We Made a Flock-Proof Car
The short version: Donut's experiment with adversarial-noise wraps against Flock cameras is a fun stunt, but it underscores how normalized ALPR surveillance has become—and how little recourse drivers actually have.
Our Take
Donut and Benn Jordan built a car wrapped in adversarial noise patterns designed to confuse AI object detection, then drove it past an actual Flock camera to see if the tech could dodge automated recognition. It's a genuinely fun watch, and credit to the crew (plus Bill Swearingen's anti-surveillance apparel work) for putting real engineering behind the idea instead of just talking about it. But the bigger story here isn't whether adversarial wraps can beat one camera on one day — it's that private companies have blanketed public roads with AI-powered tracking infrastructure thorough enough that people feel the need to armor their vehicles like it's a spy movie.
That's the part worth sitting with. Flock cameras aren't just reading plates anymore; they're logging vehicle make, color, damage, bumper stickers, and roof racks into a searchable database that police departments across the country query without a warrant. A wrap that jams one detector today doesn't change the fact that these systems are proliferating faster than any meaningful oversight, and most drivers have no idea they're being cataloged block by block. Stunts like this are a clever thought experiment, but the real fix isn't camouflage — it's forcing transparency and accountability onto the agencies and companies deploying this stuff at scale.
If you want to know what's actually watching your street, check our camera map to see documented ALPR locations near you, and if you're ready to push back on this infrastructure in your own community, our take-action page has concrete steps for doing exactly that.
This is DeFlock The USA’s original commentary. The video above is the work of Donut, published on YouTube — full credit to the creator.