A Fort Myers man was arrested in a child-abduction case after Florida police relied on a facial recognition system that flagged him as a "93% match" — and the ACLU says officers never looked much further than that.
The civil liberties organization has filed suit against two Florida police departments, alleging they treated a flawed algorithmic result as a near-certain identification rather than as a starting point for actual investigation. According to Wired, the tool at the center of the case is one of the oldest police face-recognition systems in use in the United States.
The lawsuit argues the departments compounded the technology's shortcomings by ignoring other evidence that could have cleared the man. As Ars Technica reports, the complaint charges that "police let an error-prone AI system stand in for an investigation."
Facial recognition errors have disproportionately affected Black men in previous documented wrongful-arrest cases across the country. The technology's accuracy varies widely depending on the system, the image quality, and the demographic of the person being matched — factors that critics say are rarely disclosed to juries or defendants.
The case is significant because it tests legal accountability for one of law enforcement's most widely adopted but least regulated AI tools, and could set a precedent for how much police departments can rely on algorithmic matches before they are required to corroborate them with human investigation.