Flock Cameras: 1 Billion Wrong Scans a Month.
The short version: Systematic Audit crunches Flock's own numbers to show that even a 93% accuracy rate on 20 billion monthly scans means roughly a billion errors a month.
Our Take
Credit to Systematic Audit for doing the math Flock Safety would rather you not do. The company loves to tout a 93% accuracy figure like it's a gold star, but run that percentage against 20 billion scans a month and you get an almost unfathomable volume of bad reads — misidentified plates, wrong states, garbled characters — all feeding into databases that cops, insurers, and repo men can query. A 7% error rate sounds small until you realize it's happening at a scale most people can't picture.
This is the core problem with ALPR systems: they're marketed as precision tools but deployed as dragnets. Every wrong match is a potential wrong stop, a wrongful detention, or a false flag that follows an innocent driver's plate around forever. And none of this requires a warrant, a judge, or even a crime to have occurred — the camera just has to be pointed at the road.
If a billion errors a month doesn't bother you, consider that we still don't know who's reviewing those mismatches or how long bad data lingers in these systems. Check our map to see how many of these cameras are already watching your own neighborhood, then head to our take-action page to push back before the next billion scans roll in.
This is DeFlock The USA’s original commentary. The video above is the work of Systematic Audit, published on YouTube — full credit to the creator.