FLOCK MISREAD ONE NUMBER ON HIS LICENSE PLATE. HE ENDED UP BITTEN BY A POLICE K-9 AND IN JAIL.

Our Take

Credit to Manifestation Academy for surfacing this case out of Toledo, because it's the clearest proof yet that "the algorithm said so" is not probable cause. A Flock camera flagged a "7" as a "2," software spit out a stolen-plate hit, and officers treated that mismatch as gospel instead of a tip to be verified. The result wasn't a paperwork correction — it was a man pulled over at gunpoint, bitten by a police dog, hospitalized, and booked into jail for a crime that existed only in a database error. Charges were dropped, an officer was reprimanded for skipping the basic step of actually looking at the plate, and the city paid $35,000 to make it go away. None of that undoes what happened to Brandon Upchurch.

This is the core danger we keep flagging: these networks are marketed as infallible force multipliers, but they're glorified OCR running on cheap cameras at highway speed, and OCR gets things wrong constantly. The technology doesn't just assist policing — in practice it dictates it, with human judgment reduced to an afterthought until someone's already in cuffs or in an ambulance. One misread character shouldn't be enough to justify drawing a weapon on a commuter, let alone releasing a dog on him. When the system is built to generate instant alerts and officers are trained to trust them, the mistake gets a gun pointed at it before anyone bothers to check.

Toledo isn't unique — these cameras are spreading into neighborhoods everywhere, often with zero public debate about the error rate or the consequences when they're wrong. Check our camera map to see what's already watching your routes, and head to our take-action page if you want to push back before your own town's "misread" makes the news.