AI Gets Wrong Woman Arrested 1,200 Miles Away From Crime

Our Take

Steve Lehto lays out a case that should terrify anyone who thinks these systems only ever flag the guilty: a facial recognition hit from 1,200 miles away led to an innocent woman's arrest, and by the time the dust settled she'd lost her home. This is the pattern we keep seeing with automated policing tools — a machine spits out a match, an officer treats it as gospel, and due process becomes an afterthought. Nobody asks whether the algorithm actually understood what it was looking at; they just run with the alert.

Facial recognition and ALPR networks like Flock Safety operate on the same flawed premise: that pattern-matching software is reliable enough to justify treating someone as a suspect without real corroboration. Flock's cameras are logging plates and vehicle 'fingerprints' across thousands of communities right now, feeding a searchable database that police departments increasingly lean on the same way this SD program was leaned on. When the tech is wrong — and it will be wrong — the burden falls entirely on the innocent person to prove it, often after their life has already been upended.

Credit to Lehto for walking through the legal fallout in plain terms. If you want to see how far this surveillance infrastructure has spread into your own neighborhood, check our camera map, and if you think your city council should be asking harder questions before signing another contract, our take-action page has a place to start.