Happy Returns, the reverse logistics firm owned by UPS, is piloting artificial intelligence technology to curb returns fraud, a costly problem that continues to escalate for retailers.

According to reports, the company is testing an AI-driven fraud detection system with select retail partners during the holiday season. Happy Returns estimates that return fraud costs retailers $76.5 billion annually, as the volume and sophistication of fraudulent activity continue to rise.

Happy Returns’ approach blends human oversight with machine intelligence. At the company’s processing hubs in California, Pennsylvania, and Mississippi, returns flagged by the system are first opened by human auditors, who photograph the contents. These images are then analyzed by the company’s “Return Vision” AI, which compares them with product images and return data to identify discrepancies. Human teams review the AI’s assessments and make the final determination.

“If you’re returning a pair of $300 boots and you show up with a pair of dirty old sneakers, that should be caught immediately,” said Jim Green, director of logistics and fulfillment at Everlane, one of the retailers participating in the pilot. “What Return Vision does is add an extra layer of protection for some of the not-so-obvious cases.”

Early results suggest a targeted approach. Fewer than 1% of returns processed through Happy Returns are flagged as having a high likelihood of fraud, and about 10% of those flagged cases are ultimately confirmed as fraudulent. The average value per confirmed fraud is approximately $261.

Happy Returns Chief Operating Officer Juan Hernandez-Campos said the technology is becoming increasingly critical as fraud schemes grow more complex, the report noted.