Could One AI Match Steal Your Freedom?
The short version: CaraComp's rundown of a 17-month wrongful imprisonment shows how a single AI match can become 'probable cause' in the eyes of police — the same shortcut Flock's ALPR network invites.
Our Take
CaraComp's breakdown of the St. Louis facial recognition case lands on a truth we harp on constantly: once an algorithm spits out a name or a plate number, cops and prosecutors tend to treat it as an answer, not a hint. A machine says 'match' with some confidence score, and suddenly that probability gets laundered into investigative momentum, an arrest, and — in this case — 17 months in a cell before anyone stopped to ask if the software was actually right.
Swap 'face' for 'license plate' and you're describing exactly how Flock Safety's ALPR network gets used every day. A camera flags a plate as 'stolen' or associated with a BOLO, an officer treats that flag as probable cause, and the human review that's supposed to catch false positives often never happens. The video's point about the EU AI Act's paperwork compliance not equaling real protection applies just as well here — a vendor can publish a transparency report while cops in the field are still pulling guns on innocent drivers over a bad match.
This is why we push people to know what's watching their own streets. Check our camera map to see where Flock has planted its eyes in your area, and if you want to push back on unaccountable ALPR deployment in your city, our take-action page has the tools. Credit to CaraComp for connecting AI hype to the very real people who pay for its errors.
This is DeFlock The USA’s original commentary. The video above is the work of CaraComp, published on YouTube — full credit to the creator.