Louisiana police used facial recognition technology to falsely identify and arrest Randal Reid, a Black man from Georgia, for purse thefts he did not commit, leading to his imprisonment for nearly a week before the error was discovered.
In November 2022, Louisiana law enforcement agencies used facial recognition technology to incorrectly identify Randal Reid, a 28-year-old Black man from Georgia, as a suspect in luxury purse thefts totaling over $10,000 in Jefferson Parish and Baton Rouge. The Jefferson Parish Sheriff's Office used facial recognition analysis through the Louisiana State Analytic and Fusion Exchange, which employs Clearview AI and MorphoTrak systems, to identify Reid in connection with a June 2022 theft of Chanel and Louis Vuitton purses. A Baton Rouge Police Department detective then adopted this identification to secure a second arrest warrant for a similar theft. Reid was arrested on November 25, 2022, while driving to a Thanksgiving celebration in Georgia and was held in DeKalb County jail until December 1. The facial recognition match was ultimately determined to be false when significant physical differences were noted, including a 40-pound weight difference and a mole on Reid's face that the actual suspect did not have. Reid's attorney stated that Jefferson Parish detectives 'tacitly' admitted the error and rescinded the warrant. The case highlights ongoing concerns about facial recognition technology's higher error rates for Black individuals and its use in law enforcement without adequate safeguards.
Domain classification, causal taxonomy, severity scores, and national security assessments were LLM-classified and may contain errors.
Accuracy and effectiveness of AI decisions and actions are dependent on group membership, where decisions in AI system design and biased training data lead to unequal outcomes, reduced benefits, increased effort, and alienation of users.
AI system
Due to a decision or action made by an AI system
Unintentional
Due to an unexpected outcome from pursuing a goal
Post-deployment
Occurring after the AI model has been trained and deployed