Facial Recognition in Policing: Efficiency, Error, and Racial Justice

Our Take

Credit to Ahrielle Barnett-Fitch for laying out what should be an uncomfortable truth: facial recognition gets sold to police departments as a neutral efficiency upgrade, right up until it isn't. Wrongful arrests, disproportionate error rates for Black and brown faces, and a near-total absence of public oversight aren't bugs in the system — they're the predictable result of deploying opaque tech first and asking questions never.

We cover license plate readers, not face scanners, but the pattern in this video is identical to what we track every day with Flock Safety. A private vendor builds a database, sells it to hundreds of police agencies as a 'force multiplier,' and the public finds out later — usually through a lawsuit or a journalist — how the thing actually gets used. No community vote. No independent audit. Just cameras, watchlists, and a growing appetite among departments to plug every surveillance stream into the same pipeline. Facial recognition and ALPR networks increasingly feed each other, and that convergence is exactly why piecemeal outrage over one technology at a time isn't enough.

If this video makes you uneasy about who's watching and how little say residents get, check our camera map to see what's already been installed near you, and hit take-action for concrete steps to push back before your city signs the next contract.