This video examines how U.S. police departments misuse automated systems — facial recognition, gunshot-detection tools like ShotSpotter, and license-plate readers like Flock — by treating probabilistic outputs as verified fact. It details Robert Williams’ wrongful Detroit arrest after a flawed facial-recognition match, and Michael Williams’ nearly year-long Chicago jailing based on misapplied ShotSpotter acoustic data, both ending in costly settlements. The video argues these tools are meant to generate leads, not proof, yet get “laundered” into seemingly independent evidence like witness lineups. It also covers Flock cameras’ mission creep into routine traffic enforcement, burdening taxpayers twice — funding surveillance and paying resulting tickets — while doing little to reduce serious crime, and urges viewers to push back against unchecked deployment.