Eyes wide open: How AI is reshaping grocery theft prevention.

As theft rises along the grocery aisles, retailers turn to emerging technology to mitigate losses—and pad notoriously tight margins.

Grocery merchants have endured shoplifting since the first shop opened its doors. Mitigating theft has always hinged on watchful store employees. For the most part, the snagged can of soup or bag of chips amounted to so much petty pilfering.

But grocery theft today is evolving into something more organized and frequent. According to the National Retail Federation, retailers across the board, including grocers, witnessed a 93% increase in annual shoplifting incidents between 2019 and 2023. For an industry where profitability balances on an impossibly thin profit margin, theft has become a serious problem.

Employee vigilance still matters. But thanks to artificial intelligence (AI), grocers finally have access to technologies that place eyes everywhere—and understand whether the shopper in aisle 12 slipped the package of lobster meat in his overcoat, rather than putting it back in the freezer.

"AI is giving grocers new vision—literally and strategically," says Donnafay MacDonald, research director at Info-Tech Research Group. "I sat on a call with a vendor and they said, 'Watch this video and try to detect who is stealing.' Nobody could detect it. But AI could."

Grocery retailers are just now beginning to explore these technologies, and incorporate them into their stores, from receiving decks to meat counters to checkout zones and front doors. The digital eyes don't just watch and record: They analyze behavior and alert humans to theft as it takes place in real time, and can identify people whose behavior suggests strong potential for shoplifting.

Continued below…

Theft-mitigation strategies through AI include the ability to identify repeat offenders when they walk through the front doors, notify employees and watch potential thieves as they navigate the aisles. The systems, too, help retailers understand what is being stolen most often, and when.

One zone especially ripe for AI help—checkout—was fairly theft-free for generations. But then self-checkout emerged—and now it represents a fairly busy target for thieves. Store losses range from customers swiping for potatoes, but placing a steak on the scale, to simply tossing the salmon filet or bottle of extra-virgin olive oil into the grocery bag, without scanning it at all.

Stores today are knitting AI-powered systems throughout checkout to recognize behavioral patterns; compare scanned items with video footage and weight data; and flag inconsistencies.

"Some of the loss at self-checkout can be accidental—a shopper selects regular bananas instead of the organic bananas—and that is where technology, including AI, is helping identify those missteps and correct them," says Doug Baker, vice president, industry relations at FMI—The Food Industry Association. "At the same time, self-checkout can also be a target for theft and is an area where asset protection professionals are really doubling down and rethinking their strategies, as well as seeking technology solutions."

Investing in technology for the sake of loss prevention isn't new to grocery. But the sector lags behind retailers in industries such as apparel and technology, where innovations like Radio Frequency Identification (RFID) got widely embraced. For retailers with tidy inventories of high-priced goods, RFID made good sense. Not so much, however, for attaching relatively expensive technology-laden tags to every can of soup in a grocery store.

According to Baker, 22% of food retailers that FMI surveyed are using AI for loss prevention—up from 10% in 2023. He says he anticipates investments to climb as AI becomes more sophisticated and tested.

Read on for more