The AI-Generated Pattern Hides You From Surveillance Cameras—Including Flock
Bill Swearingen, a Kansas City security researcher, says he has built patterns that can hide vehicles from Flock camera detection systems. After millions of tests, his project, noRecognition, uses reinforcement learning to generate wraps that still look normal to people but disrupt the object-detection layer that identifies cars and plates. He demonstrated it at Def Con by wrapping a Toyota Yaris and driving past a Flock camera; the footage remained visible, but the AI failed to log the vehicle. He frames the work as a privacy tool to help people avoid surveillance, especially at protests. The project targets widely deployed detection stacks, with Flock drawing particular scrutiny amid political backlash and concerns about mass plate scanning. Swearingen is crowdfunding early merchandise and future vehicle skins, though legal issues remain because plate-obstruction laws vary by state.
