Flock's License Plate Readers Are Flagging Innocent Drivers
Faulty AI matching leads to wrongful police targeting of civilians.
1How the System Works
Flock's automated license plate recognition (ALPR) systems are used by police departments across the US. Cameras scan plates and cross-check them against stolen vehicle databases. However, the system's partial matching algorithm can falsely flag innocent vehicles with similar plate structures.
2Case of Mistaken Identity
A writer driving a Jaguar Land Rover press fleet vehicle was wrongly flagged as stolen. The system read '34 10 DTM' as '34 DTM'. All JLR vehicles with similar plate structures nationwide became police targets. The incident exposed fundamental flaws in the system's design.
3Flock's Defense
The company claimed systems were 'working correctly' and blamed police procedures. CEO Thomas stated, 'The AI simply does what's asked: it found 34 DTM when asked for 34 DTM.' He admitted partial matches were implemented per law enforcement requests, with no exact match verification.
4Privacy Concerns
Flock's CEO previously called privacy advocates 'terrorists' but has since apologized. Data shows cameras are used for human searches like 'heavy-set male with black/white hat,' not just vehicles. Race and political affiliation-based searches raise civil liberties concerns.
5System's Prevalence
Over 2,000 police departments across 40+ states use Flock systems. Four similar-plated vehicles were tracked in Minnesota that same week. JLR is pressuring LAPD to correct the report. However, risks remain for commercial fleets with standardized plate structures.
6Technology and Accountability
The incident reveals critical design flaws in AI-powered surveillance. Public safety tech must incorporate safeguards for innocent civilians. Flock's 'it's not broken, it's designed that way' defense raises ethical questions about law enforcement technology development.
Original article: Schneier on Security — Flock Plaka Tanıma Sistemindeki Hata Masumları Hedef Alıyor