LAPD's Own Watchdog: 161 Innocent Drivers Pulled Over by Plate Cameras #substack #shorts

Our Take

Credit to PickBits for surfacing this one, because it's the kind of number that should end the debate over whether ALPR networks are 'just a tool.' One hundred sixty-one innocent drivers pulled over in two months, and every single plate read was accurate. The system didn't fail on recognition. It failed on data hygiene. Other agencies recovered stolen cars and never bothered to clear the record, so LAPD's cameras kept flagging the same plates as hot months or years later. That's not a technical glitch, that's a structural feature of a shared, under-governed database that nobody owns the accuracy of.

This is the part Flock's marketing never leads with: these hot lists are crowdsourced across hundreds of agencies with wildly different habits about updating them, and the officer who pulls you over has no way to know the alert is stale until you're already staring down a drawn weapon over a car that was recovered a year ago. A one-in-three error rate on felony stop alerts isn't a rounding error, it's a coin flip on whether the next person getting pulled out of their car at gunpoint did anything wrong. LAPD hadn't even audited this system since 2022, which tells you how little oversight these networks get once they're installed.

The fact that LAPD unplugged 138 cameras after its own watchdog found this should be the template, not the exception. Every department running Flock or a similar ALPR network should be forced to publish hit/miss audits on this schedule, not wait three years for an inspector general to stumble onto it. Check our camera map to see what's deployed near you, and if your local agency hasn't audited its hot-list accuracy, our take-action page has steps for demanding one.