Retailers turn to facial recognition to curb $100B in losses amid ‘sweethearting’ surge

Retail theft is on the rise in the U.S., with losses topping an estimated $100 billion annually. In addition to shoplifting and employee theft, a growing challenge is “sweethearting” — a form of insider fraud where employees covertly grant unauthorized discounts or free items to friends or family members. This type of theft often goes undetected, as employees exploit their access to systems by bypassing item scans, issuing false refunds, or misusing employee discounts.

To combat this form of fraud, Corsight AI has developed a facial recognition-based sweethearting module designed to identify and prevent sweethearting incidents.

Corsight’s global VP of sales, Gal Biton, explains that the facial intelligence platform analyzes employee-customer interactions, identifying unusual patterns, flagging frequent and suspicious contacts for further review by security teams.

“These flagged interactions often reveal sweethearting cases where customers, in cooperation with cashiers, leave with unpaid items,” Biton adds.

Corsight’s sweethearting module is part of a broader AI-driven retail security suite, which also includes features to track individuals on watchlists, detect unauthorized access to restricted areas, and monitor loitering or other behaviors associated with theft. The suite aims to tackle the multiple facets of retail crime, a growing concern not only in the U.S. but also in other regions where retailers are trialing similar technology.

In fact, SIA reports recent research shows that senior retail executives consider facial recognition technology a crucial tool in combating theft. A survey conducted by the UK-based firm Opinion Matters among 300 senior retail leaders in the U.S. and the UK found that at least 15 of the top 50 grocery stores in the U.S. have adopted facial recognition technology.

Aside from Corsight’s technology, Facewatch, a cloud-based facial recognition system, is gaining traction as it compares real-time images with a pre-existing database, in a bid to identify potential matches with individuals linked to previous incidents of theft or antisocial behavior. When a match is detected, it instantly alerts retail staff, allowing them to respond swiftly in-store, Newelectronics highlights.

Global facial recognition use in retail

A pilot in New Zealand grocery stores has shown promising early results in identifying and curbing theft through facial recognition, providing a potential model for broader adoption. In the U.K., meanwhile, lawmakers are actively debating the use of facial recognition as a viable tool for the retail sector, where theft levels have also risen. However, the technology’s spread isn’t without its challenges; in New York City, a bill is under consideration that would restrict the use of facial recognition in grocery stores, raising concerns about privacy and surveillance in public spaces.

Grocery store owners in New York City are pushing back against a proposed bill that would restrict the use of facial recognition software. Additionally, the city council introduced a measure requiring businesses to notify customers and obtain written consent before using biometric recognition, with noncompliance resulting in a $5,000 fine, Supermarket News reports.

Article Topics

biometric identification  |  biometrics  |  Corsight  |  Facewatch  |  facial recognition  |  retail biometrics  |  Security Industry Association (SIA)