Corsight Ai has developed a new person recognition system designed for retail stores, which is capable of identifying abnormal interactions between employees and clients. The tool was introduced approximately a month ago in an effort to combat “Sweethearting” cases, where employees provide discounts or free goods to friends (source).
What sets this system apart from traditional solutions is its ability to not only identify individuals on blacklists but also analyze the frequency and proximity of customer interactions with specific employees. For instance, if a client consistently approaches the same cashier, it may raise suspicions of Sweethearting. Upon detecting anomalies, the system alerts the store’s security team.
Shai Toren, Director of Corsight, highlighted that repeated customer interactions with a single employee could indicate potentially suspicious activities. Such behavior deviates from the norm where customers typically choose the nearest available cashier.
However, trade union representatives have voiced concerns that the system might unjustly interpret customer loyalty to an employee as a violation indicator. Chelsey Connor, Communications Director of the RWDSU union, emphasized that fostering customer relations to enhance sales should not be misconstrued as suspicious.
While some of Corsight’s customers have already integrated this technology, the company has refrained from disclosing specific networks utilizing the system. In recent years, major retail chains have been increasingly adopting video surveillance tools and algorithmic analyses to counter theft. According to the National Retail Federation, Sweethearting accounts for up to 29% of inventory shrinkage in the United States.
Despite Corsight’s claims that monitoring interactions is just the initial phase, critics like Caitlin Sili George from Fight for the Future have highlighted potential risks for both staff and customers. There are concerns that companies may exploit theft threats to introduce tools that could further analyze consumer behavior for marketing purposes.
Corsight has indicated that monitoring interactions represents only the beginning, with plans to develop technologies that can track customer and employee behavior with even more precision in the future.