Beyond the persistent debates over privacy violations, the accuracy and practicality of Flock’s cloud-based Automatic License Plate Recognition (ALPR) system are now under serious scrutiny. An internal audit conducted by the Roseville, California, police department uncovered a troubling number of false positives generated by the system. Take, for example, the case of American automotive journalist Joel Feder: while test-driving a Land Rover, he found himself surrounded by four police cruisers—a dramatic escalation triggered by errors in both Flock’s AI-powered license plate recognition and subsequent police enforcement procedures. The incident ultimately proved to be a false alarm, stemming from incomplete stolen license plate data entered by California law enforcement. Flock’s AI, which failed to cross-reference the full plate number, sounded the alarm, and the responding officer neglected to verify the complete license plate photo before taking action.

This isn’t an isolated issue. In Sumter County, Florida, a detective faced arrest and termination for misusing the system to access personal information without authorization. While Flock maintains that the system functions as designed, the sheer scale of its operations—processing 20 billion license plate scans monthly—means even a 1% error rate could result in 200 million misreads. Such inaccuracies carry real-world consequences, risking dangerous or unnecessary law enforcement interventions.

As Flock expands its drone-based surveillance operations, its monitoring capabilities now stretch from street-level to aerial coverage, deepening public anxiety over privacy breaches and the potential for surveillance overreach.