How Pennsylvania’s Redner’s Markets is using artificial intelligence to catch shoplifters

NORTH WALES, Montgomery County (WPVI) — You can now add theft prevention to the growing list of ways that artificial intelligence technology is being used in the real world.

A Pennsylvania-based grocery store chain has successfully used software to stop shoplifters — sometimes even predicting when they’ll strike again.

Redner’s Markets is now using AI software to help flag suspicious transactions and stop thieves at the self-checkout line.

“If they see it happen multiple times then it starts to steer them to stores and or predictive dates as to when you might see a similar transaction occur,” said Eric White, the spokesperson for Redner’s Warehouse Markets.

White says the stores have seen an increase in customers entering an item into the self-checkout computer that costs a lot less than the item they’re putting in their cart.

“She was scanning bananas and taking meats and cheeses for a catering or a charcuterie business,” said White, describing one woman who was running her own business using this method. She’s accused of stealing around $5,000

“If you’re going to form your own small business, theft shouldn’t be your main driver of getting your goods,” said White.

The store was able to catch her pattern using AI, even predicting dates as to when a similar transaction would occur again.

“The system told our agents, ‘Hey this person is most likely to be at the store at this time of day.’ And our loss prevention agent just took a shot, went up there, and sure enough found the individual doing it,” said White.

The technology is installed at all Redner’s locations and it’s catching more people now.

“These individuals should be able to pay for those groceries and they took these steps to be so creative to try and get around it and we found it anyway,” said White.

The bottom line is people stealing will ultimately drive up the cost of groceries at the store, and Redner’s hopes this technology will help them keep prices low.