Editorial: AI could help with retail theft. But at what cost?

It’s no secret that retail theft has skyrocketed since the pandemic. We’ve all seen the alarming videos of criminal gangs attacking stores, cleaning out shelves and walking off with no apparent consequences.

This year, U.S. merchants expect to report a stunning $100 billion in losses — a big increase over pre-pandemic levels. At the same time, arrests and prosecutions for retail theft in Chicago are down sharply since 2019.

At this month’s convention for Oak Brook-based Ace Hardware, exhibitor Laura Freeman of Watcher Total Protection had talked herself hoarse by the end of the opening day. Aggrieved store owners crowded her booth to quiz her about her company’s security systems, which are augmented with artificial intelligence.

AI is coming to the rescue of merchants, for better or worse. “We will see a lot more AI for shoplifter protection,” said Freeman, whose company is based in Pingree Grove. “You’ll see more and more where the system does the work.”

The old days of dumb cameras recording thieves in action for review after the fact have given way to smart systems that can detect illegal activity as it happens and send instant text alerts. The surveillance cameras constantly scan for a shopper taking something off a shelf and shoving it into a bag, for instance, which immediately triggers a warning so merchants can decide whether to question the suspect, call for help or otherwise intervene.

In the meantime, tags attached to high-value items not only trigger sensors on the way out of a store but also can be coded to the exact item and linked to video of the exact moment that item was removed from its spot in the store. As a result, merchants can spend much less time reviewing footage to zero in on the part they need.

Unfortunately, theft by retail employees is way up too. Store cashiers increasingly are being watched with smart systems that analyze their actions in real time and coordinate with data from cash registers to flag instances when mistakes are being made or theft is occurring. AI is being used to spot employees who ring up one-penny sales to open their cash drawers, or customers at those now-ubiquitous self-checkouts who try the old “banana trick,” scanning cheap items like the fruit while putting expensive items in their bags.

Some retailers with stores in high-crime locations have tried equipping their clerks with the same body cameras worn by police, mostly to deter violent confrontations that, sadly, also have become more common since the pandemic. And facial recognition makes it possible for camera systems to flag likely suspects as they enter a store so they can be tracked, based on their prior criminal records, or previous incidents of theft.

It’s amazing stuff, for sure, and three cheers for companies like Freeman’s that are helping retailers fight back against thieves. Still, this emerging technology is starting to make a visit to the friendly neighborhood store sound like submitting to Big Brother on steroids.

We realize that retailers need new tools to combat the epidemic of lawlessness costing them huge losses. And, similarly, we know police and prosecutors have a backlog of violent crimes to pursue. While we support stronger laws and enforcement to protect businesses under siege, we recognize that no one is going to rush over with lights and sirens if a shopper takes a steak while ringing up a banana at checkout.

As retailers adopt these potentially game-changing but also intrusive systems, there must be checks and balances.

Indeed, the rise in retail theft has led to much faster adoption of emerging technologies than might otherwise have occurred, according to Matt Harper, loss prevention manager at Ace. “This is going to be a necessity,” he said, especially as big-box stores install technology that makes mom-and-pop operators look like relatively soft targets for thieves. It’s no exaggeration to say an AI-enhanced arms race is underway in the retail biz.

Here’s a pleasant surprise: Illinois is ahead of the curve in dealing with the concerns these anti-theft systems raise about privacy. The 2008 Illinois Biometric Information Privacy Act regulates scans of people’s faces, eyes and fingerprints, which might have sounded like sci-fi when the law was first considered, but certainly no longer.

This page supported the law, and the pace of technological advances makes it even more important today to set clear guidelines for the ways that businesses use data.

The Illinois statute came after incidents in which Big Tech companies were failing to self-police. And while we prefer for businesses to regulate themselves, government enforcement is needed because of the high stakes involved: Consumers can always get a new credit card after a privacy breach, but fingerprints, eyes and faces can’t be reissued.

We don’t get to say this very often, but we wish more states would wake up and follow Illinois’ example. As of now, for national retailers like Ace, the landscape is messy and complex, with some states having no regulations in place and others having rules that differ from one to another. Illinois’ law may need updating, given the rapid advances afoot, but it is a worthy starting point for states doing nothing while technology runs rings around them.

Retail theft is now at epidemic levels, and AI can help. Let’s make sure we safeguard our privacy at the same time we embrace technology that helps us get a step ahead of the crooks.

_______