Share

Headlines last week screamed that Florida had “banned” Flock’s spy cameras. Governor Ron DeSantis called the license plate reader network “out of control,” and the Florida Department of Transportation (FDOT) issued a memo that seemed to pull the plug. But the reality is far narrower—and leaves most of the system untouched.
What the memo actually does
FDOT’s Engineering and Operations Memorandum No. 26r-01 revokes the general use permits it had granted to local police for automated license plate readers (ALPRs) mounted on state highway right-of-way. Agencies have 30 days to remove those cameras at their own cost, or FDOT will take them down and stop approving new ones. That’s it. The order applies only to the strip of state-owned land beside the pavement, not to the technology itself.
The legal limit that made the move possible
FDOT’s authority comes from a single statute that governs placement within the state highway system—not the cameras themselves. It’s a real estate rule, not a technology ban. Cities, counties, HOAs, shopping centers, and private toll operators can still run Flock cameras anywhere else. The memo doesn’t touch the vast majority of Flock’s national network, which covers neighborhood entrances, county roads, downtown intersections, and parking lots.
The real problem: accuracy, not location
While the location fight makes for a good press release, the deeper issue remains: false positives that can turn a routine drive into an armed felony stop. In June 2024, automotive journalist Joel Feder was pulled over in suburban Minnesota by four squad cars, guns drawn, while driving a $155,000 Range Rover press loaner. Police had been tracking the SUV for days after a Flock camera flagged it as carrying a stolen plate. The plate wasn’t stolen—a Los Angeles dealership had reported a lost manufacturer plate using its full six-character format, but someone entering the data dropped the small middle digits that New Jersey uses on dealer tags. The truncated match linked every other Jaguar Land Rover press car with the same numbering pattern. Feder wasn’t the only one flagged that week; he was just the first caught.
That case highlights a systemic flaw: dealer, manufacturer, transporter, and temporary tags use non-standard layouts that challenge optical character recognition (OCR) software. A single data-entry error two thousand miles away can trigger a false alert across dozens of departments. Flock has acknowledged the pattern. In August 2024, the company cut its default data retention from 30 days to seven, made misuse-detection mandatory for police customers, and allowed cities to block searches tied to immigration enforcement. But none of those changes fix the OCR misreading of non-standard characters. That’s an engineering problem, not a policy one.
What the governor and lawmakers are doing now
DeSantis has called for a statewide AI bill of rights, saying the real fix must come from the Legislature, not an agency memo. In February 2025, Florida lawmakers introduced a bipartisan bill (HB 1234) that would require annual audits of ALPR databases, mandate civil penalties for unauthorized searches, and create a statewide oversight board. The bill is still in committee, but it signals growing bipartisan concern. At the federal level, Senator Josh Hawley opened his own inquiry into the size of Flock’s network earlier this month. Neither effort addresses the root accuracy issue—how a plate is read in the first place.
The bottom line for drivers
For anyone driving on paper tags, dealer plates, or manufacturer press fleet vehicles, the safest assumption remains: somewhere in the country, a camera network you’ve never heard of already has an opinion about your car. Whether that opinion is correct is still mostly a matter of luck. The Florida memo didn’t turn off the network—it just moved a few cameras off state property. The broader surveillance ecosystem and its false-positive risk remain unchanged.
Join our newsletter, follow us on Instagram, and connect on Facebook for the latest on privacy and surveillance.









