Retailers nationwide are using invasive artificial intelligence to combat crime. That’s concerning for multiple reasons.

An alarming new report from Forbes sheds light on the retail industry’s “big play” to utilize artificial intelligence to combat crime. These surveillance systems, capable of automatically identifying people, license plates, and other identifying information, are already being used by more than a dozen retailers. Aside from serious concerns about privacy, people are already being wrongly arrested due to the technology, which is often inaccurate when identifying women of color (one technology has an error rate of up to 35 percent when scanning “darker-skinned women,” according to a research report). The current push behind the technology is related to rising panic about crime, with local and national headlines warning of soaring shoplifting rates and “smash and grab” robberies. According to the National Retail Federation, organized retail crime has risen by 60 percent since 2015, with nearly 70 percent of retailers reporting an increase in 2021. But as The Atlantic reported in December, it’s hard to figure out if the surge in theft is real. Not only does much of the recent reporting conflate an array of very different offenses into a single crime wave, but the crime narratives perpetuated by large corporations and law enforcement officials don’t always gel with the data and could lead to more “harmful and discriminatory carceral policies,” according to The Appeal. Meanwhile, Midwest grocery store chain Hy-Vee has taken a different approach: an in-house, armed security team. —Tina Vasquez