Take my Quiz here: Cyberguy.com

CLICK HERE TO DOWNLOAD THE FOX NEWS APP

Kurt’s key takeaways

Retail returns have changed, and so has the fraud that comes with them. Easy drop-offs and instant refunds made life better for shoppers, but they also created new vulnerabilities. Happy Returns is betting that AI combined with hands-on inspection can tip the balance back toward retailers. Early results suggest it can help, even if it is not a cure-all. As fraudsters adapt, retailers are learning they have to adapt faster.

Should retailers slow down instant refunds if it helps stop return fraud, or should convenience always come first? Let us know by writing to us at Cyberguy.com
 

Sign up for my FREE CyberGuy Report
Get my best tech tips, urgent security alerts, and exclusive deals delivered straight to your inbox. Plus, you’ll get instant access to my Ultimate Scam Survival Guide – free when you join my CYBERGUY.COM newsletter 
 

Copyright 2025 CyberGuy.com.  All rights reserved.