Rite-Aid’s ambitious rollout of facial recognition technology serves as a cautionary tale for insurance companies exploring AI innovations. The mantra, "Think Big, Start Small, Learn Fast," is often ignored, as seen in Rite-Aid’s approach of "Think Big, Start Big." This resulted in a costly failure. The technology, plagued with false positives and racial bias, inaccurately identified individuals leading to legal issues and customer dissatisfaction. Notably, Rite-Aid faced challenges due to the technology’s inability to accurately recognize faces of Black people and Latinos, leading to numerous false matches and wrongful accusations.
Rite-Aid implemented this technology broadly in 2012 across hundreds of stores without a controlled pilot. The decision, driven by the need to tackle shoplifting and the lure of breakthrough technology, turned sour. The system generated thousands of erroneous alerts, even mistaking a Black customer for a White woman, which drew the attention of the Federal Trade Commission. These errors highlight the inherent risks of adopting untested technology on a large scale.
In a recent settlement with the FTC, Rite-Aid agreed to a five-year ban on using facial recognition technology. The company, already struggling with financial losses and opioid-related lawsuits, filed for bankruptcy in October. This case underscores the importance of cautious and scaled innovation for insurance companies considering generative AI. The lesson is clear: while it’s essential to think big, starting small is crucial to avoid costly and legal pitfalls.