The video above looks at what happens when automated license plate readers get it wrong, and it raises a question every driver should be asking: what happens to you when a camera misreads your plate?
How these systems actually work
An ALPR camera, like the ones made by Flock Safety and similar vendors, does not just take a picture. It runs optical character recognition on every plate that passes, converts that into a text string, and checks it against databases of stolen vehicles, wanted persons, or other “hot list” entries. It also logs the plate, time, and location even when there is no match at all, building a travel history on anyone who drives past.
Why errors happen
OCR is not perfect. A dirty plate, bad lighting, glare, a bent frame, or a similar-looking letter and number can cause a misread. A single wrong character can turn an innocent driver’s plate into a match for a stolen vehicle or a plate tied to a warrant. Because these systems operate at highway speed and high volume, even a small error rate adds up to a meaningful number of false alerts every single day across a network of cameras.
The real-world stakes
A misread is not just a technical glitch. It can mean officers approaching a car with guns drawn because the system flagged it as stolen, or a driver detained and questioned over a match that was never accurate to begin with. These encounters carry real risk, especially when officers are told to treat an automated alert as reliable without independently verifying it before acting. The burden of proving a mistake almost always falls on the driver, not the agency running the camera.
What accountability should look like
Departments that use ALPR technology should require human verification of any match before a stop is made, keep clear audit logs of who accessed the data and why, and set short retention limits so years of travel history are not sitting in a database indefinitely. Communities deserve to know where these cameras are, who can access the data, and what happens when the system gets it wrong.
What you can do
Start by finding out what is actually deployed near you on the map, learn the basics of how this technology works on our learn page, and if you want to push for real oversight in your own community, visit take action for concrete steps.
Independent, ad-free, reader-funded
No ads, no paywalls, no selling your data. The costs are server bills, the data services behind the map, the Docket and the archives, and the computing power behind the tools being built on top of them. Every dollar readers chip in goes to exactly that.
Support us on PayPal or scan the code, or Venmo @deflocktheusa